Medicare Program; Revisions to Payment Policies Under the Physician Fee Schedule and Other Revisions to Part B for CY 2017; Medicare Advantage Bid Pricing Data Release; Medicare Advantage and Part D Medical Loss Ratio Data Release; Medicare Advantage Provider Network Requirements; Expansion of Medicare Diabetes Prevention Program Model; Medicare Shared Savings Program Requirements

AGENCY:

Centers for Medicare & Medicaid Services (CMS), HHS.

ACTION:

Final rule.

SUMMARY:

This major final rule addresses changes to the physician fee schedule and other Medicare Part B payment policies, such as changes to the Value Modifier, to ensure that our payment systems are updated to reflect changes in medical practice and the relative value of services, as well as changes in the statute. This final rule also includes changes related to the Medicare Shared Savings Program, requirements for Medicare Advantage Provider Networks, and provides for the release of certain pricing data from Medicare Advantage bids and of data from medical loss ratio reports submitted by Medicare health and drug plans. In addition, this final rule expands the Medicare Diabetes Prevention Program model.

DATES:

These regulations are effective on January 1, 2017.

FOR FURTHER INFORMATION CONTACT:

Jessica Bruton, (410) 786-5991, for issues related to identification of potentially misvalued services and any physician payment issues not identified below.

Gail Addis, (410) 786-4522, for issues related to diabetes self-management training.

Jaime Hermansen, (410) 786-2064, for issues related to moderate sedation coding and anesthesia services.

Roberta Epps, (410) 786-4503, for issues related to PAMA section 218(a) policy and the transition from traditional x-ray imaging to digital radiography.

Ann Marshall, (410) 786-3059, for primary care issues related to chronic care management (CCM), burden reduction, telehealth services and evaluation and management services.

Emily Yoder, (410) 786-1804, for issues related to resource intensive services, telehealth services and other primary care issues.

Lindsey Baldwin, (410) 786-1694, for primary care issues related to behavioral health integration services.

Geri Mondowney, (410) 786-4584, and Donta Henson, (410) 786-1947, for issues related to geographic practice cost indices.

Michael Soracoe, (410) 786-6312, for issues related to the target and phase-in provisions, the practice expense methodology, impacts, conversion factor, and the valuation of pathology and surgical procedures.

Pamela West, (410) 786-2302, for issues related to therapy.

Patrick Sartini, (410) 786-9252, for issues related to malpractice RVUs, radiation treatment, mammography and other imaging services.

Kathy Bryant, (410) 786-3448, for issues related to collecting data on resources used in furnishing global services.

Donta Henson, (410) 786-1947, for issues related to ophthalmology services.

Corinne Axelrod, (410) 786-5620, for issues related to rural health clinics or federally qualified health centers.

Simone Dennis, (410) 786-8409, for issues related to FQHC-specific market basket.

JoAnna Baldwin, (410) 786-7205, or Sarah Fulton, (410) 786-2749, for issues related to appropriate use criteria for advanced diagnostic imaging services.

Robin Usi, (410) 786-0364, for issues related to open payments.

Sean O'Grady, (410) 786-2259, or Julie Uebersax, (410) 786-9284, for issues related to release of pricing data from Medicare Advantage bids and release of medical loss ratio data submitted by Medicare Advantage organizations and Part D sponsors.

Sara Vitolo, (410) 786-5714, for issues related to prohibition on billing qualified Medicare beneficiary individuals for Medicare cost-sharing.

Michelle Peterman, (410) 786-2591, for issues related to Accountable Care Organization (ACO) participants who report PQRS quality measures separately.

Katie Mucklow, (410) 786-0537 or John Spiegel, (410) 786-1909, for issues related to Provider Enrollment Medicare Advantage Program.

Jen Zhu, (410) 786-3725, Carlye Burd, (410) 786-1972, or Nina Brown, (410) 786-6103, for issues related to Medicare Diabetes Prevention Program model expansion.

Rabia Khan or Terri Postma, (410) 786-8084 or ACO@cms.hhs.gov, for issues related to the Medicare Shared Savings Program.

Kimberly Spalding Bush, (410) 786-3232, or Fiona Larbi, (410) 786-7224, for issues related to Value-based Payment Modifier and Physician Feedback Program.

Lisa Ohrin Wilson, (410) 786-8852, or Gabriel Scott, (410) 786-3928, for issues related to physician self-referral updates.

SUPPLEMENTARY INFORMATION:

Table of Contents

I. Executive Summary and Background

A. Executive Summary

B. Background

II. Provisions of the Final Rule for PFS

A. Determination of Practice Expense Relative Value Units (PE RVUs)

B. Determination of Malpractice Relative Value Units (MRVUs)

C. Medicare Telehealth Services

D. Potentially Misvalued Services Under the Physician Fee Schedule

1. Background

2. Progress in Identifying and Reviewing Potentially Misvalued Codes

3. Validating RVUs of Potentially Misvalued Codes

4. CY 2017 Identification and Review of Potentially Misvalued Services

5. Valuing Services That Include Moderate Sedation as an Inherent Part of Furnishing the Procedure

6. Collecting Data on Resources Used in Furnishing Global Services

E. Improving Payment Accuracy for Primary Care, Care Management Services, and Patient-Centered Services

F. Improving Payment Accuracy for Services: Diabetes Self-Management Training (DSMT)

G. Target for Relative Value Adjustments for Misvalued Services

H. Phase-In of Significant RVU Reductions

I. Geographic Practice Cost Indices (GPCIs)

J. Payment Incentive for the Transition From Traditional X-Ray Imaging to Digital Radiography and Other Imaging Services

K. Procedures Subject to the Multiple Procedure Payment Reduction (MPPR) and the OPPS Cap

L. Valuation of Specific Codes

M. Therapy Caps

III. Other Provisions of the Final Rule for PFS

A. Chronic Care Management (CCM) and Transitional Care Management (TCM) Supervision Requirements in Rural Health Clinics (RHCs) and Federally Qualified Health Centers (FQHCs)

B. FQHC-Specific Market Basket

C. Appropriate Use Criteria for Advanced Diagnostic Imaging Services

D. Reports of Payments or Other Transfers of Value to Covered Recipients: Summary of Public Comments

E. Release of Part C Medicare Advantage Bid Pricing Data and Part C and Part D Medical Loss Ratio (MLR) Data

F. Prohibition on Billing Qualified Medicare Beneficiary Individuals for Medicare Cost-Sharing

G. Recoupment or Offset of Payments to Providers Sharing the Same Taxpayer Identification Number

H. Accountable Care Organization (ACO) Participants Who Report Physician Quality Reporting System (PQRS) Quality Measures Separately

I. Medicare Advantage Provider Enrollment

J. Expansion of the Diabetes Prevention Program (DPP) Model

K. Medicare Shared Savings Program

L. Value-Based Payment Modifier and Physician Feedback Program

M. Physician Self-Referral Updates

N. Designated Health Services

IV. Collection of Information Requirements

V. Regulatory Impact Analysis

Regulations Text

Acronyms

In addition, because of the many organizations and terms to which we refer by acronym in this final rule, we are listing these acronyms and their corresponding terms in alphabetical order below:

A1c Hemoglobin A1c

AAA Abdominal aortic aneurysms

ACO Accountable care organization

AMA American Medical Association

ASC Ambulatory surgical center

ATA American Telehealth Association

ATRA American Taxpayer Relief Act (Pub. L. 112-240)

AWV Annual wellness visit

BBA Balanced Budget Act of 1997 (Pub. L. 105-33)

BBRA [Medicare, Medicaid and State Child Health Insurance Program] Balanced Budget Refinement Act of 1999 (Pub. L. 106-113)

BLS Bureau of Labor Statistics

CAD Coronary artery disease

CAH Critical access hospital

CBSA Core-Based Statistical Area

CCM Chronic care management

CEHRT Certified EHR technology

CF Conversion factor

CG-CAHPS Clinician and Group Consumer Assessment of Healthcare Providers and Systems

CLFS Clinical Laboratory Fee Schedule

CoA Certificate of Accreditation

CoC Certificate of Compliance

CoR Certificate of Registration

CNM Certified nurse-midwife

CP Clinical psychologist

CPC Comprehensive Primary Care

CPEP Clinical Practice Expert Panel

CPT [Physicians] Current Procedural Terminology (CPT codes, descriptions and other data only are copyright 2015 American Medical Association. All rights reserved.)

CQM Clinical quality measure

CSW Clinical social worker

CT Computed tomography

CW Certificate of Waiver

CY Calendar year

DFAR Defense Federal Acquisition Regulations

DHS Designated health services

DM Diabetes mellitus

DSMT Diabetes self-management training

eCQM Electronic clinical quality measures

ED Emergency Department

EHR Electronic health record

E/M Evaluation and management

EMT Emergency Medical Technician

EP Eligible professional

eRx Electronic prescribing

ESRD End-stage renal disease

FAR Federal Acquisition Regulations

FDA Food and Drug Administration

FFS Fee-for-service

FQHC Federally qualified health center

FR Federal Register

FSHCAA Federally Supported Health Centers Assistance Act

GAF Geographic adjustment factor

GAO Government Accountability Office

GPCI Geographic practice cost index

GPO Group purchasing organization

GPRO Group practice reporting option

GTR Genetic Testing Registry

HCPCS Healthcare Common Procedure Coding System

HHS [Department of] Health and Human Services

HOPD Hospital outpatient department

HPSA Health professional shortage area

IDTF Independent diagnostic testing facility

IPPE Initial preventive physical exam

IPPS Inpatient Prospective Payment System

IQR Inpatient Quality Reporting

ISO Insurance service office

IT Information technology

IWPUT Intensity of work per unit of time

LCD Local coverage determination

MA Medicare Advantage

MAC Medicare Administrative Contractor

MACRA Medicare Access and CHIP Reauthorization Act of 2015 (Pub. L. 114-10)

MAP Measure Applications Partnership

MAPCP Multi-payer Advanced Primary Care Practice

MAV Measure application validity [process]

MCP Monthly capitation payment

MedPAC Medicare Payment Advisory Commission

MEI Medicare Economic Index

MFP Multi-Factor Productivity

MIPPA Medicare Improvements for Patients and Providers Act (Pub. L. 110-275)

MMA Medicare Prescription Drug, Improvement and Modernization Act of 2003 (Pub. L. 108-173, enacted on December 8, 2003)

MP Malpractice

MPPR Multiple procedure payment reduction

MRA Magnetic resonance angiography

MRI Magnetic resonance imaging

MSA Metropolitan Statistical Areas

MSPB Medicare Spending per Beneficiary

MU Meaningful use

NCD National coverage determination

NCQDIS National Coalition of Quality Diagnostic Imaging Services

NP Nurse practitioner

NPI National Provider Identifier

NPP Nonphysician practitioner

NQS National Quality Strategy

OACT CMS's Office of the Actuary

OBRA ’89 Omnibus Budget Reconciliation Act of 1989 (Pub. L. 101-239)

OBRA ’90 Omnibus Budget Reconciliation Act of 1990 (Pub. L. 101-508)

OES Occupational Employment Statistics

OMB Office of Management and Budget

OPPS Outpatient prospective payment system

OT Occupational therapy

PA Physician assistant

PAMA Protecting Access to Medicare Act of 2014 (Pub. L. 113-93)

PAMPA Patient Access and Medicare Protection Act (Pub. L. 114-115)

PC Professional component

PCIP Primary Care Incentive Payment

PE Practice expense

PE/HR Practice expense per hour

PEAC Practice Expense Advisory Committee

PECOS Provider Enrollment, Chain, and Ownership System

PFS Physician Fee Schedule

PLI Professional Liability Insurance

PMA Premarket approval

PPM Provider-Performed Microscopy

PQRS Physician Quality Reporting System

PPIS Physician Practice Expense Information Survey

PPS Prospective Payment System

PT Physical therapy

PT Proficiency Testing

PT/INR Prothrombin Time/International Normalized Ratio

PY Performance year

QA Quality Assessment

QC Quality Control

QCDR Qualified clinical data registry

QRUR Quality and Resources Use Report

RBRVS Resource-based relative value scale

RFA Regulatory Flexibility Act

RHC Rural health clinic

RIA Regulatory impact analysis

RUC American Medical Association/Specialty Society Relative (Value) Update Committee

RUCA Rural Urban Commuting Area

RVU Relative value unit

SBA Small Business Administration

SGR Sustainable growth rate

SIM State Innovation Model

SLP Speech-language pathology

SMS Socioeconomic Monitoring System

SNF Skilled nursing facility

TAP Technical Advisory Panel

TC Technical component

TIN Tax identification number

TCM Transitional Care Management

UAF Update adjustment factor

UPIN Unique Physician Identification Number

USPSTF United States Preventive Services Task Force

VBP Value-based purchasing

VM Value-Based Payment Modifier

Addenda Available Only Through the Internet on the CMS Web Site

The PFS Addenda along with other supporting documents and tables referenced in this final rule are available through the Internet on the CMS Web site at http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/PFS-Federal-Regulation-Notices.html. Click on the link on the left side of the screen titled, “PFS Federal Regulations Notices” for a chronological list of PFS Federal Register and other related documents. For the CY 2017 PFS Final Rule, refer to item CMS-1654-F. Readers who experience any problems accessing any of the Addenda or other documents referenced in this rule and posted on the CMS Web site identified above should contact Jessica Bruton at (410) 786-5991.

CPT (Current Procedural Terminology) Copyright Notice

Throughout this final rule, we use CPT codes and descriptions to refer to a variety of services. We note that CPT codes and descriptions are copyright 2015 American Medical Association. All Rights Reserved. CPT is a registered trademark of the American Medical Association (AMA). Applicable Federal Acquisition Regulations (FAR) and Defense Federal Acquisition Regulations (DFAR) apply.

I. Executive Summary and Background

A. Executive Summary

1. Purpose

This major final rule revises payment polices under the Medicare Physician Fee Schedule (PFS) and makes other policy changes related to Medicare Part B payment. These changes will be applicable to services furnished in CY 2017. In addition, this final rule includes the following provisions: Payment policy changes for Rural Health Clinics (RHCs) and Federally Qualified Health Centers (FQHCs); expansion of the Medicare Diabetes Prevention Program model; policy changes related to the Medicare Shared Savings Program; and release of pricing data submitted to CMS by Medicare Advantage (MA) organizations; and medical loss ratio reports submitted by MA plans and Part D plans. These additional policies are addressed in section III. of this final rule.

2. Summary of the Major Provisions

The statute requires us to establish payments under the PFS based on national uniform relative value units (RVUs) that account for the relative resources used in furnishing a service. The statute requires that RVUs be established for three categories of resources: Work, practice expense (PE); and malpractice (MP) expense; and, that we establish by regulation each year's payment amounts for all physicians' services paid under the PFS, incorporating geographic adjustments to reflect the variations in the costs of furnishing services in different geographic areas. In this major final rule, we establish RVUs for CY 2017 for the PFS, and other Medicare Part B payment policies, to ensure that our payment systems are updated to reflect changes in medical practice and the relative value of services, as well as changes in the statute. In addition, this final rule includes summaries of public comments and final policies regarding:

• Potentially Misvalued Codes.
• Telehealth Services.
• Establishing Values for New, Revised, and Misvalued Codes.
• Target for Relative Value Adjustments for Misvalued Services.
• Phase-in of Significant RVU Reductions.
• Chronic Care Management (CCM) and Transitional Care Management (TCM) Supervision Requirements in Rural Health Clinics (RHCs) and Federally Qualified Health Centers (FQHCs).
• FQHC-Specific Market Basket.
• Appropriate Use Criteria for Advanced Diagnostic Imaging Services.
• Reports of Payments or Other Transfers of Value to Covered Recipients: Solicitation of Public Comments.
• Release of Part C Medicare Advantage Bid Pricing Data and Part C and Part D Medical Loss Ratio (MLR) Data.
• Prohibition on Billing Qualified Medicare Beneficiary Individuals for Medicare Cost-Sharing.
• Recoupment or Offset of Payments to Providers Sharing the Same Taxpayer Identification Number.
• Accountable Care Organization (ACO) Participants Who Report Physician Quality Reporting System (PQRS) Quality Measures Separately.
• Medicare Advantage Provider Enrollment.
• Expansion of the Diabetes Prevention Program (DPP) Model.
• Medicare Shared Savings Program.
• Value-Based Payment Modifier and the Physician Feedback Program.
• Physician Self-referral Updates.
• Designated Health Services.

3. Summary of Costs and Benefits

The statute requires that annual adjustments to PFS RVUs may not cause annual estimated expenditures to differ by more than $20 million from what they would have been had the adjustments not been made. If adjustments to RVUs would cause expenditures to change by more than $20 million, we must make adjustments to preserve budget neutrality. These adjustments can affect the distribution of Medicare expenditures across specialties. In addition, several changes in this final rule will affect the specialty distribution of Medicare expenditures. When considering the combined impact of work, PE, and MP RVU changes, the projected payment impacts would be small for most specialties; however, the impact would be larger for a few specialties.

We have determined that this major final rule is economically significant. For a detailed discussion of the economic impacts, see section VI. of this final rule.

B. Background

Since January 1, 1992, Medicare has paid for physicians' services under section 1848 of the Social Security Act (the Act), “Payment for Physicians' Services.” The PFS relies on national relative values that are established for work, PE, and MP, which are adjusted for geographic cost variations. These values are multiplied by a conversion factor (CF) to convert the RVUs into payment rates. The concepts and methodology underlying the PFS were enacted as part of the Omnibus Budget Reconciliation Act of 1989 (Pub. L. 101-239, enacted on December 19, 1989) (OBRA '89), and the Omnibus Budget Reconciliation Act of 1990 (Pub. L. 101-508, enacted on November 5, 1990) (OBRA '90). The final rule published on November 25, 1991 (56 FR 59502) set forth the first fee schedule used for payment for physicians' services.

We note that throughout this major final rule, unless otherwise noted, the term “practitioner” is used to describe both physicians and nonphysician practitioners (NPPs) who are permitted to bill Medicare under the PFS for services furnished to Medicare beneficiaries.

1. Development of the Relative Values

a. Work RVUs

The work RVUs established for the initial fee schedule, which was implemented on January 1, 1992, were developed with extensive input from the physician community. A research team at the Harvard School of Public Health developed the original work RVUs for most codes under a cooperative agreement with the Department of Health and Human Services (HHS). In constructing the code-specific vignettes used in determining the original physician work RVUs, Harvard worked with panels of experts, both inside and outside the federal government, and obtained input from numerous physician specialty groups.

As specified in section 1848(c)(1)(A) of the Act, the work component of physicians' services means the portion of the resources used in furnishing the service that reflects physician time and intensity. We establish work RVUs for new, revised and potentially misvalued codes based on our review of information that generally includes, but is not limited to, recommendations received from the American Medical Association/Specialty Society Relative Value Update Committee (RUC), the Health Care Professionals Advisory Committee (HCPAC), the Medicare Payment Advisory Commission (MedPAC), and other public commenters; medical literature and comparative databases; as well as a comparison of the work for other codes within the Medicare PFS, and consultation with other physicians and health care professionals within CMS and the federal government. We also assess the methodology and data used to develop the recommendations submitted to us by the RUC and other public commenters, and the rationale for their recommendations. In the CY 2011 PFS final rule with comment period (75 FR 73328 through 73329), we discussed a variety of methodologies and approaches used to develop work RVUs, including survey data, building blocks, crosswalk to key reference or similar codes, and magnitude estimation. More information on these issues is available in that rule.

b. Practice Expense RVUs

Initially, only the work RVUs were resource-based, and the PE and MP RVUs were based on average allowable charges. Section 121 of the Social Security Act Amendments of 1994 (Pub. L. 103-432, enacted on October 31, 1994), amended section 1848(c)(2)(C)(ii) of the Act and required us to develop resource-based PE RVUs for each physicians' service beginning in 1998. We were required to consider general categories of expenses (such as office rent and wages of personnel, but excluding malpractice expenses) comprising PEs. The PE RVUs continue to represent the portion of these resources involved in furnishing PFS services.

Originally, the resource-based method was to be used beginning in 1998, but section 4505(a) of the Balanced Budget Act of 1997 (Pub. L. 105-33, enacted on August 5, 1997) (BBA) delayed implementation of the resource-based PE RVU system until January 1, 1999. In addition, section 4505(b) of the BBA provided for a 4-year transition period from the charge-based PE RVUs to the resource-based PE RVUs.

We established the resource-based PE RVUs for each physicians' service in a final rule, published on November 2, 1998 (63 FR 58814), effective for services furnished in CY 1999. Based on the requirement to transition to a resource-based system for PE over a 4-year period, payment rates were not fully based upon resource-based PE RVUs until CY 2002. This resource-based system was based on two significant sources of actual PE data: The Clinical Practice Expert Panel (CPEP) data; and the AMA's Socioeconomic Monitoring System (SMS) data. (These data sources are described in greater detail in the CY 2012 final rule with comment period (76 FR 73033).

Separate PE RVUs are established for services furnished in facility settings, such as a hospital outpatient department (HOPD) or an ambulatory surgical center (ASC), and in nonfacility settings, such as a physician's office. The nonfacility RVUs reflect all of the direct and indirect PEs involved in furnishing a service described by a particular HCPCS code. The difference, if any, in these PE RVUs generally results in a higher payment in the nonfacility setting because in the facility settings some costs are borne by the facility. Medicare's payment to the facility (such as the outpatient prospective payment system (OPPS) payment to the HOPD) would reflect costs typically incurred by the facility. Thus, payment associated with those facility resources is not made under the PFS.

Section 212 of the Balanced Budget Refinement Act of 1999 (Pub. L. 106-113, enacted on November 29, 1999) (BBRA) directed the Secretary of Health and Human Services (the Secretary) to establish a process under which we accept and use, to the maximum extent practicable and consistent with sound data practices, data collected or developed by entities and organizations to supplement the data we normally collect in determining the PE component. On May 3, 2000, we published the interim final rule (65 FR 25664) that set forth the criteria for the submission of these supplemental PE survey data. The criteria were modified in response to comments received, and published in the Federal Register (65 FR 65376) as part of a November 1, 2000 final rule. The PFS final rules published in 2001 and 2003, respectively, (66 FR 55246 and 68 FR 63196) extended the period during which we would accept these supplemental data through March 1, 2005.

In the CY 2007 PFS final rule with comment period (71 FR 69624), we revised the methodology for calculating direct PE RVUs from the top-down to the bottom-up methodology beginning in CY 2007. We adopted a 4-year transition to the new PE RVUs. This transition was completed for CY 2010. In the CY 2010 PFS final rule with comment period, we updated the practice expense per hour (PE/HR) data that are used in the calculation of PE RVUs for most specialties (74 FR 61749). In CY 2010, we began a 4-year transition to the new PE RVUs using the updated PE/HR data, which was completed for CY 2013.

c. Malpractice RVUs

Section 4505(f) of the BBA amended section 1848(c) of the Act to require that we implement resource-based MP RVUs for services furnished on or after CY 2000. The resource-based MP RVUs were implemented in the PFS final rule with comment period published November 2, 1999 (64 FR 59380). The MP RVUs are based on commercial and physician-owned insurers' malpractice insurance premium data from all the states, the District of Columbia, and Puerto Rico. For more information on MP RVUs, see section II.B.2. of this final rule.

d. Refinements to the RVUs

Section 1848(c)(2)(B)(i) of the Act requires that we review RVUs no less often than every 5 years. Prior to CY 2013, we conducted periodic reviews of work RVUs and PE RVUs independently. We completed five-year reviews of work RVUs that were effective for calendar years 1997, 2002, 2007, and 2012.

Although refinements to the direct PE inputs initially relied heavily on input from the RUC Practice Expense Advisory Committee (PEAC), the shifts to the bottom-up PE methodology in CY 2007 and to the use of the updated PE/HR data in CY 2010 have resulted in significant refinements to the PE RVUs in recent years.

In the CY 2012 PFS final rule with comment period (76 FR 73057), we finalized a proposal to consolidate reviews of work and PE RVUs under section 1848(c)(2)(B) of the Act and reviews of potentially misvalued codes under section 1848(c)(2)(K) of the Act into one annual process.

In addition to the five-year reviews, beginning for CY 2009, CMS and the RUC have identified and reviewed a number of potentially misvalued codes on an annual basis based on various identification screens. This annual review of work and PE RVUs for potentially misvalued codes was supplemented by the amendments to section 1848 of the Act, as enacted by section 3134 of the Affordable Care Act, that require the agency to periodically identify, review and adjust values for potentially misvalued codes.

e. Application of Budget Neutrality to Adjustments of RVUs

As described in section VI.C. of this final rule, in accordance with section 1848(c)(2)(B)(ii)(II) of the Act, if revisions to the RVUs cause expenditures for the year to change by more than $20 million, we make adjustments to ensure that expenditures did not increase or decrease by more than $20 million.

2. Calculation of Payments Based on RVUs

To calculate the payment for each service, the components of the fee schedule (work, PE, and MP RVUs) are adjusted by geographic practice cost indices (GPCIs) to reflect the variations in the costs of furnishing the services. The GPCIs reflect the relative costs of work, PE, and MP in an area compared to the national average costs for each component.

RVUs are converted to dollar amounts through the application of a CF, which is calculated based on a statutory formula by CMS's Office of the Actuary (OACT). The formula for calculating the Medicare fee schedule payment amount for a given service and fee schedule area can be expressed as:

Payment = [(RVU work × GPCI work) + (RVU PE × GPCI PE) + (RVU MP × GPCI MP)] × CF.

3. Separate Fee Schedule Methodology for Anesthesia Services

Section 1848(b)(2)(B) of the Act specifies that the fee schedule amounts for anesthesia services are to be based on a uniform relative value guide, with appropriate adjustment of an anesthesia conversion factor, in a manner to ensure that fee schedule amounts for anesthesia services are consistent with those for other services of comparable value. Therefore, there is a separate fee schedule methodology for anesthesia services. Specifically, we establish a separate conversion factor for anesthesia services and we utilize the uniform relative value guide, or base units, as well as time units, to calculate the fee schedule amounts for anesthesia services. Since anesthesia services are not valued using RVUs, a separate methodology for locality adjustments is also necessary. This involves an adjustment to the national anesthesia CF for each payment locality.

4. Most Recent Changes to the Fee Schedule

Section 220(d) of the Protecting Access to Medicare Act of 2014 (Pub. L. 113-93, enacted on April 1, 2014) (PAMA) added a new subparagraph (O) to section 1848(c)(2) of the Act to establish an annual target for reductions in PFS expenditures resulting from adjustments to relative values of misvalued codes. If the estimated net reduction in expenditures for a year is equal to or greater than the target for that year, the provision specifies that reduced expenditures attributable to such adjustments shall be redistributed in a budget-neutral manner within the PFS. The provision specifies that the amount by which such reduced expenditures exceed the target for a given year shall be treated as a reduction in expenditures for the subsequent year for purposes of determining whether the target for the subsequent year has been met. The provision also specifies that an amount equal to the difference between the target and the estimated net reduction in expenditures, called the target recapture amount, shall not be taken into account when applying the budget neutrality requirements specified in section 1848(c)(2)(B)(ii)(II) of the Act. The PAMA amendments originally made the target provisions applicable for CYs 2017 through 2020 and set the target for reduced expenditures at 0.5 percent of estimated expenditures under the PFS for each of those 4 years.

Subsequently, section 202 of the Achieving a Better Life Experience Act of 2014 (Division B of Pub. L. 113-295, enacted December 19, 2014) (ABLE) accelerated the application of the target, amending section 1848(c)(2)(O) of the Act to specify that target provisions apply for CYs 2016, 2017, and 2018; and setting a 1 percent target for reduced expenditures for CY 2016 and a 0.5 percent target for CYs 2017 and 2018. The implementation of the target legislation was finalized in the CY 2016 PFS final rule with comment period, and revisions are discussed in section II.G. of this final rule.

Section 1848(c)(7) of the Act, as added by section 220(e) of the PAMA, specified that for services that are not new or revised codes, if the total RVUs for a service for a year would otherwise be decreased by an estimated 20 percent or more as compared to the total RVUs for the previous year, the applicable adjustments in work, PE, and MP RVUs shall be phased in over a 2-year period. Section 220(e) of the PAMA required the phase-in of RVU reductions of 20 percent or more to begin for 2017. Section 1848(c)(7) of the Act was later amended by section 202 of the ABLE Act to require instead that the phase-in must begin in CY 2016. The implementation of the phase-in legislation was finalized in the CY 2016 PFS final rule with comment period and revisions in this year's rulemaking are discussed in section II.H. of this final rule.

II. Provisions of the Final Rule for PFS

A. Determination of Practice Expense (PE) Relative Value Units (RVUs)

1. Overview

Practice expense (PE) is the portion of the resources used in furnishing a service that reflects the general categories of physician and practitioner expenses, such as office rent and personnel wages, but excluding malpractice expenses, as specified in section 1848(c)(1)(B) of the Act. As required by section 1848(c)(2)(C)(ii) of the Act, we use a resource-based system for determining PE RVUs for each physicians' service. We develop PE RVUs by considering the direct and indirect practice resources involved in furnishing each service. Direct expense categories include clinical labor, medical supplies, and medical equipment. Indirect expenses include administrative labor, office expense, and all other expenses. The sections that follow provide more detailed information about the methodology for translating the resources involved in furnishing each service into service-specific PE RVUs. We refer readers to the CY 2010 PFS final rule with comment period (74 FR 61743 through 61748) for a more detailed explanation of the PE methodology.

2. Practice Expense Methodology

a. Direct Practice Expense

We determine the direct PE for a specific service by adding the costs of the direct resources (that is, the clinical staff, medical supplies, and medical equipment) typically involved with furnishing that service. The costs of the resources are calculated using the refined direct PE inputs assigned to each CPT code in our PE database, which are generally based on our review of recommendations received from the RUC and those provided in response to public comment periods. For a detailed explanation of the direct PE methodology, including examples, we refer readers to the Five-Year Review of Work Relative Value Units under the PFS and Proposed Changes to the Practice Expense Methodology proposed notice (71 FR 37242) and the CY 2007 PFS final rule with comment period (71 FR 69629).

b. Indirect Practice Expense per Hour Data

We use survey data on indirect PEs incurred per hour worked in developing the indirect portion of the PE RVUs. Prior to CY 2010, we primarily used the practice expense per hour (PE/HR) by specialty that was obtained from the AMA's Socioeconomic Monitoring Surveys (SMS). The AMA administered a new survey in CY 2007 and CY 2008, the Physician Practice Expense Information Survey (PPIS). The PPIS is a multispecialty, nationally representative, PE survey of both physicians and nonphysician practitioners (NPPs) paid under the PFS using a survey instrument and methods highly consistent with those used for the SMS and the supplemental surveys. The PPIS gathered information from 3,656 respondents across 51 physician specialty and health care professional groups. We believe the PPIS is the most comprehensive source of PE survey information available. We used the PPIS data to update the PE/HR data for the CY 2010 PFS for almost all of the Medicare-recognized specialties that participated in the survey.

When we began using the PPIS data in CY 2010, we did not change the PE RVU methodology itself or the manner in which the PE/HR data are used in that methodology. We only updated the PE/HR data based on the new survey. Furthermore, as we explained in the CY 2010 PFS final rule with comment period (74 FR 61751), because of the magnitude of payment reductions for some specialties resulting from the use of the PPIS data, we transitioned its use over a 4-year period from the previous PE RVUs to the PE RVUs developed using the new PPIS data. As provided in the CY 2010 PFS final rule with comment period (74 FR 61751), the transition to the PPIS data was complete for CY 2013. Therefore, PE RVUs from CY 2013 forward are developed based entirely on the PPIS data, except as noted in this section.

Section 1848(c)(2)(H)(i) of the Act requires us to use the medical oncology supplemental survey data submitted in 2003 for oncology drug administration services. Therefore, the PE/HR for medical oncology, hematology, and hematology/oncology reflects the continued use of these supplemental survey data.

Supplemental survey data on independent labs from the College of American Pathologists were implemented for payments beginning in CY 2005. Supplemental survey data from the National Coalition of Quality Diagnostic Imaging Services (NCQDIS), representing independent diagnostic testing facilities (IDTFs), were blended with supplementary survey data from the American College of Radiology (ACR) and implemented for payments beginning in CY 2007. Neither IDTFs, nor independent labs, participated in the PPIS. Therefore, we continue to use the PE/HR that was developed from their supplemental survey data.

Consistent with our past practice, the previous indirect PE/HR values from the supplemental surveys for these specialties were updated to CY 2006 using the Medicare Economic Index (MEI) to put them on a comparable basis with the PPIS data.

We also do not use the PPIS data for reproductive endocrinology and spine surgery since these specialties currently are not separately recognized by Medicare, nor do we have a method to blend the PPIS data with Medicare-recognized specialty data.

Previously, we established PE/HR values for various specialties without SMS or supplemental survey data by crosswalking them to other similar specialties to estimate a proxy PE/HR. For specialties that were part of the PPIS for which we previously used a crosswalked PE/HR, we instead used the PPIS-based PE/HR. We continue previous crosswalks for specialties that did not participate in the PPIS. However, beginning in CY 2010, we changed the PE/HR crosswalk for portable X-ray suppliers from radiology to IDTF, a more appropriate crosswalk because these specialties are more similar to each other for work time.

For registered dietician services, the resource-based PE RVUs have been calculated in accordance with the final policy that crosswalks the specialty to the “All Physicians” PE/HR data, as adopted in the CY 2010 PFS final rule with comment period (74 FR 61752) and discussed in more detail in the CY 2011 PFS final rule with comment period (75 FR 73183). We have incorporated the available utilization data for interventional cardiology, which became a recognized Medicare specialty during 2014. We finalized the use of a proxy PE/HR value for interventional cardiology in the CY 2016 final rule with comment period (80 FR 70892), as there are no PPIS data for this specialty, by crosswalking the PE/HR from Cardiology, since the specialties furnish similar services in the Medicare claims data.

Comment: A commenter questioned the validity of the PPIS survey data since it is nearly 10 years old. Several other commenters stated that CMS' estimated per-minute labor cost inputs are lower than actual labor costs.

Response: We have previously identified several concerns regarding the underlying data used in determining PE RVUs in the CY 2014 PFS final rule (78 FR 74246-74247). Even when we first incorporated the survey data into the PE methodology, many in the community expressed serious concerns over the accuracy of this or other PE surveys as a way of gathering data on PE inputs from the diversity of providers paid under the PFS. However, we currently lack another source of comprehensive data regarding PE costs, and as a result, we continue to believe that the PPIS survey data is the best data currently available. We continue to seek the best broad-based, auditable, routinely-updated source of information regarding PE costs.

c. Allocation of PE to Services

To establish PE RVUs for specific services, it is necessary to establish the direct and indirect PE associated with each service.

(1) Direct Costs

The relative relationship between the direct cost portions of the PE RVUs for any two services is determined by the relative relationship between the sum of the direct cost resources (that is, the clinical staff, medical supplies, and medical equipment) typically involved with furnishing each of the services. The costs of these resources are calculated from the refined direct PE inputs in our PE database. For example, if one service has a direct cost sum of $400 from our PE database and another service has a direct cost sum of $200, the direct portion of the PE RVUs of the first service would be twice as much as the direct portion of the PE RVUs for the second service.

(2) Indirect Costs

Section II.A.2.b. of this final rule describes the current data sources for specialty-specific indirect costs used in our PE calculations. We allocated the indirect costs to the code level on the basis of the direct costs specifically associated with a code and the greater of either the clinical labor costs or the work RVUs. We also incorporated the survey data described earlier in the PE/HR discussion. The general approach to developing the indirect portion of the PE RVUs is as follows:

• For a given service, we used the direct portion of the PE RVUs calculated as previously described and the average percentage that direct costs represent of total costs (based on survey data) across the specialties that furnish the service to determine an initial indirect allocator. That is, the initial indirect allocator is calculated so that the direct costs equal the average percentage of direct costs of those specialties furnishing the service. For example, if the direct portion of the PE RVUs for a given service is 2.00 and direct costs, on average, represented 25 percent of total costs for the specialties that furnished the service, the initial indirect allocator would be calculated so that it equals 75 percent of the total PE RVUs. Thus, in this example, the initial indirect allocator would equal 6.00, resulting in a total PE RVU of 8.00 (2.00 is 25 percent of 8.00 and 6.00 is 75 percent of 8.00).
• Next, we added the greater of the work RVUs or clinical labor portion of the direct portion of the PE RVUs to this initial indirect allocator. In our example, if this service had a work RVU of 4.00 and the clinical labor portion of the direct PE RVU was 1.50, we would add 4.00 (since the 4.00 work RVUs are greater than the 1.50 clinical labor portion) to the initial indirect allocator of 6.00 to get an indirect allocator of 10.00. In the absence of any further use of the survey data, the relative relationship between the indirect cost portions of the PE RVUs for any two services would be determined by the relative relationship between these indirect cost allocators. For example, if one service had an indirect cost allocator of 10.00 and another service had an indirect cost allocator of 5.00, the indirect portion of the PE RVUs of the first service would be twice as great as the indirect portion of the PE RVUs for the second service.
• Next, we incorporated the specialty-specific indirect PE/HR data into the calculation. In our example, if, based on the survey data, the average indirect cost of the specialties furnishing the first service with an allocator of 10.00 was half of the average indirect cost of the specialties furnishing the second service with an indirect allocator of 5.00, the indirect portion of the PE RVUs of the first service would be equal to that of the second service.

(3) Facility and Nonfacility Costs

For procedures that can be furnished in a physician's office, as well as in a facility setting, where Medicare makes a separate payment to the facility for its costs in furnishing a service, we establish two PE RVUs: Facility, and nonfacility. The methodology for calculating PE RVUs is the same for both the facility and nonfacility RVUs, but is applied independently to yield two separate PE RVUs. In calculating the PE RVUs for services furnished in a facility, we do not include resources that would generally not be provided by physicians when furnishing the service. For this reason, the facility PE RVUs are generally lower than the nonfacility PE RVUs.

(4) Services With Technical Components (TCs) and Professional Components (PCs)

Diagnostic services are generally comprised of two components: A professional component (PC) and a technical component (TC). The PC and TC may be furnished independently or by different providers, or they may be furnished together as a “global” service. When services have separately billable PC and TC components, the payment for the global service equals the sum of the payment for the TC and PC. To achieve this, we use a weighted average of the ratio of indirect to direct costs across all the specialties that furnish the global service, TCs, and PCs; that is, we apply the same weighted average indirect percentage factor to allocate indirect expenses to the global service, PCs, and TCs for a service. (The direct PE RVUs for the TC and PC sum to the global.)

(5) PE RVU Methodology

For a more detailed description of the PE RVU methodology, we refer readers to the CY 2010 PFS final rule with comment period (74 FR 61745 through 61746). We also direct interested readers to the file called “Calculation of PE RVUs under Methodology for Selected Codes” which is available on our Web site under downloads for the CY 2017 PFS final rule at http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/PFS-Federal-Regulation-Notices.html. This file contains a table that illustrates the calculation of PE RVUs as described below for individual codes.

(a) Setup File

First, we create a setup file for the PE methodology. The setup file contains the direct cost inputs, the utilization for each procedure code at the specialty and facility/nonfacility place of service level, and the specialty-specific PE/HR data calculated from the surveys.

(b) Calculate the Direct Cost PE RVUs

Sum the costs of each direct input.

Step 1: Sum the direct costs of the inputs for each service.

Step 2: Calculate the aggregate pool of direct PE costs for the current year. We set the aggregate pool of PE costs equal to the product of the ratio of the current aggregate PE RVUs to current aggregate work RVUs and the proposed aggregate work RVUs.

Step 3: Calculate the aggregate pool of direct PE costs for use in ratesetting. This is the product of the aggregate direct costs for all services from Step 1 and the utilization data for that service.

Step 4: Using the results of Step 2 and Step 3, use the conversion factor to calculate a direct PE scaling factor to ensure that the aggregate pool of direct PE costs calculated in Step 3 does not vary from the aggregate pool of direct PE costs for the current year. Apply the scaling factor to the direct costs for each service (as calculated in Step 1).

Step 5: Convert the results of Step 4 to an RVU scale for each service. To do this, divide the results of Step 4 by the CF. Note that the actual value of the CF used in this calculation does not influence the final direct cost PE RVUs, as long as the same CF is used in Step 4 and Step 5. Different CFs will result in different direct PE scaling factors, but this has no effect on the final direct cost PE RVUs since changes in the CFs and changes in the associated direct scaling factors offset one another.

(c) Create the Indirect Cost PE RVUs

Create indirect allocators.

Step 6: Based on the survey data, calculate direct and indirect PE percentages for each physician specialty.

Step 7: Calculate direct and indirect PE percentages at the service level by taking a weighted average of the results of Step 6 for the specialties that furnish the service. Note that for services with TCs and PCs, the direct and indirect percentages for a given service do not vary by the PC, TC, and global service.

We use an average of the 3 most recent years of available Medicare claims data to determine the specialty mix assigned to each code. As we stated in the CY 2016 final rule with comment period (80 FR 70894), we believe that the 3-year average will mitigate the need to use dominant or expected specialty instead of the claims data. Because we incorporated CY 2015 claims data for use in the CY 2017 proposed rates, we believe that the finalized PE RVUs associated with the CY 2017 PFS final rule provide a first opportunity to determine whether service-level overrides of claims data are necessary. Currently, in the development of PE RVUs we apply only the overrides that also apply to the MP RVU calculation. Since the proposed PE RVUs include a new year of claims into the 3-year average for the first time, we solicited comment on the proposed CY 2017 PFS rates and whether or not the incorporation of a new year of utilization data into a 3-year average mitigates the need for alternative service-level overrides such as a claims-based approach (dominant specialty) or stakeholder-recommended approach (expected specialty) in the development of PE (and MP) RVUs for low-volume codes. Prior year RVUs are available at several locations on the PFS Web site located at http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/.

Comment: Several commenters contended that even a multi-year average of claims data to determine the mix of specialties that furnish the services creates distortions and wide variability for low volume services, particularly those services with fewer than 100 annual Medicare claims. Commenters stated that low volume codes that use a specialty override appear to have stable PE and MP RVUs, while other low volume codes without overrides continue to shift from year to year. Given these fluctuations, commenters suggested that CMS implement service-level overrides to determine the specialty mix for these low volume procedures. These commenters provided a list of nearly 2000 codes and suggested specialty overrides.

Response: We appreciate commenters' interest in relatively stable PE and MP RVUs and for continuing to highlight the challenges faced when determining the specialty allocation for low volume services. Since we did not make a proposal regarding specialty overrides for low volume services, we do not believe that it would be appropriate to establish overrides for several thousand codes at this time. However, given the continued concerns, we will consider the issue, including these specific recommendations, for future rulemaking.

Step 8: Calculate the service level allocators for the indirect PEs based on the percentages calculated in Step 7. The indirect PEs are allocated based on the three components: The direct PE RVUs; the clinical labor PE RVUs; and the work RVUs.

For most services the indirect allocator is: Indirect PE percentage * (direct PE RVUs/direct percentage) + work RVUs.

There are two situations where this formula is modified:

• If the service is a global service (that is, a service with global, professional, and technical components), then the indirect PE allocator is: Indirect percentage (direct PE RVUs/direct percentage) + clinical labor PE RVUs + work RVUs.
• If the clinical labor PE RVUs exceed the work RVUs (and the service is not a global service), then the indirect allocator is: Indirect PE percentage (direct PE RVUs/direct percentage) + clinical labor PE RVUs.

(Note: For global services, the indirect PE allocator is based on both the work RVUs and the clinical labor PE RVUs. We do this to recognize that, for the PC service, indirect PEs will be allocated using the work RVUs, and for the TC service, indirect PEs will be allocated using the direct PE RVUs and the clinical labor PE RVUs. This also allows the global component RVUs to equal the sum of the PC and TC RVUs.)

For presentation purposes, in the examples in the download file called “Calculation of PE RVUs under Methodology for Selected Codes”, the formulas were divided into two parts for each service.

• The first part does not vary by service and is the indirect percentage (direct PE RVUs/direct percentage).
• The second part is either the work RVU, clinical labor PE RVU, or both depending on whether the service is a global service and whether the clinical PE RVUs exceed the work RVUs (as described earlier in this step).

Apply a scaling adjustment to the indirect allocators.

Step 9: Calculate the current aggregate pool of indirect PE RVUs by multiplying the result of step 8 by the average indirect PE percentage from the survey data.

Step 10: Calculate an aggregate pool of indirect PE RVUs for all PFS services by adding the product of the indirect PE allocators for a service from Step 8 and the utilization data for that service.

Step 11: Using the results of Step 9 and Step 10, calculate an indirect PE adjustment so that the aggregate indirect allocation does not exceed the available aggregate indirect PE RVUs and apply it to indirect allocators calculated in Step 8.

Calculate the indirect practice cost index.

Step 12: Using the results of Step 11, calculate aggregate pools of specialty-specific adjusted indirect PE allocators for all PFS services for a specialty by adding the product of the adjusted indirect PE allocator for each service and the utilization data for that service.

Step 13: Using the specialty-specific indirect PE/HR data, calculate specialty-specific aggregate pools of indirect PE for all PFS services for that specialty by adding the product of the indirect PE/HR for the specialty, the work time for the service, and the specialty's utilization for the service across all services furnished by the specialty.

Step 14: Using the results of Step 12 and Step 13, calculate the specialty-specific indirect PE scaling factors.

Step 15: Using the results of Step 14, calculate an indirect practice cost index at the specialty level by dividing each specialty-specific indirect scaling factor by the average indirect scaling factor for the entire PFS.

Step 16: Calculate the indirect practice cost index at the service level to ensure the capture of all indirect costs. Calculate a weighted average of the practice cost index values for the specialties that furnish the service. (Note: For services with TCs and PCs, we calculate the indirect practice cost index across the global service, PCs, and TCs. Under this method, the indirect practice cost index for a given service (for example, echocardiogram) does not vary by the PC, TC, and global service.)

Step 17: Apply the service level indirect practice cost index calculated in Step 16 to the service level adjusted indirect allocators calculated in Step 11 to get the indirect PE RVUs.

(d) Calculate the Final PE RVUs

Step 18: Add the direct PE RVUs from Step 5 to the indirect PE RVUs from Step 17 and apply the final PE budget neutrality (BN) adjustment. The final PE BN adjustment is calculated by comparing the sum of steps 5 and 17 of to the proposed aggregate work RVUs scaled by the ratio of current aggregate PE and work RVUs. This adjustment ensures that all PE RVUs in the PFS account for the fact that certain specialties are excluded from the calculation of PE RVUs but included in maintaining overall PFS budget neutrality. (See “Specialties excluded from ratesetting calculation” later in this section.)

(e) Setup File Information

• Specialties excluded from ratesetting calculation: For the purposes of calculating the PE RVUs, we exclude certain specialties, such as certain NPPs paid at a percentage of the PFS and low-volume specialties, from the calculation. These specialties are included for the purposes of calculating the BN adjustment. They are displayed in Table 1.

• Crosswalk certain low volume physician specialties: Crosswalk the utilization of certain specialties with relatively low PFS utilization to the associated specialties.
• Physical therapy utilization: Crosswalk the utilization associated with all physical therapy services to the specialty of physical therapy.
• Identify professional and technical services not identified under the usual TC and 26 modifiers: Flag the services that are PC and TC services but do not use TC and 26 modifiers (for example, electrocardiograms). This flag associates the PC and TC with the associated global code for use in creating the indirect PE RVUs. For example, the professional service, CPT code 93010 (Electrocardiogram, routine ECG with at least 12 leads; interpretation and report only), is associated with the global service, CPT code 93000 (Electrocardiogram, routine ECG with at least 12 leads; with interpretation and report).
• Payment modifiers: Payment modifiers are accounted for in the creation of the file consistent with current payment policy as implemented in claims processing. For example, services billed with the assistant at surgery modifier are paid 16 percent of the PFS amount for that service; therefore, the utilization file is modified to only account for 16 percent of any service that contains the assistant at surgery modifier. Similarly, for those services to which volume adjustments are made to account for the payment modifiers, time adjustments are applied as well. For time adjustments to surgical services, the intraoperative portion in the work time file is used; where it is not present, the intraoperative percentage from the payment files used by contractors to process Medicare claims is used instead. Where neither is available, we use the payment adjustment ratio to adjust the time accordingly. Table 2 details the manner in which the modifiers are applied.

We also make adjustments to volume and time that correspond to other payment rules, including special multiple procedure endoscopy rules and multiple procedure payment reductions (MPPRs). We note that section 1848(c)(2)(B)(v) of the Act exempts certain reduced payments for multiple imaging procedures and multiple therapy services from the BN calculation under section 1848(c)(2)(B)(ii)(II) of the Act. These MPPRs are not included in the development of the RVUs.

For anesthesia services, we do not apply adjustments to volume since we use the average allowed charge when simulating RVUs; therefore, the RVUs as calculated already reflect the payments as adjusted by modifiers, and no volume adjustments are necessary. However, a time adjustment of 33 percent is made only for medical direction of two to four cases since that is the only situation where a single practitioner is involved with multiple beneficiaries concurrently, so that counting each service without regard to the overlap with other services would overstate the amount of time spent by the practitioner furnishing these services.

• Work RVUs: The setup file contains the work RVUs from this final rule.

(6) Equipment Cost Per Minute

The equipment cost per minute is calculated as:

(1/(minutes per year * usage)) * price * ((interest rate/(1−(1/((1 + interest rate)^ life of equipment)))) + maintenance)

Where:

minutes per year = maximum minutes per year if usage were continuous (that is, usage = 1); generally 150,000 minutes.

usage = variable, see discussion below.

price = price of the particular piece of equipment.

life of equipment = useful life of the particular piece of equipment.

maintenance = factor for maintenance; 0.05.

interest rate = variable, see discussion below.

Usage: We currently use an equipment utilization rate assumption of 50 percent for most equipment, with the exception of expensive diagnostic imaging equipment, for which we use a 90 percent assumption as required by section 1848(b)(4)(C) of the Act.

Stakeholders have often suggested that particular equipment items are used less frequently than 50 percent of the time in the typical setting and that CMS should reduce the equipment utilization rate based on these recommendations. We appreciate and share stakeholders' interest in using the most accurate assumption regarding the equipment utilization rate for particular equipment items. However, we believe that absent robust, objective, auditable data regarding the use of particular items, the 50 percent assumption is the most appropriate within the relative value system. We welcome the submission of data that illustrates an alternative rate.

Maintenance: This factor for maintenance was finalized in the CY 1998 PFS final rule (62 FR 33164).

We continue to investigate potential avenues for determining equipment maintenance costs across a broad range of equipment items.

Comment: One commenter stated that the cost of maintaining imaging equipment exceeds the cost of general medical equipment, and that for imaging modalities the median maintenance cost is approximately 10 percent of the equipment purchase price. The commenter stated that the current 5 percent equipment maintenance rate continues to be an inadequate and outdated reflection of actual maintenance costs. The commenter also stated that information on maintenance costs is readily available to CMS through both public and private sources. The commenter did not identify these sources.

Response: As we previously stated in the CY 2016 final rule with comment period (80 FR 70897), we agree with the commenter that we do not believe the annual maintenance factor for all equipment is exactly 5 percent, and we concur that the current rate likely understates the true cost of maintaining some equipment. We also believe it likely overstates the maintenance costs for other equipment. When we solicited comments regarding sources of data containing equipment maintenance rates, commenters were unable to identify an auditable, robust data source that could be used by CMS on a wide scale. As a result, in the absence of publicly available datasets regarding equipment maintenance costs or another systematic data collection methodology for determining maintenance factor, we do not believe that we have sufficient information at present to adopt a variable maintenance factor for equipment cost per minute pricing. We continue to investigate potential avenues for determining equipment maintenance costs across a broad range of equipment items.

Interest Rate: In the CY 2013 PFS final rule with comment period (77 FR 68902), we updated the interest rates used in developing an equipment cost per minute calculation. The interest rate was based on the Small Business Administration (SBA) maximum interest rates for different categories of loan size (equipment cost) and maturity (useful life). The interest rates are listed in Table 3. (See 77 FR 68902 for a thorough discussion of this issue.) We did not propose any changes to these interest rates for CY 2017.

d. Changes to Direct PE Inputs for Specific Services

This section focuses on specific PE inputs. The direct PE inputs are included in the CY 2017 direct PE input database, which is available on our Web site under downloads for the CY 2017 PFS final rule at http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/PFS-Federal-Regulation-Notices.html.

(1) PE Inputs for Digital Imaging Services

Prior to the CY 2015 PFS rulemaking cycle, the RUC provided a recommendation regarding the PE inputs for digital imaging services. Specifically, the RUC recommended that we remove supply and equipment items associated with film technology from a previously specified list of codes since these items were no longer typical resource inputs. The RUC also recommended that the Picture Archiving and Communication System (PACS) equipment be included for these imaging services since these items are typically used in furnishing imaging services. However, since we did not receive any invoices for the PACS system prior to that year's proposed rule, we were unable to determine the appropriate pricing to use for the inputs. For CY 2015, we finalized our proposal to remove the film supply and equipment items, and to create a new equipment item as a proxy for the PACS workstation as a direct expense (79 FR 67561-67563). We used the price associated with ED021 (computer, desktop, w-monitor) to price the new item, ED050 (PACS Workstation Proxy), pending receipt of invoices to facilitate pricing specific to the PACS workstation. Subsequent to establishing payment rates for CY 2015, we received information from several stakeholders regarding pricing for items related to the digital acquisition and storage of images. We received invoices from one stakeholder that facilitated a proposed price update for the PACS workstation in the CY 2016 PFS proposed rule, and we updated the price for the PACS workstation to $5,557 in the CY 2016 PFS final rule with comment period (80 FR 70899).

In addition to the workstation used by the clinical staff for acquiring the images and furnishing the technical component (TC) of the services, a stakeholder also submitted more detailed information regarding a workstation used by the practitioner interpreting the image in furnishing the professional component (PC) of many of these services.

As we stated in the CY 2015 PFS final rule with comment period (79 FR 67563), we generally believe that workstations used by these practitioners are more accurately considered indirect costs associated with the PC of the service. However, we understand that the professional workstations for interpretation of digital images are similar in principle to some of the previous film inputs incorporated into the global and technical components of the codes, such as the view box equipment. Given that the majority of these services are reported globally in the nonfacility setting, we believe it is appropriate to include these costs as direct inputs for the associated HCPCS codes. Based on our established methodology in which single codes with professional and technical components are constructed by assigning work RVUs exclusively to the professional component and direct PE inputs exclusively to the technical components, these costs would be incorporated into the PE RVUs of the global and technical component of the HCPCS code.

We stated in the CY 2016 PFS final rule with comment period that the costs of the professional workstation may be analogous to costs related to the use of film previously incorporated as direct PE inputs for these services. We also solicited comments on whether including the professional workstation as a direct PE input for these codes would be appropriate, given that the resulting PE RVUs would be assigned to the global and technical components of the codes. Commenters responded by indicating their approval of the concept of a professional PACS workstation used for interpretation of digital images. We received invoices for the pricing of a professional PACS workstation, as well as additional invoices for the pricing of a mammography-specific version of the professional PACS workstation. The RUC also included these new equipment items in its recommendations for the CY 2017 PFS rulemaking cycle.

Based on our analysis of submitted invoices, we proposed to price the professional PACS workstation (ED053) at $14,616.93. We did not propose a change in price for the current technical PACS workstation (ED050), which will remain at a price of $5,557.00.

The price of the professional PACS workstation is based upon individual invoices submitted for the cost of a PC Tower ($1531.52), a pair of 3 MP monitors ($10,500.00 in total), a keyboard and mouse ($84.95), a UPS power backup devices for TNP ($1098.00), and a switch for PACS monitors/workstations ($1402.46).

We proposed to add the professional PACS workstation to many CPT codes in the 70000 series that use the current technical PACS workstation (ED050) and include professional work for which such a workstation would be used. We did not propose to add the equipment item to add-on codes since the base codes would include minutes for the item. We also did not propose to add the item to codes that are therapeutic in nature, as the professional PACS workstation is intended for use in diagnostic services. We therefore did not propose to add the item to codes in the Radiation Therapy section (77261 through 77799) or the Nuclear Medicine Cardiology section (78414-78499). We also did not propose to add the item to image guidance codes where the dominant provider is not a radiologist (77002, 77011, 77071, 77077, and 77081) according to the most recent year of claims data, since we believe a single workstation would be more typical in those cases. We identified approximately 426 codes to which we proposed to add a professional PACS workstation. Please see Table 4 for the full list of affected codes.

For the professional PACS workstation, we proposed to assign equipment time equal to the intraservice work time plus half of the preservice work time associated with the codes, since the work time generally reflects the time associated with the professional interpretation. We proposed half of the preservice work time for the professional PACS workstation because we do not believe that the practitioner would typically spend all of the preservice work period using the equipment. For older codes that do not have a breakdown of physician work time by service period, and only have an overall physician work time, we proposed to use half the total work time as an approximation of the intraservice work time plus one half of the preservice work time. In our review of services that contained an existing PACS workstation and had a breakdown of physician work time, we found that half of the total time was a reasonable approximation for the value of intraservice work time plus one half of preservice work time where no such breakdown existed. We also considered using an equipment time formula of the physician intraservice time plus 1 minute (as a stand-in for the physician preservice work time). We solicited public comment on the most accurate equipment time formula for the professional PACS workstation.

We solicited public comment on the proposed list of codes that would incorporate the professional PACS workstation. We were interested in public comment on the codes for which a professional PACS workstation should be included, and whether one of these professional workstations should be included for codes outside the 70000 series. In cases within the 70000 series where radiologists are not the typical specialty reporting the code, such as CPT codes 77002 and 77011, we asked whether it would be appropriate to add one of the professional PACS workstations to these services.

The following is a summary of the comments we received on the proposed addition of the professional PACS workstation, the pricing of the workstation, the list of codes that would incorporate the professional PACS workstation, and the equipment minutes to assign to the workstation.

Comment: Commenters supported the general concept of the professional PACS workstation and its addition to the proposed list of codes. Commenters stated that the professional PACS workstation is an essential component of diagnostic imaging procedures due to the switch from film to digital technology, and the professional workstation would be an appropriate inclusion as a direct PE input for these services.

Response: We appreciate the support from the commenters for the addition of the professional PACS workstation.

Comment: Many commenters addressed the subject of the proper pricing of the professional PACS workstation. Several commenters requested that CMS increase the price of the workstation to include a third and fourth monitor (for speech recognition) priced at $1,715.98, an Admin Monitor (the extra working monitor) priced at $279.27, and a Powerscribe Microphone priced at $424.00. Commenters stated that speech recognition equipment is typical for a professional PACS workstation, and that physicians typically employed a monitor with greater resolution than what would be typically used for other purposes (such as for electronic health records). Related comments contended that the proposed pricing of the workstation remained significantly less than what the average imaging facility spends on PACS technology. Other commenters disagreed with these sentiments and supported the pricing of the professional PACS workstation at the proposed rate of $14,616.93.

Response: We appreciate the feedback from the commenters regarding the proper pricing of the professional PACS workstation. When proposing a price for the professional PACS workstation, we did not include the cost of the additional monitors and the Powerscribe microphone because these items represent indirect costs under the established PE methodology and the functionality would unlikely have been included in the previously existing film inputs the professional PACS workstation is replacing. Generally, we believe that monitors used to access electronic health records and microphones used for dictation are often used by practitioners who furnish a range a PFS services, are not allocable to particular services or patients, and therefore, are included in the administrative cost category of practice expense, and therefore, are allocated to individual codes through indirect PE RVUs.

Comment: Many commenters stated that CMS should expand the list of codes with a professional PACS workstation. Commenters generally focused on three of the criteria proposed by CMS: The exclusion of the workstation from add-on services, the exclusion of therapeutic (as opposed to diagnostic) services, and the exclusion of codes outside the 70000 series. Commenters stated that add-on codes should be incorporated into the professional PACS workstation list, as they require additional time to perform, and therefore, more time with the technical PACS workstation for the technician, as well as additional time for the review and interpretation performed by the physician using the professional PACS workstation. Commenters also indicated that many therapeutic services would also require a professional PACS workstation, and disagreed with limiting the workstation to diagnostic services only. Finally, commenters supplied extensive lists of additional codes, both inside and outside of the 70000 series, where they stated that the inclusion of a professional PACS workstation was warranted.

Response: We appreciate the feedback from the commenters in helping to define the criteria for inclusion of the professional PACS workstation, along with more specific recommendations about which codes should include the workstation. After considering these comments, we will be adding the professional PACS workstation to additional suggested codes. We took the following into account in making these additions:

• We did not add the professional PACS workstation to any code that currently lacks a technical PACS workstation (ED050) or lacks a work RVU. We continue to believe that procedures which do not include a technical workstation, or do not have physician work, would not require a professional workstation.
• We did not add the professional PACS workstation to add-on codes. Because the base codes include equipment minutes for the workstation, we continue to believe it would be duplicative to add additional equipment time for the professional PACS workstation in the add-on code.
• We agree with commenters that because the clinical utility of the PACS workstation is not necessarily limited to diagnostic services, there may be therapeutic codes where it would be reasonable to assume its use to be typical. We believe that in these specific cases, the use of the professional PACS workstation has been established to be typical for the code in question by the specialties furnishing the service, as a result of the evidence provided in the comments submitted in response to our proposal. We have added the workstation to many of the therapeutic codes requested by commenters, specifically codes listed outside the 70000 series, where use of the professional PACS station is typical.
• Within the 70000 series, we reviewed each of the codes submitted by commenters. Most of these codes did not fall within one of the categories where we proposed to add the professional PACS workstation in the proposed rule: They lacked a technical PACS workstation, they were add-on codes, or they were diagnostic procedures for which radiology is not the dominant specialty providing the service. We continue to believe that the professional PACS workstation should not be added to codes that do not fall into these categories, since we believe that the image must be captured in order to for it to be interpreted, that the use of the PACS workstation in the base code reported with add-on codes would accurately capture the associated resources used, and that the PACS professional workstation is only typically used by radiologists. Based on comments, we are adding the professional workstation to only one code in the 70000 series, CPT code 73562, as it includes a technical PACS workstation, is not an add-on code, and is typically furnished by radiologists.
• For codes in the 80000 and 90000 series, we are concerned about whether it is appropriate to include the technical PACS workstation into many of these services. PACS workstations were created for imaging purposes, but many of these services that include a technical PACS workstation do not appear to make use of imaging. Although we are not removing the technical PACS workstation from these codes at this time, we do not believe that a professional PACS workstation should be added to these procedures. We will consider the inclusion of both PACS workstations for future rulemaking.

Comment: Several commenters addressed the topic of equipment time for the professional PACS workstation. Commenters requested that CMS allocate the entire preservice physician work time associated with the codes, as opposed to the proposed half of the preservice physician work time. Commenters stated that although certain physician work activities in the preservice period may not directly involve the professional workstation, even when the physician is engaged in these parallel work activities, the professional workstation is “open” to the patient at hand and cannot be used for other patients. Commenters also disagreed with the proposal to use half the total time for older codes in which there is no separation of preservice and intraservice period times. Commenters stated that using the entire physician work time would be the best option since there is no accurate way to estimate the service period times, and that it would avoid potential confusion in equipment formulas in the future.

Response: We continue to believe that the professional PACS workstation is more accurately assigned equipment time by using half of the preservice physician work time rather than the full preservice physician work time. As we stated in the proposed rule, we do not believe that the practitioner would typically spend all of the preservice work period using the equipment. Commenters agreed that the physician may not need the professional workstation for the full preservice period, but contended that the equipment would be “open” and unavailable for use by other physicians or for other patients. We disagree with this argument on clinical practice and methodological grounds. We do not agree that the professional PACS workstation would necessarily be unavailable for use by other physicians when the physician in question is not using the machine, Additionally, we note that the number of minutes assigned to the predecessor film inputs did not generally include the full number of pre-service minutes. Finally, our PE methodology is based on the resources typically used to furnish the procedure, and we typically assign time for equipment items based on when it cannot be used by another practitioner or for another patient due to its use in the given procedure. We continue to believe that half of the preservice physician work time (along with the full physician intraservice work time) is a good approximation of the time in the preservice period that the professional PACS workstation will typically be in use. As we stated in the proposed rule, we do not believe that the practitioner would typically spend all of the preservice time using the equipment, and would also spend preservice time on other activities, such as scrubbing and dressing, for example.

For older codes where there is no breakdown of work time values by service period, we do not agree with commenters that the professional PACS workstation should use the total work time. The comments do not provide a persuasive rationale for using the total work time instead of our proposed alternative, developed for consistency with codes for which we do have work time breakdowns by service period. Therefore, in the absence of service period work time detail, we continue to believe that half of the total work time is a reasonable proxy for the small number of old codes affected by this issue. We are not concerned about the potential for confusion in the future with differing equipment time formulas, as the addition of the professional PACS workstation to these codes is a one-time inclusion that will not affect the future review of this equipment.

Finally, we believe that there is a difference in the pattern of equipment usage for the professional PACS workstation between diagnostic and therapeutic codes. Generally, the intraservice work for diagnostic imaging codes describes the review of images, while the intraservice work for therapeutic services describes a broader range of activities. Therefore, although we used an equipment formula of half the preservice physician work time and the full intraservice physician work time for the diagnostic procedures, we do not believe that this same time formula would be appropriate for therapeutic procedures since the professional PACS workstation would not be in use during the intraservice portion of these services. Therefore, we will use an equipment time formula of half the preservice physician work time and half the postservice physician work time for the therapeutic codes to which we are adding a professional PACS workstation, which we believe is more consistent with the descriptions of work for the codes in question. Consistent with our ongoing efforts to improve payment accuracy for these costs, we seek recommendations from the RUC and other stakeholders on a more precise allocation methodology for equipment minutes for these procedures.

After consideration of comments received, we are finalizing our proposal to add a professional PACS workstation (ED053) to the equipment database and price it at the proposed rate of $14,616.93. We are dividing the codes that will contain a professional PACS workstation into diagnostic and therapeutic categories. For diagnostic codes, we are assigning equipment minutes equal to half the preservice physician work time and the full intraservice physician work time. For the relatively smaller group of diagnostic codes with no service period time breakdown, we are assigning equipment time equal to half of the total physician work time. For therapeutic codes, we are assigning equipment minutes equal to half the preservice physician work time and half the postservice physician work time for the second group. There are no therapeutic codes on our current list which lack a service period time breakdown. The following table lists all of the codes that include a professional PACS workstation for CY 2017, along with the equipment minutes for the workstation.

(2) Standardization of Clinical Labor Tasks

As we noted in the CY 2015 PFS final rule (79 FR 67640-67641), we continue to make improvements to the direct PE input database to provide the number of clinical labor minutes assigned for each task for every code in the database instead of only including the number of clinical labor minutes for the preservice, service, and postservice periods for each code. In addition to increasing the transparency of the information used to set PE RVUs, this improvement would allow us to compare clinical labor times for activities associated with services across the PFS, which we believe is important to maintaining the relativity of the direct PE inputs. This information would facilitate the identification of the usual numbers of minutes for clinical labor tasks and the identification of exceptions to the usual values. It would also allow for greater transparency and consistency in the assignment of equipment minutes based on clinical labor times. Finally, we believe that the information can be useful in maintaining standard times for particular clinical labor tasks that can be applied consistently to many codes as they are valued over several years, similar in principle to the use of physician preservice time packages. We believe such standards would provide greater consistency among codes that share the same clinical labor tasks and could improve relativity of values among codes. For example, as medical practice and technologies change over time, changes in the standards could be updated simultaneously for all codes with the applicable clinical labor tasks, instead of waiting for individual codes to be reviewed.

In the following paragraphs, we address a series of issues related to clinical labor tasks, particularly relevant to services currently being reviewed under the misvalued code initiative.

(a) Clinical Labor Tasks Associated With Digital Imaging

In CY 2015 PFS rulemaking, we noted that the RUC recommendation regarding inputs for digital imaging services indicated that, as each code is reviewed under the misvalued code initiative, the clinical labor tasks associated with digital technology (instead of film) would need to be addressed. When we reviewed that recommendation, we did not have the capability of assigning standard clinical labor times for the hundreds of individual codes since the direct PE input database did not previously allow for comprehensive adjustments for clinical labor times based on particular clinical labor tasks. Therefore, consistent with the recommendation, we proposed to remove film-based supply and equipment items but maintain clinical labor minutes that were assigned based on film technology.

As noted in the paragraphs above, we continue to improve the direct PE input database by specifying for each code the minutes associated with each clinical labor task. Once completed, this work would allow adjustments to be made to minutes assigned to particular clinical labor tasks related to digital technology that occur in multiple codes, consistent with the changes that were made to individual supply and equipment items. In the meantime, we believe it would be appropriate to establish standard times for clinical labor tasks associated with all digital imaging services for purposes of reviewing individual services at present, and for possible broad-based standardization once the changes to the direct PE input database facilitate our ability to adjust time across services. During the CY 2016 PFS rulemaking cycle, we proposed appropriate standard minutes for five different clinical labor tasks associated with services that use digital imaging technology. In the CY 2016 PFS final rule with comment period (80 FR 70901), we finalized appropriate standard minutes for four of those five activities, which are listed in Table 5.

We did not finalize standard minutes for the activity “Technologist QC's images in PACS, checking for all images, reformats, and dose page.” We agreed with commenters that this task may require a variable length of time depending on the number of images to be reviewed. We stated that it may be appropriate to establish several different standard times for this clinical labor task for a low/medium/high quantity of images to be reviewed, in the same fashion that the clinical labor assigned to clean a surgical instrument package has two different standard times depending on the use of a basic pack (10 minutes) or a medium pack (30 minutes). We solicited public comment and feedback on this subject, with the anticipation of including a proposal in the CY 2017 proposed rule.

We received many comments suggesting that this clinical labor activity should not have a standard time value. Commenters stated that the number of minutes varies significantly for different imaging modalities; and the time is not simply based on the quantity of images to be reviewed, but also the complexity of the images. The commenters recommended that time for this clinical labor activity should be assigned on a code by code basis. We agree with the commenters that the amount of clinical labor needed to check images in a PACS workstation may vary depending on the service. However, we do not believe that this precludes the possibility of establishing standards for clinical labor tasks as we have done in the past by creating multiple standard times, for example, those assigned to cleaning different kinds of scopes. We continue to believe that the use of clinical labor standards provides greater consistency among codes that share the same clinical labor tasks and can improve relativity of values among codes. We proposed to establish a range of appropriate standard minutes for the clinical labor activity, “Technologist QCs images in PACS, checking for all images, reformats, and dose page.” These standard minutes will be applied to new and revised codes that make use of this clinical labor activity when they are reviewed by us for valuation. We proposed 2 minutes as the standard for the simple case, 3 minutes as the standard for the intermediate case, and 4 minutes as the standard for the complex case. We proposed the simple case of 2 minutes as the standard for the typical procedure code involving routine use of imaging. These values are based upon a review of the existing minutes assigned for this clinical labor activity; we have determined that 2 minutes is the duration for most services and a small number of codes with more complex forms of digital imaging have higher values. We proposed to use 2 minutes for services involving routine X-rays (simple), 3 minutes for services involving CTs and MRIs (intermediate), and 4 minutes for the most highly complex services, which would exceed these more typical cases. We solicited comments regarding the most accurate category—simple, intermediate, or complex for existing codes, and in particular what criteria might be used to identify complex cases systematically.

The following is summary of the comments we received regarding the ongoing standardization of clinical labor tasks, and our specific proposal regarding the clinical labor task, “Technologist QCs images in PACS, checking for all images, reformats, and dose page.”

Comment: Many commenters restated their opposition to the principle of establishing standard values for clinical labor tasks. Commenters contended that clinical labor tasks were highly variable across different specialties, that the standardization process would disrupt the relativity of direct PE inputs across the PFS, and that the proposed standard times were too low and underestimated the staffing time needed to carry out the tasks in question. Commenters stressed that each code should be evaluated on an individual basis. One commenter expressed support for the overall concept regarding efforts to streamline the time for clinical labor activities.

Response: We note the objections raised by the commenters to the process of standardizing time values for clinical labor tasks. However, as we have stated previously, we believe the establishment of standards can provide greater consistency among codes that share the same clinical labor tasks, as well as improve relativity of values among codes. We also note that we do evaluate each code on an individual basis for direct PE inputs, and establishing clinical labor standards assists in that process of individual review. We continue to allow clinical labor times above the standard values for individual services, provided that there is a compelling rationale to explain why that particular service requires additional clinical labor time above and beyond the standard. We believe that establishing a range of standard minutes for this particular digital imaging clinical labor task will provide clarity and help maintain relativity across a wide range of imaging services.

Comment: One commenter requested a broad study of the actual clinical labor times associated with digital imaging.

Response: We appreciate the importance of incorporating robust, auditable, and routinely updated data sources for use in the determination of RVUs. We welcome stakeholder information on the availability of such data, while we continue to consider the best means of acquiring such data.

Comment: Several commenters addressed our specific proposal for the clinical labor task, “Technologist QCs images in PACS, checking for all images, reformats, and dose page.” Commenters requested that, short of no standard times at all, the establishment of categories for this clinical labor task should be as follows: Simple (2 min); intermediate (3 min), complex (4 min) and highly complex (5 min).

Response: We appreciate the suggestion from the commenters to adopt a categorization system very similar to our proposal, with the addition of an extra category for highly complex services valued at 5 minutes. We agree with this addition to our proposal, as it will allow for additional specificity in classifying different types of imaging services, including those that are unusually complex. However, we note that we proposed to define the simple case of 2 minutes as the standard for the typical procedure code involving routine use of imaging, and we believe only a small number of codes with more complex forms of digital imaging would typically involve more time for the task. We proposed to use 2 minutes for services involving routine X-rays (the simple case), and 3 minutes for services involving CTs and MRIs (the intermediate case). We seek recommendations from the RUC and other stakeholders and we intend to request feedback from commenters through future rulemaking to assist in identifying what we believe would be the small number of services that fall into the complex (4 min) and highly complex (5 min) categories, and the specific basis used to set the two categories apart from one another. In the meantime, we will consider individual codes on a case by case basis for this clinical labor task.

After considering the comments received, we are finalizing a range of appropriate standard minutes for the clinical labor activity, “Technologist QCs images in PACS, checking for all images, reformats, and dose page” as follows: Simple (2 min); intermediate (3 min), complex (4 min) and highly complex (5 min). We are also finalizing our criteria for determining the simple and intermediate categories as proposed.

(b) Pathology Clinical Labor Tasks

As with the clinical labor tasks associated with digital imaging, many of the currently assigned times for the specialized clinical labor tasks associated with pathology services are not consistent across codes. In reviewing past RUC recommendations for pathology services, we have not identified information that supports the judgment that the same tasks take significantly more or less time depending on the individual service for which they are performed, especially given the high degree of specificity with which the tasks are described. We continue to believe that, in general, a clinical labor task will tend to take the same amount of time to perform for one individual service as the same clinical labor task when it is performed in a clinically similar service.

Therefore, we developed standard times for clinical labor tasks that we have used in finalizing direct PE inputs in recent years, starting in the CY 2012 PFS final rule with comment period (76 FR 73213). These times were based on our review and assessment of the current times included for these clinical labor tasks in the direct PE input database. We proposed in the CY 2016 PFS proposed rule to establish standard times for a list of 17 clinical labor tasks related to pathology services, and solicited public feedback regarding our proposed standards. Many commenters stated in response to our proposal that they did not support the standardization of clinical labor activities across pathology services. Commenters stated that establishing a single standard time for each clinical labor task was infeasible due to the differences in batch size or number of blocks across different pathology procedures. Several commenters indicated that it might be possible to standardize across codes with the same batch sizes, and urged us to consider pathology-specific details, such as batch size and block number, in the creation of any future standard times for clinical labor tasks related to pathology services.

As we stated in the CY 2016 PFS proposed rule, we developed the proposed standard times based on our review and assessment of the current times included for these clinical labor tasks in the direct PE input database. We believe that, generally speaking, clinical labor tasks with the same description are comparable across different pathology procedures. We believe this to be true based on the comparability of clinical labor tasks in non-pathology services, as well as the high degree of specificity with which most clinical labor tasks for pathology services are described relative to clinical labor tasks associated with other PFS services. We concurred with commenters that accurate clinical labor times for pathology codes may be dependent on the number of blocks or batch size typically used for each individual service. However, we also believe that it is appropriate and feasible to establish “per block” standards or standards varied by batch size assumptions for many clinical labor activities that would be comparable across a wide range of individual services. We have received detailed information regarding batch size and number of blocks during review of individual pathology services on an intermittent basis in the past. We requested regular submission of these details on the PE worksheets supplied by the RUC as part of the review process for pathology services, as a means to assist in the determination of the most accurate direct PE inputs.

We also stated our belief that many of the clinical labor activities for which we proposed to establish standard times were tasks that do not depend on number of blocks or batch size. Clinical labor activities such as “Clean room/equipment following procedure” and “Dispose of remaining specimens” would typically remain standard across different services without varying by block number or batch size, with the understanding that additional time may be required above the standard value for a clinical labor task that is part of an unusually complex or difficult service. As a result, we ultimately finalized standard times for 6 of the 17 proposed clinical labor activities in the CY 2016 final rule with comment period (80 FR 70902). We have listed the finalized standard times in Table 6. We are taking no further action on the remaining 11 clinical labor activities in this final rule, pending further action by the RUC (see below).

We remain committed to the process of establishing standard clinical labor times for tasks associated with pathology services. This may include establishing standards on a per-block or per-batch basis, as we indicated during the previous rulemaking cycle. However, we are aware that the PE Subcommittee of the RUC is currently working to standardize the pathology clinical labor activities they use in making their recommendations. We believe the RUC's efforts to narrow the current list of several hundred pathology clinical labor tasks to a more manageable number through the consolidation of duplicative or highly similar activities into a single description may serve PFS relativity and facilitate greater transparency in PFS ratesetting. We also believe that the RUC's standardization of pathology clinical labor tasks would facilitate our capacity to establish standard times for pathology clinical labor tasks in future rulemaking. Therefore, we did not propose any additional changes to clinical labor tasks associated with pathology services.

(3) Equipment Recommendations for Scope Systems

During our routine reviews of direct PE input recommendations, we have regularly found unexplained inconsistencies involving the use of scopes and the video systems associated with them. Some of the scopes include video systems bundled into the equipment item, some of them include scope accessories as part of their price, and some of them are standalone scopes with no other equipment included. It is not always clear which equipment items related to scopes fall into which of these categories. We have also frequently found anomalies in the equipment recommendations, with equipment items that consist of a scope and video system bundle recommended, along with a separate scope video system. Based on our review, the variations do not appear to be consistent with the different code descriptions.

To promote appropriate relativity among the services and facilitate the transparency of our review process, during review of recommended direct PE inputs for the CY 2017 PFS proposed rule, we developed a structure that separates the scope and the associated video system as distinct equipment items for each code. Under this approach, we proposed standalone prices for each scope, and separate prices for the video systems that are used with scopes. We would define the scope video system as including: (1) A monitor; (2) a processor; (3) a form of digital capture; (4) a cart; and (5) a printer. We believe that these equipment components represent the typical case for a scope video system. Our model for this system is the “video system, endoscopy (processor, digital capture, monitor, printer, cart)” equipment item (ES031), which we proposed to re-price as part of this separate pricing approach. We obtained current pricing invoices for the endoscopy video system as part of our investigation of these issues involving scopes, which we proposed to use for this re-pricing. We understand that there may be other accessories associated with the use of scopes; we proposed to separately price any scope accessories, and individually evaluate their inclusion or exclusion as direct PE inputs for particular codes as usual under our current policy based on whether they are typically used in furnishing the services described by the particular codes.

We also proposed standardizing refinements to the way scopes have been defined in the direct PE input database. We believe that there are four general types of scopes: Non-video scopes; flexible scopes; semi-rigid scopes, and rigid scopes. Flexible scopes, semi-rigid scopes, and rigid scopes would typically be paired with one of the video scope systems, while the non-video scopes would not. The flexible scopes can be further divided into diagnostic (or non-channeled) and therapeutic (or channeled) scopes. We proposed to identify for each anatomical application: (1) A rigid scope; (2) a semi-rigid scope; (3) a non-video flexible scope; (4) a non-channeled flexible video scope; and (5) a channeled flexible video scope. We proposed to classify the existing scopes in our direct PE database under this classification system, to improve the transparency of our review process and improve appropriate relativity among the services. We plan to propose input prices for these equipment items through future rulemaking.

We proposed these changes only for the reviewed codes that make use of scopes; this applies to the codes in the Flexible Laryngoscope family (CPT codes 31572, 31573, 31574, 31575, 31576, 31577, 31578, 31579) (see section II.L) and the Laryngoplasty family (CPT codes 31551, 31552, 31553, 31554, 31580, 31584, 31587, 31591, 31592) (see section II.L) along with updated prices for the equipment items related to scopes utilized by these services. We also solicited comment on this separate pricing structure for scopes, scope video systems, and scope accessories, which we could consider proposing to apply to other codes in future rulemaking.

The following is a summary of the comments we received on this separate pricing structure for scopes, scope video systems, and scope accessories.

Comment: Many commenters addressed our general proposal to reclassify scopes and their related equipment items. Commenters expressed their support for the decision to remove the scopes from the proposed scope packages, and the proposed definition of the scope video system based on the current endoscopy video system equipment item (ES031). There were no comments opposing the general principle behind reclassifying scopes and scope equipment.

Response: We appreciate the support from the commenters for the broad project to clarify these issues related to scopes.

Comment: Many commenters also requested that CMS delay implementing the scope proposal until additional time could be devoted to the subject. Several commenters asked CMS to wait to make any changes until the RUC could form a PE Subcommittee to address this issue. For codes with proposed CY 2017 values, commenters urged CMS to adopt the RUC-recommended direct PE inputs instead of the proposed direct PE inputs, pending anticipated RUC recommendations on the subject. Another commenter requested that CMS make no change for CY 2017 for any endoscopy procedures until proper identification of the capital and disposable cost inputs could be confirmed.

Response: We appreciate commenters' interests in making certain that there is appropriate opportunity for stakeholders to provide feedback and recommendations on the reclassification of scopes and related scope equipment. This was our primary rationale for limiting proposed changes regarding these kinds of inputs to codes reviewed for the current CY 2017 rule cycle, that is, the Flexible Laryngoscope and Laryngoplasty families of codes. Because these codes are under current review; however, we believe that they should be valued according to a scheme that accurately describes the scope equipment typically used in the services. As a result, we continue to believe that our proposed classification system for scopes is the more sound methodology to use for valuation of these two families of codes for CY 2017. However, we note that we would expect to include examination of these codes as part of any broader proposal we would make regarding scope equipment items, in response to new recommendations on the subject.

We look forward to receiving recommendations from the upcoming RUC PE Subcommittee regarding scopes and related scope equipment items. We note that in order for these recommendations to be considered for CY 2018 rulemaking, we would need to receive these recommendations by the same February deadline for the submission of recommendations on code valuations.

Comment: Many commenters disagreed with the CMS proposal to price the endoscopy video system (ES031) at a price of $15,045.00. Some commenters stated that CMS should use the submitted invoices for the pricing of this equipment, which recommended a price of $49,400.00. One commenter stated that the proposed amount did not accurately reflect the current price of GI endoscopy video systems. Another commenter stated that CMS had defined the endoscopy video system as containing five items: (1) A monitor; (2) a processor; (3) a form of digital capture; (4) a cart; and (5) a printer. However, the commenter pointed out that CMS had not included a price for the digital capture device, which the commenter stressed was a significant part of the overall cost and needed to be included in the equipment's pricing. The commenter submitted a series of new invoices for endoscopy video system and requested that CMS incorporate them into the pricing of the equipment.

Response: We appreciate the feedback from the commenters about pricing, especially the submission of new data in the form of additional invoices. We agree that the cost of a digital capture device should be included in the cost of the endoscopy video system; it was our belief that the digital capture device was included in the cost of the processor. We appreciate the clarification from the commenters indicating that this is not the case, and that the digital capture device is a separately priced component of the video system. As a result, we are averaging the price of the digital capture device on the two submitted invoices and pricing it at $18,346.00. We will add this into the overall cost of the endoscopy video system.

For the other four components of the video system, we are finalizing the prices as proposed. The invoices submitted for these components indicate that they are different forms of equipment with different product IDs and different prices. For example, our price for the processor comes from a “Video Processor with keyboard & video cable” (CV-180) as opposed to the newly submitted invoice for a “Viscera Elite Video System” (OTV-S190). These are two distinct equipment items, and we do not have any data to indicate that the equipment on the newly submitted invoices is more typical in its use than the equipment that we are currently using to price the endoscopy video system.

Therefore, we are finalizing the price of the endoscopy video system at $33,391.00, based on component prices of $9,000.00 for the processor, $18,346.00 for the digital capture device, $2,000.00 for the monitor, $2,295.00 for the printer, and $1,750.00 for the cart.

Comment: A few commenters also addressed the pricing of related scope accessories. They stated that the proposed price for the fiberscope, flexible, rhinolaryngoscopy (ES020) was decreased by 33 percent based on one unrepresentative invoice and that this price undervalued the actual cost. Similarly, commenters stated that the proposed price for the stroboscopy system (ES065) at $19,100 was much lower than the manufacturer average invoice pricing. The proposed prices for the channeled and non-channeled flexible video rhinolaryngoscopes (ES064 and ES063 respectively) were also both two to three times lower than the manufacturer's average invoice price. One commenter submitted additional invoices for pricing these scopes and scope accessories.

Response: We appreciate the submission of this additional pricing data for review. Although many commenters stated that the price of the stroboscopy system was too low, only one commenter supplied additional invoices for the same equipment item that we defined in the proposed rule, the StrobeLED system, and these invoices reflected lower prices than the one we had proposed. These invoices reflected prices of $16,431.00 and $15,000.00. We are averaging these together with our previously submitted price of $19,100.00 for the stroboscopy system, which results in a new price of $16,843.87.

When we reviewed the invoices for the channeled and non-channeled flexible video rhinolaryngoscopes (ES064 and ES063 respectively), we found that the product numbers indicated that these were different equipment items than the scopes that we priced in the proposed rule. As we mentioned for the pricing of the endoscopy video system, we have no data to indicate that use of these particular rhinolaryngoscopes would be typical, as opposed to the rhinolaryngoscopes that we proposed to use to establish prices in the proposed rule. As a result, we are maintaining our current prices for these scopes pending the submission of additional information.

We similarly found that the invoices with recommended price increases for the endoscope, rigid, sinoscopy (ES013) from the current price of $2,414.17 to $4,024.00 and for the videoscope, colonoscopy (ES033) from $23,650.00 to $37,273.00 related to different equipment items that we do not believe are a better reflection of the typical case than the item we currently use. We did not propose to make price changes for these scopes, and we have not incorporated these equipment items into the new scope classification system. As we stated previously, we are currently limiting the scope changes to the CPT codes under review for CY 2017 and their associated equipment items. We will consider pricing changes for the rest of the scopes and associated scope equipment as part of the broader scope reclassification and pricing effort in future rulemaking.

We received invoices for a series of equipment items listed as “other capital inputs not included in CMS estimate” as part of this collection of invoices. Since these equipment items were not included in the original recommendations or our proposed valuations for the Flexible Laryngoscope and Laryngoplasty families of codes, we are not adding them to our equipment database at this time. We will consider the addition of these equipment items as part of the broader recommendations from the RUC PE Subcommittee on the scope classification project.

We did not receive an invoice or other data to support a change in the pricing of the fiberscope, flexible, rhinolaryngoscopy (ES020).

Comment: Many commenters objected to the use of a vendor quote for pricing of the scope equipment. Commenters requested that specialty societies should also be allowed to submit quotes for pricing as they are easier to obtain than paid invoices. Commenters also stated that the use of vendor prices created transparency issues and asked CMS to explain why they are appropriate to use rather than invoices supplied by specialties. One commenter stated that a single invoice was not an adequate sample to use as a pricing input for many types of endoscopic equipment.

Response: We are always interested in investigating multiple data sources for use in pricing supplies and equipment, provided that the information can be verified as accurate. We agree with the commenter that a single voluntarily submitted invoice may not be an adequate source for making wide ranging pricing decisions. We prefer to have pricing information from multiple data sources whenever possible, which may include information obtained from vendors of medical supplies and equipment. We continue to believe that there are risks of bias in submission of price quotes used for purposes of ratesetting. However, given the way we use these prices in the current ratesetting methodologies, we believe the risk of bias is located in submission of overstated, not understated prices. Therefore, we believe it is reasonable to assume that practitioners would generally be able acquire particular items at the prices vendors submit to CMS.

After consideration of comments received, we are finalizing our proposals as detailed in the proposed rule, with the updated prices for the endoscopy video system and the stroboscopy system.

(4) Technical Corrections to Direct PE Input Database and Supporting Files

Subsequent to the publication of the CY 2016 PFS final rule with comment period, stakeholders alerted us to several clerical inconsistencies in the direct PE database. We proposed to correct these inconsistencies as described below and reflected in the CY 2017 direct PE input database displayed on our Web site under downloads for the CY 2017 PFS proposed rule at http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/PFS-Federal-Regulation-Notices.html.

For CY 2017, we proposed the following technical corrections:

• For CPT codes 72081-72084, a stakeholder informed us that the equipment time for the PACS workstation (ED050) should be equal to the clinical labor during the service period; the equipment time formula we used for these codes for CY 2016 erroneously included 4 minutes of preservice clinical labor. We agree with the stakeholder that the PACS workstation should use the standard equipment time formula for a PACS workstation for these codes. As a result, we proposed to refine the ED050 equipment time to 21 minutes for CPT code 72081, 36 minutes for CPT code 72082, 44 minutes for CPT code 72083, and 53 minutes for CPT code 72084 to reflect the clinical labor time associated with these codes. This same commenter also indicated that a number of clinical labor activities had been entered in the database in the incorrect service period for CPT codes 37215, 50432, 50694, and 72081. These clinical labor activities were incorrectly listed in the “postservice” period instead of the “service post” period. We proposed to make these technical corrections as well so that the minutes are assigned to the appropriate service period within the direct PE input database.
• Another stakeholder alerted us that ileoscopy CPT codes 44380, 44381 and 44382 did not include the direct PE input equipment item called the Gomco suction machine (EQ235) and indicated that this omission appeared to be inadvertent. We agreed that it was. We have included the item EQ235 in the final direct PE input database for CPT code 44380 at a time of 29 minutes, for CPT code 44381 at a time of 39 minutes, and CPT code 44382 at a time of 34 minutes.

The PE RVUs displayed in Addendum B on our Web site were calculated with the inputs displayed in the CY 2017 direct PE input database.

Comment: One commenter expressed support for the proposed technical corrections to these services.

Response: We appreciate the support from the commenter. After consideration of comments received, we are finalizing these technical corrections.

Comment: Several commenters contacted CMS during the comment period after noticing that six services where CMS proposed to accept the refinement panel work RVU did not contain the updated work RVU in the Addendum B file for the proposed rule. These commenters requested that CMS address these discrepancies.

Response: We appreciate the assistance from the commenters in recognizing these discrepancies. We have corrected them and assigned the refinement panel work RVUs to the six services in question.

Comment: One commenter stated that there were potential technical errors in the clinical labor inputs for CPT codes 88329, 88331, 88360, and 88361.

Response: We have reviewed these codes and they do not contain technical errors. The clinical labor inputs were adjusted in the CY 2016 rule cycle as a result of CMS refinement (80 FR 70981-70983).

(5) Restoration of Inputs

Several of the PE worksheets included in the RUC recommendations for CY 2016 contained time for the equipment item “xenon light source” (EQ167). Because there appeared to be two special light sources already present (the fiberoptic headlight and the endoscope itself) in the services for which this equipment item was recommended by the RUC, we believed that the use of only one of these light sources would be typical and proposed to remove the xenon light equipment time. In the CY 2016 PFS final rule with comment period, we restored the xenon light (EQ167) and removed the fiberoptic headlight (EQ170) with the same number of equipment minutes for CPT codes 30300, 31295, 31296, 31297, and 92511.

We received comments expressing approval for the restoration of the xenon light. However, the commenters also stated that the two light sources were not duplicative, but rather, both a headlight and a xenon light source are required concurrently for otolaryngology procedures when scopes are utilized. The commenters requested that the fiberoptic headlight be restored to these codes.

We agreed with the commenters that the use of both light sources would be typical for these procedures. Therefore, we proposed in the CY 2017 proposed rule to add the fiberoptic headlight (EQ170) to CPT codes 30300, 31295, 31296, 31297, and 92511 at the same number of equipment minutes as the xenon light (EQ167).

Comment: One commenter expressed appreciation for the CMS proposal to restore the fiberoptic headlight to the codes in question. The commenter also stated that it had supplied invoices for LED lights, which are significantly less expensive than the xenon light source, as it was this commenter's understanding that xenon lights are no longer the typical light source for these procedures and they are no longer widely available for purchase from vendors. The commenter expressed support for retaining the xenon light as the standard light source line item for all endoscopy codes if that remained CMS' preference.

Response: We appreciate the support for our proposal from the commenter, as well as the submission of additional information regarding the typical light source for these procedures. We will add the LED light source to our equipment database at the submitted invoice price of $1,915.00. However, we will not replace the xenon light with the LED light at this time, as we believe the subject deserves further consideration. We will consider proposing this change in future rulemaking.

Comment: We received new invoices for the xenon light equipment from a different commenter which averaged out to a price of $12,298.00.

Response: We are finalizing our proposed price of $7,000.00 for the xenon light source. Since we received a comment stating that xenon lights are no longer a typical light source for procedure use, and that they have been supplanted by the use of LED lights, we are viewing the current input as a proxy item, and therefore, do not believe that it would be appropriate to increase the cost of the xenon light source at this time. We will consider making a proposal to address this subject in future rulemaking.

After consideration of comments received, we are finalizing our proposal to add the fiberoptic headlight (EQ170) to CPT codes 30300, 31295, 31296, 31297, and 92511 at the same number of equipment minutes as the xenon light (EQ167).

(6) Updates to Prices for Existing Direct PE Inputs

In the CY 2011 PFS final rule with comment period (75 FR 73205), we finalized a process to act on public requests to update equipment and supply price and equipment useful life inputs through annual rulemaking, beginning with the CY 2012 PFS proposed rule. For CY 2017, we proposed the following price updates for existing direct PE inputs:

Several commenters wrote to discuss the price of the Antibody Estrogen Receptor monoclonal (SL493). We received information including three invoices with new pricing information regarding the SL493 supply. We proposed to use this information to propose for the supply item SL493 a price of $14.00 per test, which is the average price based on the invoices that we received in total for the item.

Comment: Several commenters supported the proposed price increase and urged CMS to finalize the proposal.

Response: We appreciate the support from the commenters. After consideration of comments received, we are finalizing the price of the Antibody Estrogen Receptor monoclonal (SL493) supply at $14.00 as proposed.

We also proposed to update the price for two supplies in response to the submission of new invoices. The proposed price for “antigen, venom” supply (SH009) reflects an increase from $16.67 to $20.14 per milliliter, and the proposed price for “antigen, venom, tri-vespid” supply (SH010) reflects an increase from $30.22 to $44.05 per milliliter.

Comment: Several commenters stated that they strongly supported the proposed price updates for antigen supplies and urged CMS to finalize the proposal.

Response: We appreciate the support from the commenters. After consideration of comments received, we are finalizing the price of the “antigen, venom” (SH009) and “antigen, venom, tri-vespid” (SH010) supplies as proposed.

We proposed to remove the laser tip, diffuser fiber supply (SF030) and replace it with the laser tip, bare (single use) supply (SF029) for CPT code 31572 (formerly placeholder code 317X1). We did not propose a price change for the SF030 supply.

Comment: In reference to CPT code 52648, a commenter stated that the price for the laser tip, diffuser fiber supply (SF030) was decreasing from $850 to $197.50. The commenter stated that the methodology for this adjustment was opaque, unanticipated, and not proposed for comment in the proposed rule. The commenter stated that the $850 supply cost would be more appropriate for the laser tip, diffuser fiber supply.

Response: We stated in the CY 2017 proposed rule (81 FR 46247) that we did not believe that the submitted invoice for the laser tip, diffuser supply at $197.50 was current enough to establish a new price for the supply. As a result, we proposed to remove the laser tip, diffuser fiber supply (SF030) and replaced it with the laser tip, bare (single use) supply (SF029) for CPT code 31572 (Laryngoscopy, flexible; with ablation or destruction of lesion(s) with laser, unilateral), as we did not believe that it was appropriate to use a supply with an outdated invoice. However, we inadvertently set the price of the laser tip, diffuser fiber supply to $197.50 in the proposed direct PE input database in contradiction of our written proposal. We apologize for the confusion caused by this error. In the final direct PE input database, we are restoring the price of the laser tip, diffuser fiber supply to $850.00, since we did not intend to propose a change the price of this supply. We are also requesting the submission of additional current pricing information for the laser tip, diffuser fiber supply, given the significant difference between the $197.50 and $850.00 prices.

Comment: A commenter submitted two invoices containing pricing data for a Cook Biopsy device.

Response: While we appreciate the submission of this pricing information from the commenter, we are unable to determine which supply or equipment item these invoices were in reference to. The invoices were not mentioned in the text of the commenter's letter. We request that invoices submitted for pricing updates should contain clear documentation regarding the item in question: its name, the CMS supply/equipment code that it references (if any), the unit quantity if the item is shipped in boxes or batches, and any other information relevant for pricing.

We routinely accept public submission of invoices as part of our process for developing payment rates for new, revised, and potentially misvalued codes. Often these invoices are submitted in conjunction with the RUC-recommended values for the codes. For CY 2017, we note that some stakeholders have submitted invoices for new, revised, or potentially misvalued codes after the February deadline established for code valuation recommendations. To be considered for a given year's proposed rule, we generally need to receive invoices by the same February deadline. In similar fashion, we generally need to receive invoices by the end of the comment period for the proposed rule in order to consider them for supply and equipment pricing in the final rule for that calendar year. Of course, we consider invoices submitted as public comments during the comment period following the publication of the proposed rule when relevant for services with values open for comment, and will consider any other invoices received after February and/or outside of the public comment process as part of our established annual process for requests to update supply and equipment prices as finalized in the CY 2011 final rule with comment period (75 FR 73205).

(7) Radiation Treatment Delivery Practice Expense RVUs

Comment: Several commenters noticed that there was a 10 percent decrease in the proposed Non Facility PE RVUs for HCPCS code G6011 despite proposed changes in direct PE inputs. Commenters requested an explanation for why this decrease was taking place, and referenced section 3 of the Patient Access and Medicare Protection Act (PAMPA) (Pub. L. 114-115, enacted December 18, 2015), which requires CMS to maintain the associated “definitions, units, and inputs” for certain radiation treatment and related services for CY 2017 and CY 2018. Several commenters stated that they believed that this decrease in the PE RVU was in violation of section 1848(c)(2)(C)(i-ii) of the Act (added by section 3 of the PAMPA), which requires inputs for these services to remain unchanged for CY 2017 and 2018.

Response: We agree with the commenters that we did not propose to change any of the direct PE inputs for HCPCS code G6011, and we understand the proposed change in the nonfacility PE RVUs would generally not be expected absent a corresponding change in direct PE inputs. However, the change in the PE RVU for HCPCS code G6011 is caused by a significant shift in the specialties furnishing the service in the Medicare claims data. In the claims data we used to establish the PE RVUs for CY 2016, dermatology furnished 51 percent of the services, while radiation oncology furnished 43 percent. The most recent claims data reflects a major shift, with radiation oncology now furnishing about 85 percent of the services and dermatology only about 6 percent. The decrease in the PE RVU between CY 2016 and CY 2017 resulted from this shift in specialty mix, as the specialties actually furnishing the service, reflected in the claims data, have a higher percentage of direct PE relative to indirect PE, and therefore, a lower percentage of indirect PE, than the specialties that were previously furnishing the service in the claims data. In other words, consistent with the established methodology for allocating indirect PE to services, a specialty mix with a lower percentage of indirect PE results in fewer indirect PE RVUs being allocated and a lower overall PE RVU for the code even though the direct PE inputs have remained the same. This kind of shift is relatively unusual outside of low-volume codes, but it is consistent with our established methodology for allocating indirect PE to services. We believe that in many cases, the change in specialty utilization for a particular service would warrant a re-examination of the direct PE inputs for the service under the misvalued code initiative. Given the statutory provision that prohibits us from changing the direct PE inputs prior to CY 2019 or considering these services as potentially misvalued, we will consider this issue further for future rulemaking.

We recognize that this change would be unanticipated, but we do not believe there is a straightforward, transparent way to offset the change since the statutory provision requires that we maintain the direct inputs for the PE RVUs. We note that this change is unique among the radiation therapy and related imaging codes where the maintenance of inputs has generally resulted in payment rate stability for these services.

B. Determination of Malpractice Relative Value Units (RVUs)

1. Overview

Section 1848(c) of the Act requires that each service paid under the PFS be composed of three components: Work, PE, and malpractice (MP) expense. As required by section 1848(c)(2)(C)(iii) of the Act, beginning in CY 2000, MP RVUs are resource based. Malpractice RVUs for new codes after 1991 were extrapolated from similar existing codes or as a percentage of the corresponding work RVU. Section 1848(c)(2)(B)(i) of the Act also requires that we review, and if necessary adjust, RVUs no less often than every 5 years. In the CY 2015 PFS final rule with comment period, we implemented the third review and update of MP RVUs. For a comprehensive discussion of the third review and update of MP RVUs see the CY 2015 proposed rule (79 FR 40349 through 40355) and final rule with comment period (79 FR 67591 through 67596).

To determine MP RVUs for individual PFS services, our MP methodology is comprised of three factors: (1) Specialty-level risk factors derived from data on specialty-specific MP premiums incurred by practitioners, (2) service level risk factors derived from Medicare claims data of the weighted average risk factors of the specialties that furnish each service, and (3) an intensity/complexity of service adjustment to the service level risk factor based on either the higher of the work RVU or clinical labor RVU. Prior to CY 2016, MP RVUs were only updated once every 5 years, except in the case of new and revised codes.

As explained in the CY 2011 PFS final rule with comment period (75 FR 73208), MP RVUs for new and revised codes effective before the next 5-year review of MP RVUs were determined either by a direct crosswalk from a similar source code or by a modified crosswalk to account for differences in work RVUs between the new/revised code and the source code. For the modified crosswalk approach, we adjust (or scale) the MP RVU for the new/revised code to reflect the difference in work RVU between the source code and the new/revised work RVU (or, if greater, the difference in the clinical labor portion of the fully implemented PE RVU) for the new code. For example, if the proposed work RVU for a revised code were 10 percent higher than the work RVU for its source code, the MP RVU for the revised code would be increased by 10 percent over the source code MP RVU. Under this approach, the same risk factor is applied for the new/revised code and source code, but the work RVU for the new/revised code is used to adjust the MP RVUs for risk.

In the CY 2016 PFS final rule with comment period (80 FR 70906 through 70910), we finalized a policy to begin conducting annual MP RVU updates to reflect changes in the mix of practitioners providing services (using Medicare claims data), and to adjust MP RVUs for risk for intensity and complexity (using the work RVU or clinical labor RVU). We also finalized a policy to modify the specialty mix assignment methodology (for both MP and PE RVU calculations) to use an average of the 3 most recent years of data instead of a single year of data. We stated that under this approach, the specialty-specific risk factors would continue to be updated through notice and comment rulemaking every 5 years using updated premium data, but would remain unchanged between the 5-year reviews.

For CY 2016, we did not propose to discontinue our current approach for determining MP RVUs for new/revised codes. For the new and revised codes for which we proposed work RVUs and PE inputs, we also published the proposed MP crosswalks used to determine their MP RVUs. We address comments regarding valuation of new and revised codes in section II.L of this final rule, which makes clear the codes with interim final values for CY 2016 had newly proposed values for CY 2017, all of which were again open for comment. The MP crosswalks for new and revised codes with interim final values were established in the CY 2016 PFS final rule with comment period; we proposed these same crosswalks in the CY 2017 PFS proposed rule.

2. Updating Specialty Specific Risk Factors

The proposed CY 2017 GPCI update (eighth update), discussed in section II.E of this final rule, reflects updated MP premium data, collected for the purpose of proposing updates to the MP GPCIs. Although we could have used the updated MP premium data obtained for the purposes of the proposed eighth GPCI update to propose updates to the specialty risk factors used in the calculation of MP RVUs, this would not be consistent with the policy we previously finalized in the CY 2016 PFS final rule with comment period. In that rule, we indicated that the specialty-specific risk factors would continue to be updated through notice and comment rulemaking every 5 years using updated premium data, but would remain unchanged between the 5-year reviews. Additionally, consistent with the statutory requirement at section 1848(e)(1)(C) of the Act, only one half of the adjustment to MP GPCIs would be applied for CY 2017 based on the new MP premium data. As such, we did not think it would be appropriate to propose to update the specialty risk factors for CY 2017 based on the updated MP premium data that is reflected in the proposed CY 2017 GPCI update. Therefore, we did not propose to update the specialty-risk factors based on the new premium data collected for the purposes of the 3-year GPCI update for CY 2017 at this time. However, we solicited comment on whether we should consider doing so, perhaps as early as for 2018, prior to the fourth review and update of MP RVUs that must occur no later than CY 2020.

The following is summary of the comments we received on whether we should consider updating the specialty-risk factors based on the new premium data collected for the purposes of the 3-year GPCI update, perhaps as early as for 2018, prior to the fourth review and update of MP RVUs that must occur no later than CY 2020.

Comment: We received few comments regarding this issue. Some commenters, including the RUC, recommended that CMS use the updated MP premium data collected as part of the CY 2017 GPCI update in the creation of the MP RVUs for CY 2017. One commenter stated that CMS should follow its normal process to update MP RVUs for CY 2020. Another commenter supported the technical and policy changes that CMS made related to the MP RVUs for the CY 2016 PFS, and appreciated CMS' reluctance to change direction a year later and use the updated malpractice premium data gathered for the purpose of the GPCI update, in advance of the next 5-year review of the MP RVUs, to propose updates to the specialty risk factors used in the calculation of MP RVUs. The commenter suggested that CMS consider using the updated data to update the specialty-risk factors in the MP RVU methodology as early as CY 2018, noting that by CY 2018, the adjustment of the malpractice GPCIs would be complete, so the potential disconnect in the use of the updated premium data would no longer be an issue.

Response: We appreciate the commenters' feedback. In response to the commenters who recommended that CMS use the updated MP premium data collected as part of the CY 2017 GPCI update in the creation of the MP RVUs for CY 2017, we reiterate that we did not propose to update the specialty-risk factors based on the new premium data collected for the purposes of the 3-year GPCI update for the CY 2017 MP RVUs. Instead, we solicited comment on whether we should consider doing so prior to the next 5-year interval, perhaps as early as for CY 2018. We will consider the possibility of using the updated MP data to update the specialty risk factors used in the calculation of the MP RVUs prior to the next 5-year update in future rulemaking.

Comment: One commenter stated that CPT code 93355 should be added to the MP RVUs Invasive Cardiology Outside of Surgical Range list so that the surgical risk factor is applied when calculating the MP RVU.

Response: We did not previously propose to include this code on the list of Invasive Cardiology Outside of Surgical Range when we updated MP risk factors for CY 2015 and we did not propose the change in the CY 2017 PFS proposed rule. We will consider that request for future rulemaking in conjunction with the next update of MP risk factors.

C. Medicare Telehealth Services

1. Billing and Payment for Telehealth Services

Several conditions must be met for Medicare to make payments for telehealth services under the PFS. The service must be on the list of Medicare telehealth services and meet all of the following additional requirements:

• The service must be furnished via an interactive telecommunications system.
• The service must be furnished by a physician or other authorized practitioner.
• The service must be furnished to an eligible telehealth individual.
• The individual receiving the service must be located in a telehealth originating site.

When all of these conditions are met, Medicare pays a facility fee to the originating site and makes a separate payment to the distant site practitioner furnishing the service.

Section 1834(m)(4)(F)(i) of the Act defines Medicare telehealth services to include professional consultations, office visits, office psychiatry services, and any additional service specified by the Secretary, when furnished via a telecommunications system. We first implemented this statutory provision, which was effective October 1, 2001, in the CY 2002 PFS final rule with comment period (66 FR 55246). We established a process for annual updates to the list of Medicare telehealth services as required by section 1834(m)(4)(F)(ii) of the Act in the CY 2003 PFS final rule with comment period (67 FR 79988).

As specified at § 410.78(b), we generally require that a telehealth service be furnished via an interactive telecommunications system. Under § 410.78(a)(3), an interactive telecommunications system is defined as multimedia communications equipment that includes, at a minimum, audio and video equipment permitting two-way, real-time interactive communication between the patient and distant site physician or practitioner.

Telephones, facsimile machines, and stand-alone electronic mail systems do not meet the definition of an interactive telecommunications system. An interactive telecommunications system is generally required as a condition of payment; however, section 1834(m)(1) of the Act allows the use of asynchronous “store-and-forward” technology when the originating site is part of a federal telemedicine demonstration program in Alaska or Hawaii. As specified in § 410.78(a)(1), asynchronous store-and-forward is the transmission of medical information from an originating site for review by the distant site physician or practitioner at a later time.

Medicare telehealth services may be furnished to an eligible telehealth individual notwithstanding the fact that the practitioner furnishing the telehealth service is not at the same location as the beneficiary. An eligible telehealth individual is an individual enrolled under Part B who receives a telehealth service furnished at a telehealth originating site.

Practitioners furnishing Medicare telehealth services are reminded that these services are subject to the same non-discrimination laws as other services, including the effective communication requirements for persons with disabilities of section 504 of the Rehabilitation Act and language access for persons with limited English proficiency, as required under Title VI of the Civil Rights Act of 1964. For more information, see http://www.hhs.gov/ocr/civilrights/resources/specialtopics/hospitalcommunication.

Practitioners furnishing Medicare telehealth services submit claims for telehealth services to the MACs that process claims for the service area where their distant site is located. Section 1834(m)(2)(A) of the Act requires that a practitioner who furnishes a telehealth service to an eligible telehealth individual be paid an amount equal to the amount that the practitioner would have been paid if the service had been furnished without the use of a telecommunications system.

Originating sites, which can be one of several types of sites specified in the statute where an eligible telehealth individual is located at the time the service is being furnished via a telecommunications system, are paid a facility fee under the PFS for each Medicare telehealth service. The statute specifies both the types of entities that can serve as originating sites and the geographic qualifications for originating sites. With regard to geographic qualifications, § 410.78(b)(4) limits originating sites to those located in rural health professional shortage areas (HPSAs) or in a county that is not included in a metropolitan statistical area (MSA).

Historically, we have defined rural HPSAs to be those located outside of MSAs. Effective January 1, 2014, we modified the regulations regarding originating sites to define rural HPSAs as those located in rural census tracts as determined by the Federal Office of Rural Health Policy of the Health Resources and Services Administration (HRSA) (78 FR 74811). Defining “rural” to include geographic areas located in rural census tracts within MSAs allows for broader inclusion of sites within HPSAs as telehealth originating sites. Adopting the more precise definition of “rural” for this purpose expands access to health care services for Medicare beneficiaries located in rural areas. HRSA has developed a Web site tool to provide assistance to potential originating sites to determine their geographic status. To access this tool, see the CMS Web site at https://www.cms.gov/Medicare/Medicare-General-Information/Telehealth/index.html.

An entity participating in a federal telemedicine demonstration project that has been approved by, or received funding from, the Secretary as of December 31, 2000 is eligible to be an originating site regardless of its geographic location.

Effective January 1, 2014, we also changed our policy so that geographic status for an originating site would be established and maintained on an annual basis, consistent with other telehealth payment policies (78 FR 74400). Geographic status for Medicare telehealth originating sites for each calendar year is now based upon the status of the area as of December 31 of the prior calendar year.

For a detailed history of telehealth payment policy, see 78 FR 74399.

2. Adding Services to the List of Medicare Telehealth Services

As noted previously, in the CY 2003 PFS final rule (67 FR 79988), we established a process for adding services to or deleting services from the list of Medicare telehealth services. This process provides the public with an ongoing opportunity to submit requests for adding services. Under this process, we assign any qualifying request to make additions to the list of telehealth services to one of two categories. Revisions to criteria that we use to review requests in the second category were finalized in the CY 2012 PFS final rule (76 FR 73102). The two categories are:

• Category 1: Services that are similar to professional consultations, office visits, and office psychiatry services that are currently on the list of telehealth services. In reviewing these requests, we look for similarities between the requested and existing telehealth services for the roles of, and interactions among, the beneficiary, the physician (or other practitioner) at the distant site and, if necessary, the telepresenter, a practitioner who is present with the beneficiary in the originating site. We also look for similarities in the telecommunications system used to deliver the service; for example, the use of interactive audio and video equipment.
• Category 2: Services that are not similar to the current list of telehealth services. Our review of these requests includes an assessment of whether the service is accurately described by the corresponding code when furnished via telehealth and whether the use of a telecommunications system to furnish the service produces demonstrated clinical benefit to the patient. Submitted evidence should include both a description of relevant clinical studies that demonstrate the service furnished by telehealth to a Medicare beneficiary improves the diagnosis or treatment of an illness or injury or improves the functioning of a malformed body part, including dates and findings, and a list and copies of published peer reviewed articles relevant to the service when furnished via telehealth. Our evidentiary standard of clinical benefit does not include minor or incidental benefits.

Some examples of clinical benefit include the following:

• Ability to diagnose a medical condition in a patient population without access to clinically appropriate in-person diagnostic services.
• Treatment option for a patient population without access to clinically appropriate in-person treatment options.
• Reduced rate of complications.
• Decreased rate of subsequent diagnostic or therapeutic interventions (for example, due to reduced rate of recurrence of the disease process).
• Decreased number of future hospitalizations or physician visits.
• More rapid beneficial resolution of the disease process treatment.
• Decreased pain, bleeding, or other quantifiable symptom.
• Reduced recovery time.

For the list of telehealth services, see the CMS Web site at https://www.cms.gov/Medicare/Medicare-General-Information/Telehealth/index.html. Requests to add services to the list of Medicare telehealth services must be submitted and received no later than December 31 of each calendar year to be considered for the next rulemaking cycle. For example, qualifying requests submitted before the end of CY 2016 will be considered for the CY 2018 proposed rule. Each request to add a service to the list of Medicare telehealth services must include any supporting documentation the requester wishes us to consider as we review the request. Because we use the annual PFS rulemaking process as a vehicle for making changes to the list of Medicare telehealth services, requesters should be advised that any information submitted is subject to public disclosure for this purpose. For more information on submitting a request for an addition to the list of Medicare telehealth services, including where to mail these requests, see the CMS Web site at https://www.cms.gov/Medicare/Medicare-General-Information/Telehealth/index.html.

3. Submitted Requests To Add Services to the List of Telehealth Services for CY 2017

Under our existing policy, we add services to the telehealth list on a category 1 basis when we determine that they are similar to services on the existing telehealth list for the roles of, and interactions among, the beneficiary, physician (or other practitioner) at the distant site and, if necessary, the telepresenter. As we stated in the CY 2012 final rule with comment period (76 FR 73098), we believe that the category 1 criteria not only streamline our review process for publicly requested services that fall into this category, but also expedite our ability to identify codes for the telehealth list that resemble those services already on this list.

We received several requests in CY 2015 to add various services as Medicare telehealth services effective for CY 2017. The following presents a discussion of these requests, and our decisions regarding additions to the CY 2017 telehealth list. Of the requests received, we found that four services were sufficiently similar to ESRD-related services currently on the telehealth list to qualify on a category 1 basis. Therefore, we proposed to add the following services to the telehealth list on a category 1 basis for CY 2017:

• CPT codes 90967 (End-stage renal disease (ESRD) related services for dialysis less than a full month of service, per day; for patients younger than 2 years of age; 90968 (End-stage renal disease (ESRD) related services for dialysis less than a full month of service, per day; for patients 2-11 years of age; 90969 (End-stage renal disease (ESRD) related services for dialysis less than a full month of service, per day; for patients 12-19 years of age); and 90970 (End-stage renal disease (ESRD) related services for dialysis less than a full month of service, per day; for patients 20 years of age and older).

As we indicated in the CY 2015 final rule with comment period (80 FR 41783), for the ESRD-related services (CPT codes 90963-90966) added to the telehealth list for CY 2016, the required clinical examination of the catheter access site must be furnished face-to-face “hands on” (without the use of an interactive telecommunications system) by a physician, CNS, NP, or PA. This requirement also applies to CPT codes 90967-90970.

While we did not receive a specific request, we also proposed to add two advance care planning services to the telehealth list. We have determined that these services are similar to the annual wellness visits (HCPCS codes G0438 & G0439) currently on the telehealth list:

• CPT codes 99497 (advance care planning including the explanation and discussion of advance directives such as standard forms (with completion of such forms, when performed), by the physician or other qualified health care professional; first 30 minutes, face-to-face with the patient, family member(s), or surrogate); and 99498 (advance care planning including the explanation and discussion of advance directives such as standard forms (with completion of such forms, when performed), by the physician or other qualified health care professional; each additional 30 minutes (list separately in addition to code for primary procedure)).

We also received requests to add services to the telehealth list that do not meet our criteria for Medicare telehealth services. We did not propose to add the following procedures for observation care, emergency department visits, critical care E/M, psychological testing, and physical, occupational and speech therapy, for the reasons noted:

a. Observation Care: CPT Codes—

• 99217 (observation care discharge day management (this code is to be utilized to report all services provided to a patient on discharge from “observation status” if the discharge is on other than the initial date of “observation status.” To report services to a patient designated as “observation status” or “inpatient status” and discharged on the same date, use the codes for observation or inpatient care services [including admission and discharge services, 99234-99236 as appropriate.]));
• 99218 (initial observation care, per day, for the evaluation and management of a patient which requires these three key components: A detailed or comprehensive history; a detailed or comprehensive examination; and medical decision making that is straightforward or of low complexity. Counseling and coordination of care with other physicians, other qualified health care professionals, or agencies are provided consistent with the nature of the problem(s) and the patient's and family's needs. Usually, the problem(s) requiring admission to “observation status” are of low severity. Typically, 30 minutes are spent at the bedside and on the patient's hospital floor or unit);
• 99219 (initial observation care, per day, for the evaluation and management of a patient, which requires these three key components: A comprehensive history; a comprehensive examination; and medical decision making of moderate complexity. Counseling and coordination of care with other physicians, other qualified health care professionals, or agencies are provided consistent with the nature of the problem(s) and the patient's and family's needs. Usually, the problem(s) requiring admission to “observation status” are of moderate severity. Typically, 50 minutes are spent at the bedside and on the patient's hospital floor or unit);
• 99220 (initial observation care, per day, for the evaluation and management of a patient, which requires these three key components: A comprehensive history; a comprehensive examination; and medical decision making of high complexity. Counseling and coordination of care with other physicians, other qualified health care professionals, or agencies are provided consistent with the nature of the problem(s) and the patient's and family's needs. Usually, the problem(s) requiring admission to “observation status” are of high severity. Typically, 70 minutes are spent at the bedside and on the patient's hospital floor or unit);
• 99224 (subsequent observation care, per day, for the evaluation and management of a patient, which requires at least two of these three key components: Problem focused interval history; problem focused examination; medical decision making that is straightforward or of low complexity. Counseling and coordination of care with other physicians, other qualified health care professionals, or agencies are provided consistent with the nature of the problem(s) and the patient's and family's needs. Usually, the patient is stable, recovering, or improving. Typically, 15 minutes are spent at the bedside and on the patient's hospital floor or unit);
• 99225 (subsequent observation care, per day, for the evaluation and management of a patient, which requires at least two of these three key components: An expanded problem focused interval history; an expanded problem focused examination; medical decision making of moderate complexity. Counseling and coordination of care with other physicians, other qualified health care professionals, or agencies are provided consistent with the nature of the problem(s) and the patient's and family's needs. Usually, the patient is responding inadequately to therapy or has developed a minor complication. Typically, 25 minutes are spent at the bedside and on the patient's hospital floor or unit);
• 99226 (subsequent observation care, per day, for the evaluation and management of a patient, which requires at least two of these three key components: A detailed interval history; a detailed examination; medical decision making of high complexity. Counseling and coordination of care with other physicians, other qualified health care professionals, or agencies are provided consistent with the nature of the problem(s) and the patient's and family's needs. Usually, the patient is unstable or has developed a significant complication or a significant new problem. Typically, 35 minutes are spent at the bedside and on the patient's hospital floor or unit);
• 99234 (observation or inpatient hospital care, for the evaluation and management of a patient including admission and discharge on the same date, which requires these three key components: A detailed or comprehensive history; a detailed or comprehensive examination; and medical decision making that is straightforward or of low complexity. Counseling and coordination of care with other physicians, other qualified health care professionals, or agencies are provided consistent with the nature of the problem(s) and the patient's and family's needs. Usually the presenting problem(s) requiring admission are of low severity. Typically, 40 minutes are spent at the bedside and on the patient's hospital floor or unit);
• 99235 (observation or inpatient hospital care, for the evaluation and management of a patient including admission and discharge on the same date, which requires these three key components: A comprehensive history; a comprehensive examination; and medical decision making of moderate complexity. Counseling and coordination of care with other physicians, other qualified health care professionals, or agencies are provided consistent with the nature of the problem(s) and the patient's and family's needs. Usually the presenting problem(s) requiring admission are of moderate severity. Typically, 50 minutes are spent at the bedside and on the patient's hospital floor or unit);
• 99236 (observation or inpatient hospital care, for the evaluation and management of a patient including admission and discharge on the same date, which requires these three key components: A comprehensive history; a comprehensive examination; and medical decision making of high complexity. Counseling and coordination of care with other physicians, other qualified health care professionals, or agencies are provided consistent with the nature of the problem(s) and the patient's and family's needs. Usually the presenting problem(s) requiring admission are of high severity. Typically, 55 minutes are spent at the bedside and on the patient's hospital floor or unit);

The request to add these observation services referenced various studies supporting the use of observation units. The studies indicated that observation units provide safe, cost effective care to patients that need ongoing evaluation and treatment beyond the emergency department visit by having reduced hospital admissions, shorter lengths of stay, increased safety and reduced cost. Additional studies cited indicated that observation units reduce the work load on emergency department physicians, and reduce emergency department overcrowding.

In the CY 2005 PFS proposed rule (69 FR 47510), we considered a request but did not propose to add the observation CPT codes 99217-99220 to the list of Medicare telehealth services on a category two basis for the reasons described in that rule. The most recent request did not include any information that would cause us to question the previous evaluation under the category one criterion, which has not changed, regarding the significant differences in patient acuity between these services and services on the telehealth list. While the request included evidence of the general benefits of observation units, it did not include specific information demonstrating that the services described by these codes provided clinical benefit when furnished via telehealth, which is necessary for us to consider these codes on a category two basis. Therefore, we did not propose to add these services to the list of approved telehealth services.

b. Emergency Department Visits: CPT Codes—

• 99281 (emergency department visit for the evaluation and management of a patient, which requires these three key components: A problem focused history; a problem focused examination; and straightforward medical decision making. Counseling and coordination of care with other physicians, other qualified health care professionals, or agencies are provided consistent with the nature of the problem(s) and the patient's and family's needs. Usually, the presenting problem(s) are self-limited or minor);
• 99282 (emergency department visit for the evaluation and management of a patient, which requires these three key components: An expanded problem focused history; an expanded problem focused examination; and medical decision making of low complexity. Counseling and coordination of care with other physicians, other qualified health care professionals, or agencies are provided consistent with the nature of the problem(s) and the patient's and family's needs. Usually, the presenting problem(s) are of low to moderate severity);
• 99283 (emergency department visit for the evaluation and management of a patient, which requires these three key components: An expanded problem focused history; an expanded problem focused examination; and medical decision making of moderate complexity. Counseling and coordination of care with other physicians, other qualified health care professionals, or agencies are provided consistent with the nature of the problem(s) and the patient's and family's needs. Usually, the presenting problem(s) are of moderate severity);
• 99284 (emergency department visit for the evaluation and management of a patient, which requires these three key components: A detailed history; a detailed examination; and medical decision making of moderate complexity. Counseling and coordination of care with other physicians, other qualified health care professionals, or agencies are provided consistent with the nature of the problem(s) and the patient's and family's needs. Usually, the presenting problem(s) are of high severity, and require urgent evaluation by the physician, or other qualified health care professionals but do not pose an immediate significant threat to life or physiologic function); and
• 99285 (emergency department visit for the evaluation and management of a patient, which requires these three key components within the constraints imposed by the urgency of the patient's clinical condition and mental status: A comprehensive history; a comprehensive examination; and medical decision making of high complexity. Counseling and coordination of care with other physicians, other qualified health care professionals, or agencies are provided consistent with the nature of the problem(s) and the patient's and family's needs. Usually, the presenting problem(s) are of high severity and pose an immediate significant threat to life or physiologic function).

In the CY 2005 PFS proposed rule (69 FR 47510), we considered a request but did not propose to add the emergency department visit CPT codes 99281-99285 to the list of Medicare telehealth services for the reasons described in that rule.

The current request to add the emergency department E/M services stated that the codes are similar to outpatient visit codes (CPT codes 99201-99215) that have been on the telehealth list since CY 2002. As we noted in the CY 2005 PFS final rule, while the acuity of some patients in the emergency department might be the same as in a physician's office; we believe that, in general, more acutely ill patients are more likely to be seen in the emergency department, and that difference is part of the reason there are separate codes describing evaluation and management visits in the Emergency Department setting. The practice of emergency medicine often requires frequent and fast-paced patient reassessments, rapid physician interventions, and sometimes the continuous physician interaction with ancillary staff and consultants. This work is distinctly different from the pace, intensity, and acuity associated with visits that occur in the office or outpatient setting. Therefore, we did not propose to add these services to the list of approved telehealth services on a category one basis.

The requester did not provide any studies supporting the clinical benefit of managing emergency department patients with telehealth which is necessary for us to consider these codes on a category two basis. Therefore, we did not propose to add these services to the list of approved telehealth services on a category two basis.

Many requesters of additions to the telehealth list urged us to consider the potential value of telehealth for providing beneficiaries access to needed expertise. We note that if clinical guidance or advice is needed in the emergency department setting, a consultation may be requested from an appropriate source, including consultations that are currently included on the list of telehealth services.

c. Critical Care Evaluation and Management: CPT Codes—

• 99291 (critical care, evaluation and management of the critically ill or critically injured patient; first 30-74 minutes); and 99292 (critical care, evaluation and management of the critically ill or critically injured patient; each additional 30 minutes (list separately in addition to code for primary service).

We previously considered and rejected adding these codes to the list of Medicare telehealth services in the CY 2009 PFS final rule (74 FR 69744) on a category 1 basis because, due to the acuity of critically ill patients, we did not believe critical care services are similar to any services on the current list of Medicare telehealth services. In that rule, we said that critical care services must be evaluated as category 2 services. Because we considered critical care services under category 2, we needed to evaluate whether these are services for which telehealth can be an adequate substitute for a face-to-face encounter, based on the category 2 criteria at the time of that request. We had no evidence suggesting that the use of telehealth could be a reasonable surrogate for the face-to-face delivery of this type of care.

The American Telemedicine Association (ATA) submitted a new request for CY 2016 that cited several studies to support adding these services on a category 2 basis. To qualify under category 2, we would need evidence that the service furnished via telehealth is still described accurately by the requested code and produces a clinical benefit for the patient via telehealth. However, in reviewing the information provided by the ATA and a study titled, “Impact of an Intensive Care Unit Telemedicine Program on Patient Outcomes in an Integrated Health Care System,” published July 2014 in JAMA Internal Medicine, which found no evidence that the implementation of ICU telemedicine significantly reduced mortality rates or hospital length of stay, which could be indicators of clinical benefit. Therefore, we stated that we do not believe that the submitted evidence demonstrates a clinical benefit to patients. Therefore, we did not propose to add these services on a category 2 basis to the list of Medicare telehealth services for CY 2016 (80 FR 71061).

This year, requesters cited additional studies to support adding critical care services to the Medicare telehealth list on a category 2 basis. Eight of the studies dealt with telestroke and one with teleneurology. Telestroke is an approach that allows a neurologist to provide remote treatment to vascular stroke victims. Teleneurology offers consultations for neurological problems from a remote location. It may be initiated by a physician or a patient, for conditions such as headaches, dementia, strokes, multiple sclerosis and epilepsy.

However, according to the literature, the management of stroke via telehealth requires more than a single practitioner and is distinct from the work described by the above E/M codes, 99291 and 99292. One additional study cited involved pediatric patients, while another noted that the Department of Defense has used telehealth to provide critical care services to hospitals in Guam for many years. Another reference study indicated that consulting intensivists thought that telemedicine consultations were superior to telephone consultations. In all of these cases, we believe the evidence demonstrates that interaction between these patients and distant site practitioners can have clinical benefit. However, we do not agree that the kinds of services described in the studies are those that are included in the above critical care E/M codes 99291 and 99292. We note that CPT guidance makes clear that a variety of other services are bundled into the payment rates for critical care, including gastric intubations and vascular access procedures among others We do not believe these kinds of services are furnished via telehealth. Public comments, included cited studies, can be viewed at https://www.regulations.gov/#!documentDetail;D=CMS-2015-0081-0002. Therefore, we did not propose to add CPT codes 99291 or 99292 to the list of Medicare telehealth services for CY 2017.

However, we are persuaded by the requests that we recognize the potential benefit of critical care consultation services that are furnished remotely. We note that there are currently codes on the telehealth list that could be reported when consultation services are furnished to critically ill patients. In consideration of these public requests, we recognize that there may be greater resource costs involved in furnishing these services relative to the existing telehealth consultation codes. We also agree with the requesters that there may be potential benefits of remote care by specialists for these patients. For these reasons, we think it would be advisable to create a coding distinction between telehealth consultations for critically ill patients, for example stroke patients, relative to telehealth consultations for other hospital patients. Such a coding distinction would allow us to recognize the additional resource costs in terms of time and intensity involved in furnishing such services, under the conditions where remote, intensive consultation is required to provide access to appropriate care for the critically ill patient. We recognize that the current set of E/M codes, including current CPT codes 99291 and 99292, may not adequately describe such services because current E/M coding presumes that the services are occurring in-person, in which case the expert care would be furnished in a manner described by the current codes for critical care.

Therefore, we proposed to make payment through new HCPCS codes G0508 and G0509, initial and subsequent, used to describe critical care consultations furnished via telehealth. This new coding would provide a mechanism to report an intensive telehealth consultation service, initial or subsequent, for the critically ill patient, such as a stroke patient, under the circumstance when a qualified health care professional has in-person responsibility for the patient but the patient benefits from additional services from a distant-site consultant specially trained in providing critical care services. We proposed limiting these services to once per day per patient. Like the other telehealth consultations, these services would be valued relative to existing E/M services.

More details on the new coding (G0508 and G0509) and valuation for these services are discussed in section II.L. of this final rule and the final RVUs for this service are included in Addendum B of this final rule, including a summary of the public comments we received and our responses to the comments. Like the other telehealth consultation codes, we proposed that these services would be added to the telehealth list and would be subject to the geographic and other statutory restrictions that apply to telehealth services.

d. Psychological Testing: CPT Codes—

• 96101 (psychological testing (includes psychodiagnostic assessment of emotionality, intellectual abilities, personality and psychopathology, e.g., MMPI, Rorschach, WAIS), per hour of the psychologist's or physician's time, both face-to-face time administering tests to the patient and time interpreting these test results and preparing the report);
• 96102 psychological testing (includes psychodiagnostic assessment of emotionality, intellectual abilities, personality and psychopathology, e.g., MMPI and WAIS), with qualified health care professional interpretation and report, administered by technician, per hour of technician time, face-to-face);
• 96118 Neuropsychological testing (e.g., Halstead-Reitan neuropsychological battery, Wechsler memory scales and Wisconsin card sorting test), per hour of the psychologist's or physician's time, both face-to-face time administering tests to the patient and time interpreting these test results and preparing the report); and,
• 96119 Neuropsychological testing (e.g., Halstead-Reitan neuropsychological battery, Wechsler memory scales and Wisconsin card sorting test), with qualified health care professional interpretation and report, administered by technician, per hour of technician time, face-to-face).

Requesters indicated that there is nothing in the Minnesota Multiphasic Personality Inventory (MMPI), the Rorschach inkblot test, the Wechsler Adult Intelligence Scale (WAIS), the Halstead-Reitan Neuropsychological Battery and Allied Procedures, or the Wisconsin Card Sorting Test (WCST), that cannot be done via telehealth nor is different than neurological tests done for Parkinson's disease, seizure medication side effects, gait assessment, nor any of the many neurological examinations done via telehealth with the approved outpatient office visit and inpatient visit CPT codes currently on the telehealth list. As an example, requesters indicated that the MPPI is administered by a computer, which generates a report that is interpreted by the clinical psychologist, and that the test requires no interaction between the clinician and the patient.

We previously considered the request to add these codes to the Medicare telehealth list in the CY 2015 final rule with comment period (79 FR 67600). We decided not to add these codes, indicating that these services are not similar to other services on the telehealth list because they require close observation of how a patient responds. We noted that the requesters did not submit evidence supporting the clinical benefit of furnishing these services via telehealth so that we could evaluate them on a category 2 basis. While we acknowledge that requesters believe that some of these tests require minimal, if any, interaction between the clinician and patient, we disagree. We continue to believe that successful completion of the tests listed as examples in these codes require the clinical psychologist to closely observe the patient's response, which cannot be performed via telehealth. Some patient responses, for example, sweating and fine tremors, may be missed when the patient and examiner are not in the same room. Therefore, we did not propose to add these services to the list of Medicare telehealth services for CY 2017.

e. Physical and Occupational Therapy and Speech-Language Pathology Services: CPT Codes—

• 92507 (treatment of speech, language, voice, communication, and auditory processing disorder; individual); and, 92508 (treatment of speech, language, voice, communication, and auditory processing disorder; group, 2 or more individuals); 92521 (evaluation of speech fluency (e.g., stuttering, cluttering)); 92522 (evaluation of speech sound production (e.g., articulation, phonological process, apraxia, dysarthria)); 92523 (evaluation of speech sound production (e.g., articulation, phonological process, apraxia, dysarthria); with evaluation of language comprehension and expression (e.g., receptive and expressive language)); 92524 (behavioral and qualitative analysis of voice and resonance); (evaluation of oral and pharyngeal swallowing function); 92526 (treatment of swallowing dysfunction or oral function for feeding); 92610 (evaluation of oral and pharyngeal swallowing function); CPT codes 97001 (physical therapy evaluation); 97002 (physical therapy re-evaluation); 97003 (occupational therapy evaluation); 97004 (occupational therapy re-evaluation); 97110 (therapeutic procedure, 1 or more areas, each 15 minutes; therapeutic exercises to develop strength and endurance, range of motion and flexibility); 97112 (therapeutic procedure, 1 or more areas, each 15 minutes; neuromuscular reeducation of movement, balance, coordination, kinesthetic sense, posture, or proprioception for sitting or standing activities); 97116 (therapeutic procedure, 1 or more areas, each 15 minutes; gait training (includes stair climbing)); 97532 (development of cognitive skills to improve attention, memory, problem solving (includes compensatory training), direct (one-on-one) patient contact, each 15 minutes); 97533 (sensory integrative techniques to enhance sensory processing and promote adaptive responses to environmental demands, direct (one-on-one) patient contact, each 15 minutes); 97535 (self-care/home management training (e.g., activities of daily living (adl) and compensatory training, meal preparation, safety procedures, and instructions in use of assistive technology devices/adaptive equipment) direct one-on-one contact, each 15 minutes); 97537 (community/work reintegration training (e.g., shopping, transportation, money management, avocational activities or work environment/modification analysis, work task analysis, use of assistive technology device/adaptive equipment), direct one-on-one contact, each 15 minutes); 97542 (wheelchair management (e.g., assessment, fitting, training), each 15 minutes); 97750 (physical performance test or measurement (e.g., musculoskeletal, functional capacity), with written report, each 15 minutes); 97755 (assistive technology assessment (e.g., to restore, augment or compensate for existing function, optimize functional tasks and maximize environmental accessibility), direct one-on-one contact, with written report, each 15 minutes); 97760 Orthotic(s) management and training (including assessment and fitting when not otherwise reported), upper extremity(s), lower extremity(s) and/or trunk, each 15 minutes); 97761 (prosthetic training, upper and lower extremity(s), each 15 minutes); and 97762 (checkout for orthotic/prosthetic use, established patient, each 15 minutes).

The statute defines who is an authorized practitioner of telehealth services. Physical therapists, occupational therapists and speech-language pathologists are not authorized practitioners of telehealth under section 1834(m)(4)(E) of the Act, as defined in section 1842(b)(18)(C) of the Act. Because the above services are predominantly furnished by physical therapists, occupational therapists and speech-language pathologists, we do not believe it would be appropriate to add them to the list of telehealth services at this time. One requester suggested that we can add telehealth practitioners without legislation, as evidenced by the addition of nutritional professionals. However, we do not believe we have such authority and note that nutritional professionals are included as practitioners in the definition at section 1834(b)(18)(C)(vi) of the Act, and thus, are within the statutory definition of telehealth practitioners. Therefore, we did not propose to add these services to the list of Medicare telehealth services for CY 2017.

In summary, we proposed to add the following codes to the list of Medicare telehealth services beginning in CY 2017 on a category 1 basis:

• ESRD-related services 90967 through 90970. The required clinical examination of the catheter access site must be furnished face-to-face “hands on” (without the use of an interactive telecommunications system) by a physician, CNS, NP, or PA.
• Advance care planning (CPT codes 99497 and 99498).
• Telehealth Consultations for a Patient Requiring Critical Care Services (G0508 and G0509).

The following is summary of the comments we received regarding the proposed addition of services to the list of Medicare telehealth services:

Comment: Many commenters supported one or more of our proposals to add ESRD-related services (CPT codes 90967, 90968, 90969 and 90970) and advance care planning services (CPT codes 99497 and 99498) to the list of Medicare telehealth services for CY 2017.

Response: We appreciate the commenters' support for the proposed additions to the list of Medicare telehealth services. After consideration of the public comments received, we are finalizing our proposal to add these services to the list of Medicare telehealth services for CY 2017 on a category 1 basis.

Comment: Many commenters also supported the proposal to make payment through new codes, initial and subsequent, used to describe critical care consultations furnished via telehealth. Commenters indicated that the codes will improve patient outcomes and quality of care.

Response: We thank the commenters for their support. We believe the new coding G0508 and G0509 would provide a mechanism to report an intensive telehealth consultation service, initial or subsequent, for the critically ill patient, for example a stroke patient, under the circumstance when a qualified health care professional has in-person responsibility for the patient but the patient benefits from additional services from a distant-site consultant specially trained in furnishing critical care services. After consideration of the public comments received, we are finalizing our proposal to add these critical care consultation services to the list of Medicare telehealth services for CY 2017 on a category 1 basis. We are finalizing these services as limited to once per day per patient.

We are also finalizing our proposal to make payment for these critical care consultation services through new codes G0508 and G0509, initial and subsequent, used to describe critical care consultations furnished via telehealth. More details on the new coding and valuation for these services are discussed in section II.L. of this final rule and the final RVUs for this service are included in Addendum B of this final rule. Like the other telehealth consultation codes, we proposed and are finalizing that these services would be added to the telehealth list and would be subject to the geographic and other statutory restrictions that apply to telehealth services.

Comment: Several commenters agreed with our decision not to add psychological and neuropsychological testing services to the telehealth list, noting that the face-to-face contact between the psychologist or technician and the beneficiary is critical for detecting behaviors related to test taking, such as movements or other nonverbal signals that could be missed by using current telehealth media.

A few commenters disagreed with our decision not to add psychological and neuropsychological testing services. Commenters cited general benefits, such as increased access to care, improved health outcomes, and as a remedy to address provider shortages. One commenter maintained that the requested codes are similar to many neurological examinations done via telehealth with the approved outpatient office visit and inpatient visit CPT codes currently on the telehealth list.

Response: As noted above, we previously considered the request to add these codes to the telehealth list, on a category 1 basis, in the CY 2015 final rule with comment period (79 FR 67600). We decided not to add these codes, indicating that these services are not similar to other services on the telehealth list because they require close observation of how a patient responds. Commenters provided no evidence of clinical benefit, which is necessary to support adding these services on a category 2 basis. Therefore, we are not adding these services to the list of Medicare list telehealth services for CY 2017.

Comment: A few commenters disagreed with our decision not to add observation care and emergency department visits. Commenters cited general benefits, such as improved quality of care, reduced physician workload, reduced emergency department overcrowding, and reduced shortage of available specialty services. Concerning CPT codes 99281-99283, one commenter indicated that none of these codes include what is categorized as a “detailed” or “comprehensive” history or exam; none of these codes include complexity in medical decision making that is categorized as “high;” and none of these codes include presenting problems of “high” or “high severity/immediate significant threat to life or physiological function.”

Response: As noted above, we previously considered and rejected adding these codes to the list of Medicare telehealth services in the CY 2005 PFS final rule (69 FR 66276) on a category 1 basis because of the difference in typical patient acuity relative to any services on the current list of Medicare telehealth services. While CPT codes 99281-99283 may not include a detailed or comprehensive history or exam or a high level of medical decision making, we do not agree that these codes are similar to outpatient visit codes (CPT codes 99201-99215) currently on the list of Medicare telehealth services. As previously stated, more acutely ill patients are more likely to be seen in the emergency department, and that difference is part of the reason there are separate codes describing evaluation and management visits in the Emergency Department setting. The work in an Emergency Department setting is distinctly different from the pace, intensity, and acuity associated with visits that occur in the office or outpatient setting. Commenters provided no evidence of clinical benefit for these services when furnished via telehealth specifically, which is necessary to support adding these services on a category 2 basis. Therefore, we are not adding these services to the list of Medicare telehealth services for CY 2017.

We remind stakeholders that if consultative telehealth services are required for patients where emergency department or observation care services would ordinarily be reported, multiple codes describing consultative services are currently on the telehealth list and can be used to bill for such telehealth services.

Comment: Concerning various services primarily furnished by physical therapists, occupational therapists, and speech-language pathologists, commenters recognized that a statutory change is required to allow such services to be added to the list of Medicare telehealth services.

Response: We appreciate commenters recognizing the statutory limitation on adding these services. Therefore, we are not adding these services to the list of Medicare telehealth services for CY 2017.

4. Place of Service (POS) Code for Telehealth Services

We have received multiple requests from various stakeholders to establish a POS code to identify services furnished via telehealth. These requests have come from other payers, but may also be related to confusion concerning whether to use the POS where the distant site physician is located or the POS where the patient is located. The process for establishing POS codes is managed by the POS Workgroup within CMS, is available for use by all payers, and is not contingent upon Medicare PFS rulemaking. We noted in the CY 2017 proposed rule (81 FR 46184) that, if such a POS code were created, in order to make it valid for use in Medicare, we would have to determine the appropriate payment rules associated with the code. Therefore, we proposed how a POS code for telehealth would be used under the PFS with the expectation that, if such a code is available, it would be used as early as January 1, 2017. We proposed that the physicians or practitioners furnishing telehealth services would be required to report the telehealth POS code to indicate that the billed service is furnished as a telehealth service from a distant site. As noted below, since the publication of the CY 2017 proposed rule, the telehealth POS code has been created.

Our requirement for physicians and practitioners to use the telehealth POS code to report that telehealth services were furnished from a distant site would improve payment accuracy and consistency in telehealth claims submission. Currently, for services furnished via telehealth, we have instructed practitioners to report the POS code that would have been reported had the service been furnished in person. However, some practitioners use the POS where they are located when the service is furnished, while others use the POS corresponding to the patient's location.

Under the PFS, the POS code determines whether a service is paid using the facility or non-facility practice expense relative value units (PE RVUs). The facility rate is paid when a service is furnished in a location where Medicare is making a separate facility payment to an entity other than the physician or practitioner that is intended to reflect the facility costs associated with the service (clinical staff, supplies and equipment). We note that in accordance with section 1834(m)(2)(B) of the Act, the payment amount for the telehealth facility fee paid to the originating site is a national fee, paid without geographic or site of service adjustments that generally are made for payments to different kinds of Medicare providers and suppliers. In the case of telehealth services, we believe that facility costs (clinical staff, supplies, and equipment) associated with furnishing the service would generally be incurred by the originating site, where the patient is located, and not by the practitioner at the distant site. The statute requires Medicare to pay a fee to the site that hosts the patient. This is analogous to the circumstances under which the facility PE RVUs are used to pay for services under the PFS. Therefore, we proposed to use the facility PE RVUs to pay for telehealth services reported by physicians or practitioners with the telehealth POS code. We note that there are only three codes on the telehealth list with a difference greater than 1.0 PE RVUs between the facility PE RVUs and the non-facility PE RVUs. We did not anticipate that this proposal would result in a significant change in the total payment for the majority of services on the telehealth list. Moreover, many practitioners already use a facility POS when billing for telehealth services (those that report the POS of the originating site where the beneficiary is located). The policy to use the telehealth POS code for telehealth services would not affect payment for telehealth services for these practitioners.

The POS code for telehealth would not apply to originating sites billing the facility fee. Originating sites are not furnishing a service via telehealth since the patient is physically present in the facility. Accordingly, the originating site would continue to use the POS code that applies to the type of facility where the patient is located.

We also proposed a change to § 414.22(b)(5)(i)(A) that addresses the PE RVUs used in different settings. These revisions would improve clarity regarding our current policies. Specifically, we proposed to amend this section to specify that the facility PE RVUs are paid for practitioner services furnished via telehealth under § 410.78. In addition, we proposed a change to resolve any potential ambiguity and clarify that payment under the PFS is made at the facility rate (facility PE RVUs) when services are furnished in a facility setting paid by Medicare, including in off-campus provider based departments. As proposed, the regulation reflected the policy being proposed, for CY 2017 only, to pay the physician the nonfacility rate for services furnished in an off-campus provider based department that was not excepted under section 603 of the Bipartisan Budget Act of 2015. Finally, to streamline the existing regulation, we also proposed to delete § 414.32 of our regulation that refers to the calculation of payments for certain services prior to 2002.

The following is summary of the comments we received regarding the proposal to use a POS code for services furnished via telehealth:

Comment: Many commenters supported the proposal to use the POS code for telehealth, indicating that it would clarify and simplify billing requirements, improve payment accuracy and consistency in telehealth claims submissions, and provide more reliable data regarding telehealth services.

Response: We appreciate the support for this proposal.

Comment: One commenter asked us to reconsider the proposal, noting that the AMA's CPT Editorial Panel has adopted a telehealth modifier for those medical services that are currently covered telehealth services by Medicare or other payers, which obviates the need for the POS code.

Response: The POS code was requested by other payers, and we continue to believe that adopting it for use in the Medicare program would provide consistency in reporting and identifying services furnished via telehealth. We have had longstanding HCPCS modifiers for telehealth. While these modifiers were not adopted by CPT, they have been available for use by other payers. Despite the availability of these HCPCS modifiers noting telehealth services, payers have requested creation of the new POS code. Therefore, we do not understand why introduction of a new CPT modifier as opposed to a HCPCS modifier would obviate the need for a POS code. Instead, we agree with other payers that the POS code would provide consistency in reporting and identifying services furnished via telehealth, since it eliminates the need for service-specific rules regarding appropriate POS reporting for telehealth services.

Comment: Another commenter stated that use of the POS code, or originating site restrictions, would place additional administrative barriers for telepsychiatric access.

Response: We note that the POS is a required field on the professional claim, regardless of whether the service is furnished via telehealth. Since a selection needs to be made, we believe that requiring the selection of a specific code is no more burdensome than requiring the claim to specify the POS appropriate to either the setting of the telehealth patient or the setting of the distant site practitioner. The POS code does not entail any new originating site restrictions.

Comment: Various commenters asked for clarification of the following:

• Whether the POS code would replace the GT modifier.
• Whether the description of telehealth as a service furnished via an interactive audio and video telecommunications system applies to the POS code as it does to the GT modifier.
• How to ensure proper payment when the distant site practitioner is at a facility, but the patient is not.

Response: Under current policy, use of the GT and GQ modifiers certifies that the service meets the telehealth requirements, and would continue to be required. The POS code would be used in addition to the GT and GQ modifiers. We did not propose to implement a change in the requirement to use either the GT and GQ modifier because at the time of the proposed rule, we did not know whether the telehealth POS code would be made effective for January 1, 2017. However, because under our proposal the POS code would serve to identify telehealth services furnished under section 1834(m) of the Act via an interactive audio and video telecommunications system, we believe that we should consider eliminating the required use of the GT and GQ telehealth modifiers, and we may revisit this question through future rulemaking. Like the modifiers, use of the POS code certifies that the service meets the telehealth requirements. Distant site providers will be paid using the facility PE RVUs, regardless of their location. The setting of the patient does not affect the payment to the distant site provider.

Comment: Commenters also asked for clarification that the proposal to adopt the telehealth POS relates solely to payment, and not to licensure requirements. The commenter noted that practitioners who furnish telehealth services must adhere to the standard of care and licensure rules, regulations and laws of the state where the patient is located, just as the practitioner would in a traditional face-to-face encounter.

Response: The commenters are correct that the purpose of our POS proposal is to assist in determining proper payment. It will also help us to accurately track telehealth utilization and spending. The proposal to adopt the telehealth POS code has no bearing on state licensure requirements or other state regulations. We appreciate the commenters' request for clarification.

Comment: Several commenters supported the proposal to use the facility PE RVUs for telehealth services. One commenter said paying some telehealth services at non-facility rates creates undesirable financial incentives to prefer telehealth services over services that are furnished in person at the originating site.

Response: We appreciate the support for the proposal and agree with the commenter's articulation regarding the importance of developing payment rates that reflect the relative resource costs of furnishing the services and that do not create unintended financial incentives.

Comment: Many other commenters opposed the proposal. Commenters stated that it would result in lower fees for telehealth services furnished by psychologists. Commenters also stated that PE costs increase for services furnished via telehealth due to the costs of HIPAA-compliant telecommunication equipment.

One commenter remarked that use of a POS code should not be the basis for reducing payments and that many codes would experience a significant payment change. The commenter noted that a 1.0 RVU reduction would result in a $36 payment reduction for the service. One commenter stated CMS should propose budget neutral PE and originating site fees, based on data, for CY 2018. One commenter noted that there are no facility PE RVUs for several codes.

Response: We do not believe that use of the telehealth POS code produces a significant payment change in the vast majority of circumstances. For distant site practitioners who are already paid using the facility PE RVUs and for services where there is no payment difference between the facility and non-facility PE RVUs, there will be no change in payment as a result of the telehealth POS code.

There is utilization data for 56 of the 81 codes on the telehealth list. For these codes, 20 are not paid differently based on site of service, and 27 codes are paid differently by fewer than 0.5 RVUs. There are only three codes on the telehealth list with a difference greater than 1.0 PE RVUs between the facility PE RVUs and the non-facility PE RVUs.

Concerning psychotherapy and psychological testing services, we note that for the vast majority of psychiatric services the difference between the two rates is very small. For example, the difference between the facility and non-facility national rates for 45 minutes of psychotherapy is 0.02 RVUs per service: Less than $1.00. The differences between the facility PE RVUs and non-facility PE RVUS ranges from 0.01-0.03 RVUs for nine of the psychological testing codes on the Medicare telehealth list, and 0.12 RVUs lower for two other codes. We do not consider these reductions significant, nor do we have any evidence that practice expense costs are greater for furnishing such services via telehealth than for furnishing a face-to-face service. Commenters provided no evidence that practice expense costs for services furnished via telehealth are greater, due to the requirement for HIPAA-compliant equipment, than for furnishing in-person services, even in the facility setting.

There are a few HCPCS codes on the telehealth list that do not have a calculated facility PE RVU. For these services, the non-facility PE RVUs would serve as a proxy, and therefore, there would be no payment change for these codes.

Finally, we note that the originating site facility fee is established by statute (section 1834(m)(2)(B) of the Act) and is not affected by this proposal.

We note that we believe that payment using the facility PE RVUs for telehealth services is consistent our belief that the direct practice expense costs are generally incurred at the location of the beneficiary and not by the distant site practitioner. After reviewing the current list of telehealth services in the context of the comments, we continue to believe this is accurate.

After consideration of the public comments received, we are finalizing our proposal to use the POS code for telehealth and to use the facility PE RVUs to pay for telehealth service reported by physicians or practitioners with the telehealth POS code for CY 2017. However, we understand commenters' concerns and will consider the concerns regarding use of the facility payment rate as we monitor utilization of telehealth services. We will welcome information from stakeholders regarding any potential unintended consequences of the payment policy. We will also consider the applicability of the facility rate to any codes newly added to the list of telehealth services.

We have updated the POS code list on our Web site at https://www.cms.gov/Medicare/Coding/place-of-service-codes/Place_of_Service_Code_Set.html to include POS 02: Telehealth (Descriptor: The location where health services and health related services are provided or received, through telecommunication technology). The new code will be used for services furnished on or after January 1, 2017.

We are finalizing proposed revisions to our regulation at § 414.22(b)(5)(i)(A) that addresses the PE RVUs used in different settings as described above, except that we are not finalizing the proposed change that would have resulted in the payment of the nonfacility rate for services furnished in off-campus provider based departments that are not excepted under Section 603 of the Bipartisan Budget Act of 2015 since we are finalizing that payments to such non-excepted PBDs will be made under the PFS. In a separate interim final rule with comment period issued in conjunction with the CY 2017 OPPS/ASC final rule with comment period (see Medicare Program: Hospital Outpatient Prospective Payment and Ambulatory Surgical Center Payment Systems and Quality Reporting Programs; Organ Procurement Organization Reporting and Communication; Transplant Outcome Measures and Documentation Requirements; Electronic Health Record (EHR) Incentive Programs; Payment to Certain Off-Campus Outpatient Departments of a Provider; Hospital Value-Based Purchasing (VBP) Program; Establishment of Physician Fee Schedule Payment Rates for Nonexcepted Items and Services Billed by Applicable Departments of a Hospital), we are finalizing other payment policies for nonexcepted items and services furnished by such non-excepted off-campus provider based departments. Accordingly, physicians furnishing services in such provider-based departments will continue to be paid the facility rate. We are also finalizing the proposal to delete § 414.32 of our regulation that refers to the calculation of payments for certain services prior to 2002.

We remind the public that we are currently soliciting requests to add services to the list of Medicare telehealth services. To be considered during PFS rulemaking for CY 2018, these requests must be submitted and received by December 31, 2016. Each request to add a service to the list of Medicare telehealth services must include any supporting documentation the requester wishes us to consider as we review the request. For more information on submitting a request for an addition to the list of Medicare telehealth services, including where to mail these requests, we refer readers to the CMS Web site at https://www.cms.gov/Medicare/Medicare-General-Information/Telehealth/index.html.

5. Telehealth Originating Site Facility Fee Payment Amount Update

Section 1834(m)(2)(B) of the Act establishes the Medicare telehealth originating site facility fee for telehealth services furnished from October 1, 2001 through December 31, 2002, at $20.00. For telehealth services furnished on or after January 1 of each subsequent calendar year, the telehealth originating site facility fee is increased by the percentage increase in the MEI as defined in section 1842(i)(3) of the Act. The originating site facility fee for telehealth services furnished in CY 2016 is $25.10. The MEI increase for 2017 is 1.2 percent and is based on the most recent historical update through 2016Q2 (1.6 percent), and the most recent historical MFP through calendar year 2015 (0.4 percent). Therefore, for CY 2017, the payment amount for HCPCS code Q3014 (Telehealth originating site facility fee) is 80 percent of the lesser of the actual charge or $25.40. The Medicare telehealth originating site facility fee and the MEI increase by the applicable time period is shown in Table 6.

D. Potentially Misvalued Services Under the Physician Fee Schedule

1. Background

Section 1848(c)(2)(B) of the Act directs the Secretary to conduct a periodic review, not less often than every 5 years, of the RVUs established under the PFS. Section 1848(c)(2)(K) of the Act requires the Secretary to periodically identify potentially misvalued services using certain criteria and to review and make appropriate adjustments to the relative values for those services. Section 1848(c)(2)(L) to the Act also requires the Secretary to develop a process to validate the RVUs of certain potentially misvalued codes under the PFS, using the same criteria used to identify potentially misvalued codes, and to make appropriate adjustments.

As discussed in section II.B. of this final rule, each year we develop appropriate adjustments to the RVUs taking into account recommendations provided by the American Medical Association/Specialty Society Relative Value Scale Update Committee (RUC), the Medicare Payment Advisory Commission (MedPAC), and others. For many years, the RUC has provided us with recommendations on the appropriate relative values for new, revised, and potentially misvalued PFS services. We review these recommendations on a code-by-code basis and consider these recommendations in conjunction with analyses of other data, such as claims data, to inform the decision-making process as authorized by the law. We may also consider analyses of work time, work RVUs, or direct PE inputs using other data sources, such as Department of Veteran Affairs (VA), National Surgical Quality Improvement Program (NSQIP), the Society for Thoracic Surgeons (STS), and the Physician Quality Reporting System (PQRS) databases. In addition to considering the most recently available data, we also assess the results of physician surveys and specialty recommendations submitted to us by the RUC for our review. We also consider information provided by other stakeholders. We conduct a review to assess the appropriate RVUs in the context of contemporary medical practice. We note that section 1848(c)(2)(A)(ii) of the Act authorizes the use of extrapolation and other techniques to determine the RVUs for physicians' services for which specific data are not available and requires us to take into account the results of consultations with organizations representing physicians who provide the services. In accordance with section 1848(c) of the Act, we determine and make appropriate adjustments to the RVUs.

In its March 2006 Report to the Congress (http://www.medpac.gov/documents/reports/Mar06_EntireReport.pdf?sfvrsn=0), MedPAC discussed the importance of appropriately valuing physicians' services, noting that misvalued services can distort the market for physicians' services, as well as for other health care services that physicians order, such as hospital services. In that same report MedPAC postulated that physicians' services under the PFS can become misvalued over time. MedPAC stated, “When a new service is added to the physician fee schedule, it may be assigned a relatively high value because of the time, technical skill, and psychological stress that are often required to furnish that service. Over time, the work required for certain services would be expected to decline as physicians become more familiar with the service and more efficient in furnishing it.” We believe services can also become overvalued when PE declines. This can happen when the costs of equipment and supplies fall, or when equipment is used more frequently than is estimated in the PE methodology, reducing its cost per use. Likewise, services can become undervalued when physician work increases or PE rises.

As MedPAC noted in its March 2009 Report to Congress (http://www.medpac.gov/documents/reports/march-2009-report-to-congress-medicare-payment-policy.pdf?sfvrsn=0), in the intervening years since MedPAC made the initial recommendations, CMS and the RUC have taken several steps to improve the review process. Also, section 1848(c)(2)(K)(ii) of the Act augments our efforts by directing the Secretary to specifically examine, as determined appropriate, potentially misvalued services in the following categories:

• Codes that have experienced the fastest growth.
• Codes that have experienced substantial changes in practice expenses.
• Codes that describe new technologies or services within an appropriate time period (such as 3 years) after the relative values are initially established for such codes.
• Codes which are multiple codes that are frequently billed in conjunction with furnishing a single service.
• Codes with low relative values, particularly those that are often billed multiple times for a single treatment.
• Codes that have not been subject to review since implementation of the fee schedule.
• Codes that account for the majority of spending under the physician fee schedule.
• Codes for services that have experienced a substantial change in the hospital length of stay or procedure time.
• Codes for which there may be a change in the typical site of service since the code was last valued.
• Codes for which there is a significant difference in payment for the same service between different sites of service.
• Codes for which there may be anomalies in relative values within a family of codes.
• Codes for services where there may be efficiencies when a service is furnished at the same time as other services.
• Codes with high intra-service work per unit of time.
• Codes with high practice expense relative value units.
• Codes with high cost supplies.
• Codes as determined appropriate by the Secretary.

Section 1848(c)(2)(K)(iii) of the Act also specifies that the Secretary may use existing processes to receive recommendations on the review and appropriate adjustment of potentially misvalued services. In addition, the Secretary may conduct surveys, other data collection activities, studies, or other analyses, as the Secretary determines to be appropriate, to facilitate the review and appropriate adjustment of potentially misvalued services. This section also authorizes the use of analytic contractors to identify and analyze potentially misvalued codes, conduct surveys or collect data, and make recommendations on the review and appropriate adjustment of potentially misvalued services. Additionally, this section provides that the Secretary may coordinate the review and adjustment of any RVU with the periodic review described in section 1848(c)(2)(B) of the Act. Section 1848(c)(2)(K)(iii)(V) of the Act specifies that the Secretary may make appropriate coding revisions (including using existing processes for consideration of coding changes) that may include consolidation of individual services into bundled codes for payment under the physician fee schedule.

2. Progress in Identifying and Reviewing Potentially Misvalued Codes

To fulfill our statutory mandate, we have identified and reviewed numerous potentially misvalued codes as specified in section 1848(c)(2)(K)(ii) of the Act, and we plan to continue our work examining potentially misvalued codes in these areas over the upcoming years. As part of our current process, we identify potentially misvalued codes for review, and request recommendations from the RUC and other public commenters on revised work RVUs and direct PE inputs for those codes. The RUC, through its own processes, also identifies potentially misvalued codes for review. Through our public nomination process for potentially misvalued codes established in the CY 2012 PFS final rule with comment period, other individuals and stakeholder groups submit nominations for review of potentially misvalued codes as well.

Since CY 2009, as a part of the annual potentially misvalued code review and Five-Year Review process, we have reviewed over 1,671 potentially misvalued codes to refine work RVUs and direct PE inputs. We have assigned appropriate work RVUs and direct PE inputs for these services as a result of these reviews. A more detailed discussion of the extensive prior reviews of potentially misvalued codes is included in the CY 2012 PFS final rule with comment period (76 FR 73052 through 73055). In the CY 2012 PFS final rule with comment period, we finalized our policy to consolidate the review of physician work and PE at the same time (76 FR 73055 through 73958), and established a process for the annual public nomination of potentially misvalued services.

In the CY 2013 PFS final rule with comment period, we built upon the work we began in CY 2009 to review potentially misvalued codes that have not been reviewed since the implementation of the PFS (so-called “Harvard-valued codes”). In CY 2009, we requested recommendations from the RUC to aid in our review of Harvard-valued codes that had not yet been reviewed, focusing first on high-volume, low intensity codes (73 FR 38589). In the fourth Five-Year Review (76 FR 32410), we requested recommendations from the RUC to aid in our review of Harvard-valued codes with annual utilization of greater than 30,000. In the CY 2013 PFS final rule with comment period, we identified specific Harvard-valued services with annual allowed charges that total at least $10,000,000 as potentially misvalued. In addition to the Harvard-valued codes, in the CY 2013 PFS final rule with comment period we finalized for review a list of potentially misvalued codes that have stand-alone PE (codes with physician work and no listed work time and codes with no physician work that have listed work time).

In the CY 2016 PFS final rule with comment period, we finalized for review a list of potentially misvalued services, which included eight codes in the neurostimulators analysis-programming family (CPT 95970-95982). We also finalized as potentially misvalued 103 codes identified through our screen of high expenditure services across specialties.

3. Validating RVUs of Potentially Misvalued Codes

Section 1848(c)(2)(L) of the Act requires the Secretary to establish a formal process to validate RVUs under the PFS. The Act specifies that the validation process may include validation of work elements (such as time, mental effort and professional judgment, technical skill and physical effort, and stress due to risk) involved with furnishing a service and may include validation of the pre-, post-, and intra-service components of work. The Secretary is directed, as part of the validation, to validate a sampling of the work RVUs of codes identified through any of the 16 categories of potentially misvalued codes specified in section 1848(c)(2)(K)(ii) of the Act. Furthermore, the Secretary may conduct the validation using methods similar to those used to review potentially misvalued codes, including conducting surveys, other data collection activities, studies, or other analyses as the Secretary determines to be appropriate to facilitate the validation of RVUs of services.

In the CY 2011 PFS proposed rule (75 FR 40068) and CY 2012 PFS proposed rule (76 FR 42790), we solicited public comments on possible approaches, methodologies, and data sources that we should consider for a validation process. A summary of the comments along with our responses are included in the CY 2011 PFS final rule with comment period (75 FR 73217) and the CY 2012 PFS final rule with comment period (73054 through 73055).

We contracted with two outside entities to develop validation models for RVUs.

Given the central role of time in establishing work RVUs and the concerns that have been raised about the current time values used in rate setting, we contracted with the Urban Institute to develop empirical time estimates based on data collected from several health systems with multispecialty group practices. The Urban Institute collected data by directly observing the delivery of services and through the use of electronic health records for services selected by the contractor in consultation with CMS and is using this data to produce objective time estimates. We expect the final Urban Institute report will be made available on the CMS Web site later this year.

The second contract is with the RAND Corporation, which used available data to build a validation model to predict work RVUs and the individual components of work RVUs, time and intensity. The model design was informed by the statistical methodologies and approach used to develop the initial work RVUs and to identify potentially misvalued procedures under current CMS and RUC processes. RAND consulted with a technical expert panel on model design issues and the test results. The RAND report is available under downloads on the Web site for the CY 2015 PFS final rule with comment period at http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/PFS-Federal-Regulation-Notices-Items/CMS-1612-FC.html.

After posting RAND's report on the models and results on our Web site, we received comments indicating that the models did not adequately address global surgery services due to the lack of available data on included visits. Therefore, we modified the RAND contract to include the development of G-codes that could be used to collect data about post-surgical follow-up visits on Medicare claims to meet the requirements in section 1848(c)(8)(B) of the Act regarding collection of data on global services. Our discussion related to this data collection requirement is in section II.D.6. Also, the data from this project would provide information that would allow the time for these services to be included in the model for validating RVUs.

4. CY 2017 Identification and Review of Potentially Misvalued Services

a. 0-Day Global Services That Are Typically Billed With an Evaluation and Management (E/M) Service With Modifier 25

Because routine E/M is included in the valuation of codes with 0-, 10-, and 90-day global periods, Medicare only makes separate payment for E/M services that are provided in excess of those considered included in the global procedure. In such cases, the physician would report the additional E/M service with Modifier 25, which is defined as a significant, separately identifiable E/M service performed by the same physician on the day of a procedure above and beyond other services provided or beyond the usual preservice and postservice care associated with the procedure that was performed. Modifier 25 allows physicians to be paid forE/M services that would otherwise be denied as bundled.

In reviewing misvalued codes, both CMS and the RUC have often considered how frequently particular codes are reported with E/M codes to account for potential overlap in resources. Some stakeholders have expressed concern with this policy especially with regard to the valuation of 0-day global services that are typically billed with a separate E/M service with the use of Modifier 25. For example, when we established our valuation of the osteopathic manipulative treatment (OMT) services, described by CPT codes 98925-98929, we did so with the understanding that these codes are usually reported withE/M codes.

For our CY 2017 proposal (81 FR 46187), we investigated Medicare claims data for CY 2015 and found that 19 percent of the codes that described 0-day global services were billed over 50 percent of the time with an E/M with Modifier 25. Since routine E/M is included in the valuation of 0-day global services, we believed that the routine billing of separate E/M services may have indicated a possible problem with the valuation of the bundle, which is intended to include all the routine care associated with the service.

In the proposed rule (81 FR 46187), we stated that reviewing the procedure codes typically billed with an E/M with Modifier 25 may be one avenue to appropriate valuation for these services. Therefore, we developed and proposed a screen for potentially misvalued codes that identified 0-day global codes billed with an E/M 50 percent of the time or more, on the same day of service, with the same physician and same beneficiary. We included a list of codes with total allowed services greater than 20,000. There are 83 codes that met the proposed criteria for the screen and were proposed as potentially misvalued. We also sought comment regarding additional ways to address appropriate valuations for all services that are typically billed with an E/M with Modifier 25.

The following is the summary of the comments we received.

Comment: Several commenters disagreed with CMS' assertion that there is a possible valuation problem with the bundle when an E/M with Modifier 25 is typically reported on the same day of service as a 0-day global procedure. Commenters stated that billing an E/M with Modifier 25 pays physicians for the justifiable and appropriate services they render to patients; allowing for a patient-centered approach to care. Some commenters considered the possibility that there could be fraudulent billing practices when reporting an E/M with Modifier 25 and a few offered various solutions for rectifying the problem from a program integrity perspective. For example, one commenter suggested that further education on the appropriate use of Modifier 25 or penalty for misuse would be effective alternatives to combat inappropriate billing while another commenter suggested investigating the diagnosis coding for services.

Commenters overwhelmingly opposed any change to billing policies or standard valuation for 0-day services that are billed with an E/M with Modifier 25.

Response: We appreciate commenters' perspective on this issue. While we understand the commenters' views, since routine E/M is included in the valuation of 0-day global services we continue to believe that the routine billing of separate E/M services may still indicate a possible problem with the valuation of the global period or the assignment of the global period for particular codes, given that the period is intended to include all the routine care associated with the service. As discussed below, we are finalizing some of the 0-day global services as potentially misvalued. We will also continue to consider this issue for future rulemaking.

Comment: Several commenters expressed appreciation for the identification of an objective screen and reasonable query. While some commenters were accepting of the screen as proposed, others stated their preference for the screen to be withdrawn entirely or limited in scope, with some commenters suggesting the screen be limited to the codes that met the criteria and for which the overlap had not already been considered by the RUC in developing recommended values. Several thousand commenters suggested withdrawing or limiting the scope of this screen, particularly as it pertains to the OMT codes.

Response: Section 1848(c)(2)(K) of the Act requires the Secretary to periodically identify potentially misvalued services and to review and make appropriate adjustments to the relative values for those services. Section 1848(c)(2)(K) of the Act identifies several categories of services as potentially misvalued, including codes for services where there may be efficiencies when a service is furnished at the same time as other services, along with codes as determined appropriate by the Secretary. Based on the comments received, we understand that stakeholders would have us identify as potentially misvalued only those individual codes with obvious overlapping resource costs when typically reported with an E/M, rather than consider the issue of misvaluation of the global period more broadly. In response to these comments, we are finalizing the use of our screen for 0-day global services that are typically billed with an E/M with Modifier 25 as a mechanism for identifying services that are potentially misvalued.

Because we recognize that the primary purpose in displaying lists of misvalued codes in rulemaking has been to seek recommendations regarding appropriate valuation from stakeholders, including the RUC, for 2017 we are only identifying the services for which we believe there might be the kind of misvaluation the RUC and the medical specialty societies recognize. Based on the comments from these organizations, we believe that for codes reviewed in the past 5 years, the RUC has already addressed that kind of misvaluation. In other words, commenters have made clear that external review of these services is likely to be limited to clear overlap in resource costs, but will not address the broader concerns we have about developing rates for services that include routine E/M when evaluation and management is also routinely separately reported. As a result, we will continue to consider that issue for future rulemaking. We note that we are required under statute to improve the valuation of the 10- and 90-day global periods, and therefore, we will consider this issue in that context, as well.

Comment: While some commenters supported our review of the 83 codes that were proposed as potentially misvalued through the screen, the majority of commenters, including the RUC, stated that the codes detailed in Table 7 did not meet the criteria for the screen because they were either reviewed in the last 5 years and/or are not typically reported with an E/M, and therefore, should be removed. While commenters largely disagreed on the list of proposed codes, most agreed that the services they believed met the screen criteria should be reviewed.

Response: After considering the comments received, we are significantly reducing the number of codes identified as potentially misvalued. We agree with commenters that the majority of the codes that we are not finalizing have been recently reviewed. Due to a drafting error in the proposed rule, we stated that we had exempted codes that had been reviewed in the past 5 years. While that exclusion has been standard for many other misvalued code screens, we did not intend to apply it in this case, given our concerns with the valuation of the global period when E/M visits are routinely reported at the same time. As displayed in the proposed rule, the list of codes reflected our intention to include codes that have been recently reviewed. Regardless, we understand based on comments that any review by stakeholders for recently reviewed codes would be likely to result in similar valuation. Therefore, we do not believe that we should include codes reviewed in the past 5 years on this list of misvalued codes, given the limited nature of the likely review. Regarding the accuracy of which of the codes are typically reported with E/M codes, we note that our review included analysis was based on more recent, full claims data than had yet been made public. In the interest of transparency, we are finalizing the list of services based on the publically available data.

b. End-Stage Renal Disease Home Dialysis Services (CPT Codes 90963 Through 90970)

In the CY 2004 PFS final rule with comment period (68 FR 63216), we established new Level II HCPCS G-codes for end-stage renal disease (ESRD) services and established payment for those codes through monthly capitation payment (MCP) rates. For ESRD center-based patients, payment for the G-codes varied based on the age of the beneficiary and the number of face-to-face visits furnished each month (for example, 1 visit, 2-3 visits and 4 or more visits). We believed that many physicians would provide 4 or more visits to center-based ESRD patients and a small proportion will provide 2-3 visits or only one visit per month. Under the MCP methodology, to receive the highest payment, a physician would have to provide at least four ESRD-related visits per month. However, payment for home dialysis MCP services only varied by the age of beneficiary. Although we did not initially specify a frequency of required visits for home dialysis MCP services, we stated that we expect physicians to provide clinically appropriate care to manage the home dialysis patient.

The CPT Editorial Panel created new CPT codes to replace the G-codes for monthly ESRD-related services, and we accepted the new codes for use under the PFS in CY 2009. The CPT codes created were 90963-90966 for monthly ESRD-related services for home dialysis patient and CPT codes 90967-90970 for dialysis with less than a full month of services.

In a GAO report titled “END-STAGE RENAL DISEASE Medicare Payment Refinements Could Promote Increased Use of Home Dialysis” dated October 2015, http://www.gao.gov/products/GAO-16-125, the GAO stated that experts and stakeholders they interviewed indicated that home dialysis could be clinically appropriate for at least half of patients. Also, at a meeting in 2013, the chief medical officers of 14 dialysis facility chains jointly estimated that a realistic target for home dialysis would be 25 percent of dialysis patients. The GAO noted that CMS data showed that about 10 percent of adult Medicare dialysis patients use home dialysis as of March 2015.

In the report, the GAO noted that CMS intended for the existing payment structure to create an incentive for physicians to prescribe home dialysis, because the monthly payment rate for managing the dialysis care of home patients, which requires a single in-person visit, was approximately equal to the rate for managing and providing two to three visits to ESRD center-based patients. However, GAO found that, in 2013, the rate of $237 for managing home patients was lower than the average payment of $266 and maximum payment of $282 for managing ESRD center-based patients. The GAO stated that this difference in payment rates may discourage physicians from prescribing home dialysis.

Physician associations and other physicians GAO interviewed stated that the visits with home patients are often longer and more comprehensive than in-center visits; this is in part because physicians may conduct visits with individual home patients in a private setting, but they may be able to more easily visit multiple in-center patients on a single day as they receive dialysis. The physician associations GAO interviewed also said that they may spend a similar amount of time outside of visits to manage the care of home patients and that they are required to provide at least one visit per month to perform a complete assessment of the patient.

It is important to note that, as stated in the CY 2011 PFS final rule with comment period (75 FR 73296), we believe that furnishing monthly face-to-face visits is an important component of high quality medical care for ESRD patients being dialyzed at home and generally would be consistent with the current standards of medical practice. However, we also acknowledged that extenuating circumstances may arise that make it difficult for the MCP physician (or NPP) to furnish a visit to a home dialysis patient every month. Therefore, we allow Medicare contractors the discretion to waive the requirement for a monthly face-to-face visit for the home dialysis MCP service on a case-by-case basis, for example, when the MCP physician's (or NPP's) notes indicate that the MCP physician (or NPP) actively and adequately managed the care of the home dialysis patient throughout the month.

The GAO recommended, and we agreed, that CMS examine Medicare policies for monthly payments to physicians to manage the care of dialysis patients and revise them if necessary to ensure that these policies are consistent with our goal of encouraging the use of home dialysis among patients for whom it is appropriate. Therefore, we proposed to identify CPT codes 90963 through 90970 as potentially misvalued codes based on the volume of claims submitted for these services relative to those submitted for facility ESRD services.

The following is summary of the comments we received.

Comment: Commenters supported the proposal to identify these codes as potentially misvalued and supported CMS' goal of encouraging the use of home dialysis among patients for whom it is appropriate. Some commenters suggested we establish parity between payment for four ESRD-related visits per month for in-center dialysis patients and payment for the care of home dialysis patients for an entire month. One commenter cautioned that CMS should also consider factors other than payment that play a critical role in whether a patient decides to use a home dialysis modality as outlined in a recent GAO report and requested that CMS work closely with nephrologists on this issue. One commenter encouraged CMS to focus on incentives for the adult population separately from pediatrics as they see no benefit from reanalysis of the pediatric home and daily dialysis CPT codes 90963-90965 and 90967-90969.

Response: We appreciate all of the comments and agree that CPT codes 90963 through 90970 should be identified as potentially misvalued. After considering the comments, we are finalizing the addition of CPT codes 90963 through 90970 to the list of potentially misvalued codes. We will also continue to consider these issues for future rulemaking.

c. Direct PE Input Discrepancies

i. Appropriate Direct PE Inputs Involved in Procedures Involving Endoscopes

In the proposed rule (81 FR 46190), we stated that stakeholders had raised concerns about potential inconsistencies with the inputs and the prices related to endoscopic procedures in the direct PE database. Upon review, we noted that there are 45 different pieces of endoscope related-equipment and 25 different pieces of endoscope related-supplies that are currently associated with these services. Relative to other kinds of equipment items in the direct PE input, these items are much more varied and used for many fewer services. Given the frequency with which individual codes can be reviewed and the importance of standardizing inputs for purposes of maintaining relativity across PFS services, we believed that this unusual degree of variation was likely to result in code misvaluation. To facilitate efficient review of this particular kind of misvaluation, and because we believed that stakeholders would prefer the opportunity to contribute to such standardization, we requested that stakeholders like the AMA RUC review and make recommendations on the appropriate endoscopic equipment and supplies typically provided in all endoscopic procedures for each anatomical body region, along with their appropriate prices.

The following is summary of the comments we received.

Comment: Many commenters stated that the RUC is the appropriate resource for the review of appropriate direct PE inputs involved in procedures involving endoscopes and urged CMS to work with the RUC to address this issue. Additionally, the RUC stated that due to the complexity of this issue and the need to incorporate input from various specialty societies that the RUC planned to form a workgroup of the PE subcommittee to review the issue.

Response: We appreciate the comments and will review any recommendation provided to us by the RUC for use in future rulemaking, consistent with our normal review processes.

ii. Appropriate Direct PE Inputs in the Facility Post-Service Period When Post-Operative Visits Are Excluded

In the proposed rule (81 FR 46190), we identified a potential inconsistency in instances where there are direct PE inputs included in the facility postservice period even though post-operative visit is not included in a service. We identified 13 codes affected by this issue and stated that we were unclear if the discrepancy was caused by inaccurate direct PE inputs or inaccurate post-operative data in the work time file. We requested that stakeholders including the AMA RUC review these discrepancies and provide their recommendations on the appropriate direct PE inputs for the codes.

The following is summary of the comments we received.

Comment: The RUC stated that for CPT codes 21077 (Impression and preparation of eye socket prosthesis), 21079 (Impression and custom preparation of temporary oral prosthesis), 21080 (Impression and custom preparation of permanent oral prosthesis), 21081 (Impression and custom preparation of lower jaw bone prosthesis), 21082 (Impression and custom preparation of prosthesis for roof of mouth enlargement), 21083 (Impression and custom preparation of roof of mouth prosthesis), and 21084 (Impression and custom preparation of speech aid prosthesis) the practice expense time in the postservice period in the facility setting is completely distinct from the physician post-operative visit and that time must be accounted for the manufacture and fitting of the prosthetics. The RUC stated that the following codes all had inaccurate post-operative data in the work time file and provided recommendations on appropriate post-operative visits: CPT codes 28636 (Insertion of hardware to foot bone dislocation with manipulation, accessed through the skin), 28666 (Insertion of hardware to toe joint dislocation with manipulation, accessed through the skin), 43652 (Incision of vagus nerves of stomach using an endoscope), 47570 (Connection of gall bladder to bowel using an endoscope), and 66986 (Exchange of lens prosthesis). Additionally, another commenter stated that CPT code 46900 (Chemical destruction of anal growths) also had inaccurate post-operative data in the work time file and provided a recommendation on the appropriate post-operative visit.

Response: We thank stakeholders for their comments. We will review the recommendations provided to us by the AMA RUC and other commenters and will consider for future rulemaking, consistent with our normal review processes.

d. Insertion and Removal of Drug Delivery Implants—CPT Codes 11981 and 11983

In the proposed rule (81 FR 46190), we stated that stakeholders had urged CMS to create new coding describing the insertion and removal of drug delivery implants for buprenorphine hydrochloride, formulated as a 4 rod, 80 mg, long acting subdermal drug implant for the treatment of opioid addiction. The stakeholders suggested that current coding describing insertion and removal of drug delivery implants was too broad and that new coding was needed to account for specific additional resource costs associated with particular treatment. We identified existing CPT codes 11981 (Insertion, non-biodegradable drug delivery implant), 11982 (Removal, non-biodegradable drug delivery implant), and 11983 (Removal with reinsertion, non-biodegradable drug delivery implant) as potentially misvalued codes and sought comment and information regarding whether the current resource inputs in work and practice expense for the codes appropriately accounted for variations in the service relative to which devices and related drugs are inserted and removed.

The following is summary of the comments we received.

Comment: One commenter stated that CMS should create distinct codes and payment levels for a four-rod implant as opposed to the one-rod implant detailed in CPT codes 11981-11983. In contrast, another commenter stated that the identified codes adequately describe the work and practice expense for drug implant delivery and removal services. Additionally, another commenter stated the codes should be removed from the potentially misvalued list. The RUC stated that a coding change proposal had been submitted for the services under the CPT process and that the RUC anticipated providing relevant recommendations for CY 2018.

Response: We thank stakeholders for their comments. We will review new coding and recommended valuations for future rulemaking, consistent with our normal review processes.

5. Valuing Services That Include Moderate Sedation as an Inherent Part of Furnishing the Procedure

The CPT manual identifies more than 400 diagnostic and therapeutic procedures (listed in Appendix G) for which the CPT Editorial Panel has determined that moderate sedation is an inherent part of furnishing the procedure. In developing RVUs for these services, we include the relative resources associated with moderate sedation in the valuation since the CPT codes include moderate sedation as an inherent part of the procedure. Therefore, practitioners only report the procedure code when furnishing the service. Endoscopic procedures constitute a significant portion of the services identified in Appendix G. In the CY 2015 PFS proposed rule (79 FR 40349), we noted that it appeared that practice patterns for endoscopic procedures were changing, with anesthesia increasingly being separately reported for these procedures, meaning that the relative resources associated with sedation were no longer incurred by the practitioner reporting the Appendix G procedure. We indicated that, in order to reflect apparent changes in medical practice, we were considering establishing a uniform approach to the appropriate valuation of all Appendix G services for which moderate sedation is no longer inherent, rather than addressing the issue at the procedure level as individual codes are revalued. We solicited public comment on approaches to the appropriate valuation of these services.

In the CY 2016 PFS proposed rule (80 FR 41707), we again solicited public comment and recommendations on approaches to address the appropriate valuation of moderate sedation related to Appendix G services. Following our comment solicitation, the CPT Editorial Panel created CPT codes for separately reporting moderate sedation services in association with the elimination of Appendix G from the CPT manual for CY 2017. This coding change would provide for payment for moderate sedation services only in cases where they are furnished. In addition to providing recommended values for the new codes used to separately report moderate sedation, the RUC provided a methodology for revaluing all services previously identified in Appendix G, without moderate sedation, in order to make appropriate corresponding adjustments for the procedural services. The RUC recommended this methodology to address moderate sedation valuation generally instead of recommending that it be addressed as individual codes are reviewed. The RUC's recommended methodology would remove work RVUs for moderate sedation from Appendix G codes based on a code-level assessment of whether the procedures are typically furnished to straightforward patients or more difficult patients. Based on its recommended methodology, the RUC recommended removal of fewer RVUs from each of the procedural services than it recommended for valuing the moderate sedation services. If we were to use the RUC-recommended values for both the moderate sedation codes and the Appendix G procedural codes without refinement, overall payments for these procedures, when moderate sedation is furnished, would increase relative to the current payment.

We direct readers to section II.L. of this final rule, which includes more detail regarding our valuation of the new moderate sedation codes, our methodology for revaluation of the procedural codes previously identified in Appendix G, and discussion and responses to the public comments we received regarding our proposal. We believe that the RVUs assigned under the PFS should reflect the overall relative resources of PFS services, regardless of how many codes are used to report the services. Therefore, our methodology for valuation of Appendix G procedural services maintains current resource assumptions for the procedures when furnished with moderate sedation and redistributes the RVUs associated with moderate sedation (previously included in the Appendix G procedural codes) to other PFS services. We believe that this methodology for revaluation of Appendix G services without moderate sedation is consistent with our general principle that the overall relative resources for the procedures do not change based solely on changes in coding.

We also noted in the CY 2017 PFS proposed rule that stakeholders presented information to CMS regarding specialty group survey data for physician work. The stakeholders shared survey results for physician work involved in furnishing moderate sedation that demonstrated a significant bimodal distribution between procedural services furnished by gastroenterologists (GI) and procedural services furnished by other specialties. Since we believe that gastroenterologists furnish the highest volume of services previously identified in Appendix G, and services primarily furnished by gastroenterologists prompted the concerns that led to our identification of changes in medical practice and potentially duplicative payment for these codes, we have addressed the variations between GI and other specialties in our review of the new moderate sedation CPT codes and their recommended values. We again direct readers to section II.L. of this final rule where we discuss our establishment of an endoscopy-specific moderate sedation G-code that augments the new CPT codes for moderate sedation, the public comments we received, and our finalized valuations reflecting the differences in the physician survey data between GI and other specialties.

6. Collecting Data on Resources Used in Furnishing Global Services

a. Background

(1) Current Payment Policy for Global Packages

Under the PFS, certain services, such as surgery, are valued and paid for as part of global packages that include the procedure and the services typically furnished in the periods immediately before and after the procedure. For each of these global packages, we establish a single PFS payment that includes payment for particular services that we assume to be typically furnished during the established global period. There are three primary categories of global packages that are labeled based on the number of post-operative days included in the global period: 0-day; 10-day; and 90-day. The 0-day global packages include the surgical procedure and the pre-operative and post-operative services furnished by the physician on the day of the service. The 10-day global packages include these services and, in addition, visits related to the procedure during the 10 days following the day of the procedure. The 90-day global packages include the same services as the 0-day global codes plus the pre-operative services furnished one day prior to the procedure and post-operative services during the 90 days immediately following the day of the procedure. Section 40.1 of Chapter 12 of the Claims Processing Manual (Pub. 100-04) defines the global surgical package to include the following services related to the surgery when furnished during the global period by the same physician or another practitioner in the same group practice:

• Pre-operative Visits: Pre-operative visits after the decision is made to operate beginning with the day before the day of surgery for major procedures and the day of surgery for minor procedures;
• Intra-operative Services: Intra-operative services that are normally a usual and necessary part of a surgical procedure;
• Complications Following Surgery: All additional medical or surgical services required of the surgeon during the post-operative period of the surgery because of complications that do not require additional trips to the operating room;
• Post-operative Visits: Follow-up visits during the post-operative period of the surgery that are related to recovery from the surgery;
• Post-surgical Pain Management: By the surgeon; and
• Miscellaneous Services: Items such as dressing changes; local incisional care; removal of operative pack; removal of cutaneous sutures and staples, lines, wires, tubes, drains, casts, and splints; insertion, irrigation and removal of urinary catheters, routine peripheral intravenous lines, nasogastric and rectal tubes; and changes and removal of tracheostomy tubes.

In the CY 2015 PFS proposed and final rules we extensively discussed the problems with accurate valuation of 10- and 90-day global packages. Our concerns included the fact that we do not use actual data on services furnished to update the rates, questions regarding the accuracy of our current assumptions about typical services, whether we will be able to adjust values on a regular basis to reflect changes in the practice of medicine and health care delivery, and how our global payment policies affect what services are actually furnished (79 FR 67582 through 67585). In finalizing a policy to transform all 10- and 90-day global codes to 0-day global codes in CY 2017 and CY 2018, respectively, to improve the accuracy of valuation and payment for the various components of global packages, including pre- and post-operative visits and the procedure itself, we stated that we were adopting this policy because it is critical that PFS payment rates be based upon RVUs that reflect the relative resources involved in furnishing the services. We also stated our belief that transforming all 10- and 90-day global codes to 0-day global packages would:

• Increase the accuracy of PFS payment by setting payment rates for individual services that more closely reflect the typical resources used in furnishing the procedures;
• Avoid potentially duplicative or unwarranted payments when a beneficiary receives post-operative care from a different practitioner during the global period;
• Eliminate disparities between the payment for E/M services in global periods and those furnished individually;
• Maintain the same-day packaging of pre- and post-operative physicians' services in the 0-day global packages; and
• Facilitate the availability of more accurate data for new payment models and quality research.

(2) Data Collection & Revaluation of Global Packages Required by MACRA

Section 523(a) of the Medicare Access and CHIP Reauthorization Act of 2015 (MACRA) (Pub. L. 114-10, enacted April 16, 2015) added section 1848(c)(8)(A) of the Act, which prohibits the Secretary from implementing the policy, described above, that would have transformed all 10-day and 90-day global surgery packages to 0-day global packages.

Section 1848(c)(8)(B) of the Act, which was also added by section 523(a) of the MACRA, requires us to collect data to value surgical services. Section 1848(c)(8)(B)(i) of the Act requires us to develop, through rulemaking, a process to gather information needed to value surgical services from a representative sample of physicians, and requires that the data collection begin no later than January 1, 2017. The collected information must include the number and level of medical visits furnished during the global period and other items and services related to the surgery and furnished during the global period, as appropriate. This information must be reported on claims at the end of the global period or in another manner specified by the Secretary. Section 1848(c)(8)(B)(ii) of the Act requires that, every 4 years, we reassess the value of this collected information; and allows us to discontinue the collection of this information if the Secretary determines that we have adequate information from other sources to accurately value global surgical services. Section 1848(c)(8)(B)(iii) of the Act specifies that the Inspector General shall audit a sample of the collected information to verify its accuracy. Section 1848(c)(9) of the Act (added by section 523(b) of the MACRA) authorizes the Secretary, through rulemaking, to delay up to 5 percent of the PFS payment for services for which a physician is required to report information under section 1848(c)(8)(B)(i) of the Act until the required information is reported.

Section 1848(c)(8)(C) of the Act, which was also added by section 523(a) of the MACRA, requires that, beginning in CY 2019, we must use the information collected as appropriate, along with other available data, to improve the accuracy of valuation of surgical services under the PFS.

(3) Public Input

As noted above, section 1848(c)(8)(C) of the Act mandates that we use the collected data to improve the accuracy of valuation of surgery services beginning in 2019. We described in the CY 2015 PFS final rule (79 FR 67582 through 67591) the limitations and difficulties involved in the appropriate valuation of the global packages, especially when the resources and the related values assigned to the component services are not defined. To gain input from stakeholders on implementation of this data collection, we sought comment on various aspects of this task in the CY 2016 proposed rule (80 FR 41707 through 41708). We solicited comments from the public regarding the kinds of auditable, objective data (including the number and type of visits and other services furnished during the post-operative period by the practitioner furnishing the procedure) needed to increase the accuracy of the valuation and payment for 10- and 90-day global packages. We also solicited comment on the most efficient means of acquiring these data as accurately and efficiently as possible. For example, we sought information on the extent to which individual practitioners or practices may currently maintain their own data on services furnished during the post-operative period, and how we might collect and objectively analyze those data and use the results for increasing the accuracy of the values beginning in CY 2019.

We received many comments in response to the comment solicitation in the CY 2016 proposed rule regarding potential methods of valuing the individual components of the global surgical package. A large number of comments expressed strong support for our proposal to hold an open door forum or town hall meetings with the public. In response, we held a national listening session on January 20, 2016. Prior to the listening session, the topics for which guidance was being sought were sent electronically to those who registered for the session and made available on our Web site. The topics were:

• Capturing the types of services typically furnished during the global period.
• Determining the representative sample for the claims-based data collection.
• Determining whether we should collect data on all surgical services or, if not, which services should be sampled.
• Potential for designing data collection elements to interface with existing infrastructure used to track follow-up visits within the global period.
• Consideration of using the 5 percent withhold until required information is furnished to encourage reporting.

The 658 participants in the national listening session provided valuable information on this task. A written transcript and an audio recording of this session are available at https://www.cms.gov/Outreach-and-Education/Outreach/NPC/National-Provider-Calls-and-Events-Items/2016-01-20-MACRA.html.

b. Data Collection Required To Accurately Value Global Packages

Resource-based valuation of individual physicians' services is a critical foundation for Medicare payment to physicians. It is essential that the RVUs under the PFS be based as closely and accurately as possible on the actual resources used in furnishing specific services to make appropriate payment and preserve relativity among services. For global surgical packages, this requires using objective data on all of the resources used to furnish the services that are included in the package. Not having such data for some components may significantly skew relativity and create unwarranted payment disparities within the PFS.

The current valuations for many services valued as global packages are based upon the total package as a unit rather than by determining the resources used in furnishing the procedure and each additional service/visit and summing the results. As a result, we do not have the same level of information about the components of global packages as we do for other services. To value global packages accurately and relative to other procedures, we need accurate information about the resources—work, PEs and malpractice—used in furnishing the procedure, similar to what is used to determine RVUs for all services. In addition we need the same information on the post-operative services furnished in the global period (and pre-operative services the day before for 90-day global packages). Public comments about our CY 2015 proposal to value all global services as 0-day global services and pay separately for additional post-operative services when furnished indicated that there were no reliable data available on the value of the underlying procedure that did not also incorporate the value of the post-operative services, reinforcing our view that more data are needed across the board.

While we believe that most of the services furnished in the global period are visits for follow-up care, we do not have accurate information on the number and level of visits typically furnished because those billing for global services are not required to submit claims for post-operative visits. A May 2012 Office of Inspector General (OIG) report, titled Cardiovascular Global Surgery Fees Often Did Not Reflect the Number of Evaluation and Management Services Provided (http://oig.hhs.gov/oas/reports/region5/50900054.pdf) found that for 202 of the 300 sampled cardiovascular global surgeries, the Medicare payment rates were based on a number of visits that did not reflect the actual number of services provided. Specifically, physicians provided fewer services than the visits included in the payment calculation for 132 global surgery services and provided more services than were included in the payment calculations for 70 services. Similar results were found in OIG reports titled “Musculoskeletal Global Surgery Fees Often Did Not Reflect The Number Of Evaluation And Management Services Provided” (http://oig.hhs.gov/oas/reports/region5/50900053.asp) and “Review of Cataract Global Surgeries and Related Evaluation and Management Services, Wisconsin Physicians Service Insurance Corporation Calendar Year 2003, March 2007.” (http://oig.hhs.gov/oas/reports/region5/50600040.pdf).

Claims data plays a major role in PFS ratesetting. Specifically, Medicare claims data are a primary driver in the allocation of indirect PE RVUs and MP RVUs across the codes used by particular specialties, and in making overall budget neutrality and relativity adjustments. In most cases, a claim must be filed for all visits. Such claims provide information such as the place of service, the type and, if relevant, the level of the service, the date of the service, and the specialty of the practitioner furnishing the services. Because we have not required claims reporting of visits included in global surgical packages, we do not have any of this information for the services bundled in the package.

In addition to the lack of information about the number and level of visits actually furnished, the current global valuations rely on crosswalks to E/M visits, based upon the assumption that the resources, including work, used in furnishing pre- and post-operative visits are similar to those used in furnishing E/M visits. We are unaware of any studies or surveys that verify this assertion. Although we generally value the visits included in global packages using the same direct PE inputs as are used for E/M visits, for services for which the RUC recommendations include specific PE inputs in addition to those typically included for E/M visits, we generally use the additional inputs in the global package valuation. In contrast, when a visit included in a global package would use fewer resources than a comparable E/M service, the RUC generally does not include recommendations to decrease the PE inputs of the visit included in the global package, and we have not generally made comparable reductions. Another inconsistency with our current global package valuation approach is that even though we effectively assume that the E/M codes are appropriate for valuing pre- and post-operative services, the indirect PE inputs used for calculating payments for global services are based upon the specialty mix furnishing the global service, not the specialty mix of the physicians furnishing the E/M services, resulting in a different valuation for the E/M services contained in global packages than for separately billable E/M services. There is a critical need to obtain complete information if we are to value global packages accurately and in a way that preserves relativity across the fee schedule.

In response to the requirement of section 1848(c)(8)(B)(i) of the Act that we develop, through rulemaking, a process to gather information needed to value surgical services, we proposed a rigorous data collection effort to provide us the data needed to accurately value the 4,200 codes with a 10- or 90-day global period. Using our authority under sections 1848(c)(2)(M) and (c)(8)(B)(i) of the Act, we proposed to gather the data needed to determine how to best structure global packages with post-operative care that is typically delivered days, weeks or months after the procedure and whether there are some procedures for which accurate valuation for packaged post-operative care is not possible. Finally, we indicated that these data would provide useful information to assess the resources used in furnishing pre- and post-operative care in global periods. To accurately do so, we need to know the volume and costs of the resources typically used.

We proposed a three-pronged approach to collect timely and accurate data on the frequency of and the level of pre- and post-operative visits and the resources involved in furnishing the pre-operative visits, post-operative visits, and other services for which payment is included in the global surgical payment. By analyzing these data, we would not only have the most comprehensive information available on the resources used in furnishing these services, but also would be able to determine the appropriate packages for such services. Specifically, the proposal included:

• A requirement for claims-based reporting about the number and level of pre- and post-operative visits furnished for 10- and 90-day global services.
• A survey of a representative sample of practitioners about the activities involved in and the resources used in providing a number of pre- and post-operative visits during a specified, recent period of time, such as two weeks.
• A more in-depth study, including direct observation of the pre- and post-operative care delivered in a small number of sites, and a separate survey module for practitioners practicing in ACOs.

The information collected and analyzed through the activities would be the first comprehensive look at the volume and level of services in a global period, and the activities and inputs involved in furnishing global services. The data from these activities would ultimately inform our revaluation of global surgical packages as required by statute.

To expand awareness of the proposal for data collection, we held a national listening session in which CMS reviewed the proposal for participants. Subsequent to this national listening session, we held a town hall meeting at the CMS headquarters in which participants, in person and virtual, shared their views on the proposal with CMS. The transcript from these town halls is available on the CMS Web site with the CY 2017 final rule downloads.

(1) Statutory Authority for Data Collection

As described in this section of the final rule, section 1848(c)(8)(B)(i) of the Act requires us to develop, through rulemaking, a process to gather information needed to value surgical services from a representative sample of physicians. The statute requires that the collected information include the number and level of medical visits furnished during the global period and other items and services related to the surgery and furnished during the global period, as appropriate.

In addition, section 1848(c)(2)(M) of the Act, which was added to the Act by section 220 of the PAMA, authorizes the Secretary to collect or obtain information on resources directly or indirectly related to furnishing services for which payment is made under the PFS. Such information may be collected or obtained from any eligible professional or any other source. Information may be collected or obtained from surveys of physicians, other suppliers, providers of services, manufacturers, and vendors. That section also authorizes the Secretary to collect information through any other mechanism determined appropriate. When using information gathered under this authority, the statute requires the Secretary to disclose the information source and discuss the use of such information in the determination of relative values through notice and comment rulemaking.

As described in this section of the final rule, to gain information to assist CMS in determining the appropriate packages for global services and to revalue those services, CMS needs more information on the resources used in furnishing such services. Through the claims-based data collection and the study we are finalizing in this final rule, we would have better information about the actual number of services furnished to Medicare beneficiaries to use in valuation for these codes than has been typically available, such as from RUC surveys that reflect practitioner's estimates of the number of services typically furnished. We anticipate that such efforts would inform how to more regularly collect data on the resources used in furnishing physicians' services. To the extent that such mechanisms prove valuable, they may be used to collect data for valuing other services. To achieve this significant data collection, we proposed to collect data under the authority of both section 1848(c)(8)(B) and (c)(2)(M) of the Act.

(2) Claims-Based Data Collection

We proposed a claims-based data collection that would have required all those providing 10- or 90-day global services to report on services furnished during the global period using a series of G-codes specially created for this purpose, beginning January 1, 2017.

In response to the comments submitted on the proposal, we are finalizing a claims-based data collection that differs from this proposal in the following significant ways:

• CPT code 99024 will be used for reporting post-operative services rather than the proposed set of G-codes. Reporting will not be required for pre-operative visits included in the global package or for services not related to patient visit.
• Reporting will be required only for services related to codes reported annually by more than 100 practitioners and that are reported more than 10,000 times or have allowed charges in excess of $10 million annually.
• Practitioners are encouraged to begin reporting post-operative visits for procedures furnished on or after January 1, 2017, but the mandatory requirement to report will be effective for services related to global procedures furnished on or after July 1, 2017.
• Only practitioners who practice in groups with 10 or more practitioners in Florida, Kentucky, Louisiana, Nevada, New Jersey, North Dakota, Ohio, Oregon, and Rhode Island will be required to report. Practitioners who only practice in smaller practices or in other geographic areas are encouraged to report data, if feasible.

Given that the data collection will be limited to only some states, a subset of global services, and only to those who practice in larger practices the information collected through claims for global packages services will not parallel the claims data that are available in pricing other PFS services. However, we believe that the information collected through this data collection will be a significant improvement over the information currently available to value these services and will be supplemented with information obtained through other mechanisms.

In the following sections, we discuss the comments on each element of our data collection proposal, our responses and our final decision.

(a) Information To Be Reported

A key element of claims-based reporting is using codes that appropriately reflect the services furnished. In response to the comment solicitation in the CY 2016 PFS proposed rule and the input received via the January 2016 listening session, we received numerous recommendations for the information to be reported on claims. The most frequently recommended approach was for practitioners to report the existing CPT code for follow-up visits included in the surgical package (CPT 99024—Postoperative follow-up visit, normally included in the surgical package, to indicate that an E/M service was performed during a postoperative period for a reason(s) related to the original procedure). Others suggested using this code for outpatient visits and using length of stay data to estimate the number of inpatient visits during the global period. In response to our concerns that CPT code 99024 would provide only the number of visits and not the level of visits as required by the statute, one commenter suggested using modifiers in conjunction with CPT code 99024 to indicate the level of the visit furnished. Others recommended using existing CPT codes for E/M visits to report post-operative care. One commenter suggested that CMS analyze data from a sample of large systems and practices that are using electronic health records that require entry of some CPT code for every visit to capture the number of post-operative visits. After noting that the documentation requirements and PEs required for post-operative visits differ from those of E/M visits outside the global period, one commenter encouraged us to develop a separate series of codes to capture the work of the post-operative services and to measure, not just estimate, the number and complexity of visits during the global period.

Other commenters opposed the use of a new set of codes or the use of modifiers to report post-operative visits. Commenters also noted several issues for us to consider in developing data collection mechanisms, including that many post-operative services do not have CPT codes to bill separately, that surgeons perform a wide range of collaborative care services, and that patient factors, including disease severity and comorbidities, influence what post-operative care is furnished.

To assist us in determining appropriate coding for claims-based reporting, we added a task to the RAND validation contract for developing a model to validate the RVUs in the PFS, which was awarded in response to a requirement in the Affordable Care Act. Comments that we received on the validation report suggested the models did not adequately address global surgery services due to the lack of available data on visits included in the global package. Therefore, we modified the validation contract to include the development of G-codes that could be used to collect data about post-surgical follow-up visits on Medicare claims for valuing global services under MACRA so that this time could be included in the model for validating RVUs.

To inform its work on developing coding for claims-based reporting, the contractor conducted interviews with surgeons and other physicians/non-physician practitioners (NPP) who provide post-operative care. A technical expert panel (TEP), convened by the contractor, reviewed the findings of the interviews and provided input on how to best capture care provided in the post-operative period on claims.

In summarizing the input from the interviews and the TEP, the contractor indicated that several considerations were important in developing a claims-based method for capturing post-operative services. First, a simple system to facilitate reporting was needed. Since it was reported that a majority of post-operative visits are straightforward, the contractor found that a key for any proposed system is identifying the smaller number of complex post-operative visits. Another consideration was not using the existing CPT E/M structure to capture postoperative care because of concerns that E/M codes are inadequately designed to capture the full scope of post-operative care and that using such codes might create confusion. Another consideration was that the TEP was most enthusiastic about a set of codes that used site of care, time, and complexity to report visits. The contractor also believed it was important to distinguish—particularly in the inpatient setting—between circumstances where a surgeon is providing primary versus secondary management of a patient. Finally, a mechanism for reporting the postoperative care occurs outside of in-person visits and by clinical staff was needed. The report noted that in the inpatient setting in particular, surgeons spend considerable time reviewing test results and coordinating care with other practitioners.

After reviewing various approaches, a set of time-based, post-operative visit codes that could be used for reporting care provided during the post-operative period was recommended.

The recommended codes distinguish services by the setting of care and whether they are furnished by a physician/NPP or by clinical staff. All codes are intended to be reported in 10-minute increments. A copy of the report is available on the CMS Web site under downloads for the CY 2017 PFS proposed rule with comment period at http://www.cms.gov/physicianfeesched/downloads/.

We proposed the following no-pay codes be used for reporting on claims the services actually furnished but not paid separately because they are part of global packages.

(i) Coding for Inpatient Global Service Visits

Our proposal included three codes for reporting inpatient pre- and post-operative visits that distinguish the intensity involved in furnishing the services. Under this proposal, visits that involve any combination or number of the services listed in Table 10, which were recommended by the contractor as those in a typical visit, would be reported using GXXX1. Based on the findings from the interviews and the TEP, the report indicated that the vast majority of inpatient post-operative visits would be expected to be reported using GXXX1.

Under our proposal, inpatient pre- and post-operative visits that are more complex than typical visits but do not qualify as critical illness visits would be coded using GXXX2 (Inpatient visit, complex, per 10 minutes, included in surgical package). To report this code, the practitioner would be required to furnish services beyond those included in a typical visit and have documentation that indicates what services were provided that exceeded those included in a typical visit. In the proposed rule, we noted some circumstances that might merit the use of the complex visit code are secondary management of a critically ill patient where another provider such as an intensivist is providing the primary management, primary management of a particularly complex patient such as a patient with numerous comorbidities or high likelihood of significant decline or death, management of a significant complication, or complex procedures outside of the operating room (For example, significant debridement at the bedside).

The highest level of inpatient pre- and post-operative visits, critical illness visits (GXXX3—Inpatient visit, critical illness, per 10 minutes, included in surgical package) would be reported when the physician is providing primary management of the patient at a level of care that would be reported using critical care codes if it occurred outside of the global period. This involves acute impairment of one or more vital organ systems such that there is a high probability of imminent or life threatening deterioration in the patient's condition.

Similar to how time is now counted for the existing CPT critical care codes, we proposed that all time spent engaged in work directly related to the individual patient's care would count toward the time reported with the inpatient visit codes; this includes time spent at the immediate bedside or elsewhere on the floor or unit, such as time spent with the patient and family members, reviewing test results or imaging studies, discussing care with other staff, and documenting care.

(ii) Coding for Office and Other Outpatient Global Services Visits

For the three codes in our proposal that would be used for reporting post-operative visits in the office or other outpatient settings, codes, time would be defined as the face-to-face time with patient, which reflects the current rules for time-based outpatient codes.

Like GXXX1, GXXX5 (Office or other outpatient visit, typical, per 10 minutes, included in surgical package) would be used for reporting any combination of activities in Table 10 under our proposal.

We proposed only face-to-face time spent by the practitioner with the patient and their family members would count toward the time reported with the office visit codes.

(iii) Coding for Services Furnished via Electronic Means

Services that are furnished via phone, the internet, or other electronic means outside the context of a face-to-face visit would be reported using GXXX7 when furnished by a practitioner and GXXX8 when provided by clinical staff under our proposal. We proposed that practitioners would not report these services if they are furnished the day before, the day of, or the day after a visit as we believe these would be included in the pre- and post-service activities in the typical visit. However, we proposed that these codes be used to report non-face-to-face services provided by clinical staff prior to the primary procedure since global surgery codes are typically valued with assumptions regarding pre-service clinical labor time. Given that some practitioners have indicated that services they furnish commonly include activities outside the face-to-face service, we believed it was important to capture information about those activities in both the pre- and post-service periods. We also believed these requirements to report on clinical labor time are consistent with and no more burdensome than those used to report clinical labor time associated with chronic care management services, which similarly describe care that takes place over more than one patient encounter.

In addition, we proposed for services furnished via interactive telecommunications that meet the requirements of a Medicare telehealth service visit, the appropriate global service G-code for the services would be reported with the GT modifier to indicate that the service was furnished “via interactive audio and video telecommunications systems.”

(iv) Rationale for Use of G-Codes

After considering the contractor report, the comments in response to the comment solicitation in the CY 2016 proposed rule and other stakeholder input that we have received, and our needs for data to fulfill our statutory mandate and to value surgical services appropriately, we proposed this new set of codes because we believe it provides us the most robust data upon which to determine the most appropriate way and amounts to pay for PFS surgical services. We noted that these proposed codes would provide data of the kind that can reasonably collected through claims data and that reflect what we believe are key issues in the valuation of post-operative care—where the service is provided, who furnishes the service, its relative complexity, and the time involved in the service.

We solicited public comments about all aspects of these codes, including the nature of the services described, the time increment, and any other areas of interest to stakeholders. We noted particular interest in any pre- or post-operative services furnished that could not be appropriately captured by these codes. We solicited comments on whether the proposed codes were appropriate for collecting data on pre-operative services. We also sought comment on any activities that should be added to the list of activities in Table 10 to reflect typical pre-operative visit activities.

(v) Alternative Approach to Coding

In making the proposal for G-codes, we noted that many stakeholders had expressed strong support for the use of CPT code 99024 (Postoperative follow-up visit, normally included in the surgical package, to indicate that an evaluation and management service was performed during a post-operative period for a reason(s) related to the original procedure) to collect data on post-operative care. In response to stakeholders noting that practitioners are familiar with this existing CPT code and the burden on practitioners would be minimized by only having to report that a visit occurred, not the level of the visit, we noted that we did not believe that this code alone would provide the information that we need for valuing surgical services nor do we believe it alone can meet the statutory requirement that we collect data on the number and level of visits. Given the strong support for the use of CPT code 99024, we solicited comments specifically on how we could use this code to capture the statutorily required data on the number and level of visits and the data that we would need to value global services in the future.

We also discussed in the proposed rule our concern that using CPT code 99024 with modifiers to indicate to which of the existing levels of E/M codes the visit corresponds may not accurately capture what drives greater complexity in post-operative visits. We noted that as outlined in the contractor's report, E/M billing requirements are built upon complexity in elements such as medical history, review of systems, family history, social history, and how many organ systems are examined. In the context of a post-operative visit, many of these elements may be irrelevant. The contractor's report also notes that there was significant concern from interviewees and the expert panel about documentation that is required for reporting E/M codes. Specifically, they stated that documentation requirements for surgeons to support the relevant E/M visit code would place undue administrative burden on surgeons given that many surgeons currently use minimal documentation when they provide a postoperative visit. We also noted that to value surgical packages accurately we need to understand the activities involved in furnishing post-operative care and as discussed above, we lack information that would demonstrate that activities involved in post-operative care are similar to those in E/M services. In addition, the use of modifiers to report levels of services is more difficult to operationalize than using unique HCPCS codes. However, we sought comments on whether, and if so, why, practitioners would find it easier to report CPT code 99024 with modifiers corresponding to the proposed G-code levels rather than the new G-codes, as proposed. We also sought comment on whether practitioners would find it difficult to use this for pre-operative visits since the CPT code descriptor specifically defines it as a “post-operative follow-up” service.

We also sought comment on whether time of visits could alone be a proxy for the level of visit. If pre- and post-operative care varies only by the time the practitioner spends on care so that time could be a proxy for complexity of the service, then we could use the reporting of CPT code 99024 in 10-minute increments to meet the statutory requirement of collecting claims-based data on the number and level of visits. In addition to comments on whether time is an accurate proxy for level of visit, we solicited comment on the feasibility and desirability of reporting CPT code 99024 in 10-minute increments.

The following is a summary of the comments that we received on our proposal to use G-codes for reporting the services furnished during the pre- and post-operative periods of 10- and 90-day global services.

Comment: Many commenters offered critiques of the G-codes. Most objected to reporting using the proposed G-codes. Some commenters raised concerns with the code definitions. These included: Lack of alignment with clinical workflow, failure to adequately account for variation in complexity and medical decision-making, and use of the term “typical” to define visits in a different way than the term is generally used in PFS valuations. One commenter suggested that CMS should require care plans for outpatient visits in the post-operative period. It was also suggested that the complex visit code could be improved by using a term other than “complex” in the definition. A commenter questioned whether that vast majority of cases would be complex instead of “typical,” since the definition of “complex” included management of a patient with multiple comorbidities and most Medicare beneficiaries have multiple comorbidities. A commenter also suggested that CMS refine the G-codes to distinguish physician visits from NPP visits. In addition, several commenters objected to the proposed G-codes for on-line and telephone services because they believed it would be nearly impossible to track these data and extremely burdensome to do so. Commenters indicated that the G-codes were not well-defined overall and should not be used without testing to determine their validity.

Response: We appreciate the detailed comments on the design of the G-codes and the concerns regarding their limitations in appropriately reflecting the services furnished in 10- and 90-day global periods. These comments provide information for how the G-codes could be modified to better reflect services furnished in global periods, however, as is discussed at the outset of this section, we are not using the proposed G-codes for this data collection effort.

Comment: Most commenters objected to using codes based on time increments and the proposed 10-minute increments, specifically. Some stated that reporting of services by time did not reflect the way surgeons practiced and would divert practitioners from patient care. One commenter stated that it was not feasible for practitioners to collect time data for every task that they or their clinical staff performed. Another stated that requesting physicians and/or their staff to use a stop-watch to, in effect, conduct time and motion studies for all their non-operating room patient care activities is an incredible burden. Another stated that reporting time in 10-minute increments “is untenable,” noting that, except for a few specialties, physicians do not think of providing care in terms of timed increments. The commenter added that surgeons, in particular, are not accustomed to reporting time for all pre- and post-operative visits and to do so would be a huge disruption to workflow. In addition to objections about the burden of reporting time data, some commenters objected to the use of time data as a factor in valuations.

Three organizations commented that it was appropriate to collect time data, but recommended that we do so based upon 15-minute increments as these were more familiar to physicians than the proposed 10-minue increments. In addition, some other groups, including MedPAC, agreed that data on time was needed for valuations.

Response: Time is a key factor in valuing physician services under the physician fee schedule. Section 1848(c)(1) of the Act defines the work component as the portion of the resources used in furnishing the service that reflects physician time and intensity in furnishing the service. We also note that time-based codes are used by practitioners for a range of services in the PFS including psychotherapy, anesthesiology and critical care services. Critical care services are notable because these services are likely to be furnished intermittently as many commenters suggested is typical for post-operative follow-up services. Since issues have not been raised about the difficulty of using the current critical care codes, it is unclear why reporting of time would be burdensome and disruptive of care in this area. We have no reason to believe the documentation of time is more difficult or burdensome for those furnishing 10- and 90-day global services than for other practitioners. However, based on the comments, it is clear that many practitioners would perceive reporting of time involved in furnishing these services to be a significant increase in burden relative to existing practice. Before implementing a change considered by so many to be so burdensome, we are exploring other ways of obtaining information that can be used to improve the accuracy of valuing these services. Accordingly, we are not finalizing, at this time, the requirement to use time-based codes.

Comment: Many commenters stated that the use of these codes would be costly, requiring extensive education of practitioners and staff and necessitating updates to EHR systems and billing software. Some also noted the cost of processing additional claims. Many commenters noted that this would be particularly difficult as this additional administrative burden would come at the same time practitioners are adjusting the Merit-based Incentive Payment System (MIPS). One commenter provided the results of a survey of surgical practitioners in 20 specialties in which 30 percent of respondents stated that the cost of integrating the new G-codes into their practice would cost more than $100,000 and only about 10 percent stated that it would cost less than $25,000.

Some commenters expressed specific concern about the documentation burden that would come from using these codes. On the other hand, other commenters suggested that providers of visits during the global surgical services should be held to the same documentation standards as providers of E/M services. One stated that the “administrative burden on surgeons should be no different and certainly no less than that on non-surgeons when it comes to documenting a visit with a patient. If many surgeons currently use minimal documentation when they provide a post-operative visit that is no excuse for expecting the same inadequate level of documentation going forward. To require anything less than the same level of documentation for all clinicians providing E/M services would be irresponsible and unfair and would defeat the very purpose of documenting the actual types and extent of these services in the post-operative period.”

Response: The need for accurate, complete and useful data must be balanced with administrative burden and cost. We articulated that using a select number of G-codes based on time would impose a burden on providers, but that burden is necessary for us to comply with the statutory requirement to gather the data necessary to value global procedures. We note that CPT routinely incorporates more than 100 new codes in annual updates, and for this reason we did not anticipate that the inclusion of eight new G-codes was likely to present significant challenges to EHR systems or other infrastructure. Based on the comments we received, however, it is clear that the majority of stakeholders believe the burden is much greater than we had assumed. In general, we agree with commenters that comparable documentation is appropriate for all physicians furnishing and being paid by Medicare for similar services.

Comment: Several commenters noted that the difficulties of using these codes would affect the accuracy of the data reported. One commenter stated that the G-code proposal would be impossible to implement and “at the very least” would yield incomplete and unreliable results.

Response: We agree with commenters that implementation burden is an important consideration in determining how practitioners should report on care provided in the post-operative period and that if practitioners find the reporting requirements to be excessive and require great expenditures to incorporate into their practice, the accuracy of the data could be undermined. We considered this in determining the final policy described below.

Comment: Some commenters criticized the proposed G-codes because they were not directly linked to E/M codes or comparable to existing E/M codes. On the other hand, some commenters preferred the codes describing such visits not be linked or comparable to E/M codes to avoid confusion or unintentional, inappropriate payments. One commenter stated that the follow-up work performed within the global periods and the continuity work performed by cognitive physicians should not be represented by the same codes. Another commenter stated that the care required by a patient recovering from a procedure is fundamentally different from the typical follow-up of an established outpatient or inpatient, especially when there are multiple simultaneous interacting conditions, a single metastable chronic illness, or one or more acute exacerbated chronic illnesses that requires inpatient care and expertise.

Response: Commenters' belief that the work in follow-up visits included in the global package is not necessarily well described by the work of current E/M codes is worth exploring. Current data does not allow us to determine the validity of these commenters' assertion but given its importance, we believe it is critical to gather data on whether follow-up visits provided in the post-operative period are different than other E/M services. To the extent the services in the post-operative period are different from other E/M services, it would not make sense to use E/M codes in valuing global services as is ostensibly the case under the current process the RUC uses in developing recommended values for PFS services.

Comment: Most commenters supported using CPT codes, rather than the proposed G-codes. A few pointed to the existing E/M codes, but most recommended that any claims-based reporting use CPT code 99024, an existing CPT code that describes post-operative services in a global period. Commenters noted that since this is a current CPT code the administrative burden would be much less than that associated with using the proposed new G-codes. These commenters suggest that practitioners are likely already familiar with the code, some already use it to track services within their practice, and some others already report it to other payers. Also, they suggest that because EHR and billing systems already include CPT code 99024, it will be less costly to implement than the proposal. Some also preferred using CPT code 99024 because unlike the proposed G-codes it does not require the reporting of time units.

Most commenters disagreed that time could be a proxy for the complexity of the visit and objected to reporting time for the same reasons discussed above. These commenters did not agree that CPT code 99024 could be reported in time units as a proxy for collecting the required information about the level of visits.

Three organizations disagreed, however, stating that time is a sufficient proxy for work relativity in post-operative visits and that the number units of CPT code 99024 could reflect the complexity involved. These commenters recommended reporting data in 15-minute intervals, rather than the proposed 10-minute increments, stating that physicians are familiar with 15-minute increments and thus the use of 15-minute increments would greatly reduce the administrative burden. They recommended that CMS clearly define how time is to be reported and suggested that the 8-minute rule is already a familiar concept that could be used.

Many commenters suggested that other approaches, such as a survey, clinical registries, or on-line portals be used to collect data on level of visits.

Several commenters stated that CMS should not collect data on the level of visits based on these commenters' perspective that there is no problem with the level of visits currently used in the valuation of global packages. One commenter pointed out that only 1 percent of all established patient office visits used in valuing 10-day and 90-day global surgery packages have a visit level above a CPT code 99213. Another commenter suggested that the survey be used to collect data on the level of visits. Others suggested that RUC surveys be used to measure level of visits.

Response: We understand that stakeholders believe that using CPT code 99024 rather than the proposed G-codes will significantly lower administrative burden and lower costs related to the collection of this data. We do not have data showing that the level of visits used in valuation of global packages are correct or incorrect; to the best of our knowledge, this has never been assessed outside of the RUC process. While the current valuations for global packages rely primarily on CPT codes 99212 and 99213 for the visit component, we do not agree that this means that the levels are accurate. Further, as some commenters have made clear, there is not consensus among stakeholders that the post-operative visits are equivalent to other E/M visits. Additionally, the relationship between the number and level of visits assumed to be in the global period and the overall work RVUs for the global codes is often unclear. For all of these reasons, we disagree with commenters that we do not need to collect data on the level of services.

In addition to the statutory reference to collecting data on the level of visits, we believe that code valuations can be more accurate with more complete information. While we continue to believe that data only on the number of visits furnished would not provide data on both the number and level of visits needed for valuation of services, data on the number of visits alone is an important input in valuing global packages and having accurate data on the number of visits could be a useful first step in analyzing the global packages.

After considering the comments, we are finalizing a requirement to report post-operative visits furnished during 10- and 90-day global periods. However, rather than using the proposed set of G-codes for this reporting, we are requiring that CPT code 99024 be used to report such visits. We will not, at this time, require time units or modifiers to distinguish levels of visits to be reported. Since this code is specifically limited to post-operative care, we are only requiring reporting of post-operative visits. We expect that the reporting of this information through Medicare claims will provide us with information about the actual number of visits furnished during the post-operative periods for many services reported using global codes. Because the number of visits is a major factor in valuation of global services, we believe that examination of such information, when available, can improve the accuracy of the global codes. The use of a simple code that practitioners are familiar with should facilitate the submission of accurate information. We expect practitioners to note the visit in the medical chart documenting the post-operative visit.

Since CPT code 99024 will only provide data on the number of visits and no data on the level or resources used in furnishing the visit, we believe this is only the first step in gathering the data required by Section 1848(c)(8). The proposed G-codes could have provided information to better understand the resources used in furnishing services during global periods and in valuation of such services assuming that they could be accurately reported. However, widespread concerns from groups representing the practitioners that would be reporting these services, including concerns about the burdens regarding and the inability of physicians to track time and the need to learn a new 8-code coding system, persuade us that we should pursue less burdensome ways of obtaining information. We will assess whether these methods will lead to the collection of necessary data, including data on time and intensity, of these services.

As suggested by commenters, we will explore whether the data collected from the survey that we are conducting, which is discussed later in this preamble, can provide information on the level of visits and other resources needed to value surgical services accurately. Stakeholders should be aware that since this a new approach for collecting data, and one that has not been used previously, we are concerned that additional or different reporting will be necessary to collect data on the number and level of visits and other information needed to value surgical services as required by Section 1848(c)(8).

b. Reporting of Claims

We proposed that the G-codes detailed above would be reported for services related to and within 10- and 90-day global periods for procedures furnished on or after January 1, 2017. Services related to the procedure furnished following recovery and otherwise within the relevant global period would be required to be reported. These codes would be included on claims filed through the usual process. Through this mechanism, we would collect all of the information reported on a claim for services, including information about the practitioner, service furnished, date of service, and the units of service. By not imposing special reporting requirements on these codes, we proposed to allow practitioners the flexibility to report the services on a rolling basis as they are furnished or to report all of the services on one claim once all have been furnished, as long as the filed claims meet the requirements for filing claims.

We did not propose any special requirements for inclusion of additional data on claims that could be used for linking the post-operative care furnished to a particular service. To use the data reported on post-operative visits for analysis and valuation, we proposed to link the data reported on post-operative care to the related procedure using date of service, practitioner, beneficiary, and diagnosis. While we believed this approach to matching would allow us to accurately link the preponderance of G-codes to the related procedure, we sought comment on the extent to which post-operative care may not be appropriately linked to related procedures whether we should consider using additional variables to link these aspects of the care, and whether additional data should be required to be reported to enable a higher percentage of matching.

The following is summary of the comments we received on our proposal to require reporting on pre- and post-operative care associated with all procedures with 10- and 90-day global periods.

Comment: Many commenters objected to the proposal to require reporting on post-operative services for all 10- or 90-day global services. Some suggested that many of the global services are low volume and have little impact on Medicare spending. It was also noted that it would be difficult to obtain a meaningful sample of low-volume services. Others discussed the burden of reporting on all services. The RUC recommended that CMS only require reporting on services that are furnished by more than 100 providers and that either are furnished more than 10,000 times or have allowed charges of more than $10 million annually to obtain meaningful data for valuation. The RUC noted that many procedures were infrequently furnished and thus useful data would not be obtained. This position was supported by a significant number of commenters. In response to the stated concern about having complete data when more than one surgical service is furnished during the global period, a commenter pointed out that a review of the 2014 Medicare 5 percent sample file shows that, two surgical global codes are performed on the same date of service, by the same physician, only 18 percent of the time.

Response: The commenters are correct that the vast majority of 10- and 90-day procedures are furnished infrequently and thus have little effect on Medicare expenditures or direct impact on the valuations of other services under the PFS. We proposed to collect data on all procedures since we believed the data we collected would be more accurate if physicians reported on all services as it would be routine and would not have required physicians to determine at each pre- and post-operative visit whether or not reporting the service was required. Moreover, as pointed out by commenters, we believe that reporting on all applicable services would have provided more complete data when multiple surgeries occurred during the global period.

Having specific data on all procedures would provide specific information for each service that Medicare pays for using a global period. In assessing the likely benefit of the additional data as compared to the burden of reporting based on the comments we received, we agree with commenters that collecting the data from high volume/high cost procedures could provide adequate information to improve the accuracy of valuation of global packages overall. Even if all practitioners reported data on all procedures, it is likely that we would not receive enough data on low-volume services for the data to be reliable for use in valuations. There are more than 1,500 services that are furnished less than 100 times per year. Because of this, data that we could collect on these services would be extremely limited. We also find that data on services with low volumes are not reliable due to variability from year to year. Since we often value related services by extrapolating data on one service to other services in the family, with adjustments as necessary to reflect variations in the procedure, the data gathered on high-volume services could similarly be used to value low-volume services in the same family. As a result we, believe that the data on high-volume services can improve the accuracy of values for all 10- and 90-day services.

After consideration of the comments, we are implementing a requirement for reporting on services that are furnished by more than 100 practitioners and are either furnished more than 10,000 times or have allowed charges of more than $10 million annually as recommended by the RUC and many other commenters. Under this policy, we estimate that we would collect data on about 260 codes that describe approximately 87 percent of all furnished 10- and 90-day global services and about 77 percent of all Medicare expenditures for 10- and 90-day global services under the PFS. Given that this data would provide information on the codes describing the vast majority of 10- and 90-day global services and expenditures, it will provide significant data for valuation. For 2017, we will use the CY 2014 claims data to determine the codes for which reporting is required and display the list on the CMS Web site. In subsequent years, we will update the list to reflect more recent claims data and publish a list of codes prior to the beginning of the reporting year. The services for which reporting is required will include successor codes to those deleted or modified since CY 2014 for which reporting would have been required if the code had not been deleted or modified.

The following is summary of the comments we received on our proposal to require claims-based reporting for services related to procedures furnished on or after January 1, 2017.

Comments: Many commenters expressed concerns regarding the difficulty of making the changes required to implement this new reporting by January 1, 2017. Some commenters noted that this change was coming at the same time as the new MIPS program. Some commenters stated that the statute required a process to be in place by January 1, 2017, but that CMS has flexibility regarding when to begin the required reporting. Some commenters suggested that CMS consider conducting the proposed survey before implementing any claims-based reporting.

Response: We proposed to begin required reporting on January 1, 2017, based upon the statutory language regarding both the collection and use of the data for revaluation of services. We understand that some practices will need to make modifications to their EHR and billing systems to report this data to us. We also acknowledge that an opportunity for testing the systems and training will enhance the quality of data that we receive.

After consideration of comments, we are encouraging practitioners to begin reporting data on post-operative services for procedures furnished on or after January 1, 2017. However, the requirement to report will become mandatory for post-operative services related to procedures furnished on or after July 1, 2017 rather than as of January 1, 2017, as proposed. This delay will not negatively impact the use value of the collected data since we expect that data received early in the year might be less complete than data submitted once practitioners adjusted to the requirements. Also, by allowing time for practitioners to adjust EHR and billing software, to test such systems and to train staff, we think the quality of the data will be enhanced by providing flexibility with regard to the effective date of the requirement. Finally, because we are limiting required reporting to high-volume codes, meaningful data for CY 2017 should be available from 6 months of reporting. Our systems can now accept the post-operative visit data so practitioners can begin submitting such claims at any time.

c. Special Provisions for Teaching Physicians

We sought comment on whether special provisions are needed to capture the pre- and post-operative services provided by residents in teaching settings. If the surgeon is present for the key portion of the visit, should the surgeon report the joint time spent by the resident and surgeon with the patient? If the surgeon is not present for the key portion of the visit, should the resident report the service? If we value services without accounting for services provided by residents that would otherwise be furnished by the surgeon in non-teaching settings, subsequent valuations based upon the data we collect may underestimate the resources used, particularly for the types of surgeries typically furnished in teaching facilities. However, there is also a risk of overvaluing services if the reporting includes services that are provided by residents when those services would otherwise be furnished by a physician other than the surgeon, such as a hospitalist or intensivist, and as such, should not be valued in the global package.

Comment: We received only a few comments on this issue. Some commenters suggested using the CMS policies that apply to other services that teaching surgeons report to CMS for the reporting of CPT code 99024. More specifically, when the appropriate conditions are met they would use the GC or GE modifier to identify those services in which surgical residents are involved. One of these suggested that once we have the data we discuss with stakeholders how to use the data involving residents in future valuations. Others suggested that we capture data on resident's time as it could be important for valuation, especially for the more complex cases in a teaching facility setting. Some urged that we provide clear guidance on when the resident's time could be reported. One commenter stated that teaching physicians should be exempt from reporting requirements.

Response: These comments reinforce the importance of collecting data from teaching physicians and to do so using the existing Medicare rules that teaching physicians use in reporting services in which residents are involved in furnishing. Because we are finalizing data collection using CPT code 99024, the issues regarding the reporting of time data are no longer relevant.

After consideration of the comments, we are finalizing a requirement that teaching physicians will be subject to the reporting requirements in the same way that other physicians are. Such physicians should report CPT code 99024 only when the services furnished would meet the general requirements for reporting services and should use the GC or GE modifier as appropriate.

e. Who Reports

In both the comments on the CY 2016 proposed rule and in input from the January 2016 national listening session, there was a great deal of discussion regarding the challenges that we are likely to encounter in obtaining adequate data to support appropriate valuation. Some indicated that a broad sample and significant cooperation from physicians would be necessary to understand what is happening as part of the global surgical package. One commenter suggested that determining a representative sample would be difficult and, due to the variability related to the patient characteristics, it would be easier to have all practitioners report. Many suggested that we conduct an extensive analysis across surgical specialties with a sample that is representative of the entire physician community and covers the broad spectrum of the various types of physician practice to avoid problems that biased or inadequate data collection would cause. Suggestions of factors to account for in selecting a sample include specialty, practice size (including solo practices), practice setting, volume of claims, urban, rural, type of surgery, and type of health care delivery systems. Another commenter pointed out that small sample sizes may lead to unreliable data. Some commenters stated that requiring all practitioners to report this information is unreasonable and would be an insurmountable burden. A participant acknowledged that it would be difficult for practitioners to report on only certain procedures, while another stated that this would not be an administrative burden.

After considering the input of stakeholders on the CY 2016 proposed rule and at the January 2016 national listening session discussed above, we proposed that any practitioner who furnishes a procedure that is a 10- or 90-day global service report the pre- and post-operative services furnished on a claim using the proposed G-codes. We agreed with stakeholders that it would be necessary to obtain data from a broad, representative sample. However, as we struggled to develop a nationally representative sampling approach that would result in statistically reliable and valid data, it became apparent that we do not have adequate information about how post-operative care is delivered, how it varies and, more specifically, what drives variation in post-operative care to develop a sampling frame. In its work to develop the coding used for its study, the contractor found a range of opinions on what drives variation in post-operative care. (The report is available on the CMS Web site under downloads for the CY 2017 PFS proposed rule with comment period at http://www.cms.gov/physicianfeesched/downloads/.) Without information on what drives variation in pre- and post-operative care, we would have to speculate about the factors upon which to base a sample or assume that the variation in such care results from the same variables as are frequently identified for explaining variation in health care and clinical practice. In addition, we expressed concern about whether a sample could provide sufficient volume to value accurately the global package, except in the case of a few high-volume procedures.

In addition to concerns about achieving a statistically representative sample of all practitioners nationally, we noted in the proposed rule significant operational concerns with limiting data collection to a subset of practitioners or a subset of services. These include how to gain sufficient information on practitioners to stratify the sample, how to identify the practitioners who must report, and for those who practice in multiple settings or with multiple groups in which settings the practitioner would report. We concluded that establishing the rules to govern which post-operative care should be reported based on our proposed G-codes would be challenging for us to develop and difficult for physicians to apply in the limited time between the issuance of the CY 2017 PFS final rule with comment period and the beginning of reporting on January 1, 2017. We do not believe that the same problems apply to the same extent to our final policy to use a single code that already exists to report services described only by codes reported in high volumes. For example, implementation of new sets of codes associated with annual PFS updates are often supported by informational and educational efforts undertaken by national organizations, like the national medical specialty societies. Given that many practitioners are already familiar with CPT code 99024 (as noted by many commenters), the need for such efforts is significantly mitigated.

We also noted in the proposed rule that the more robust the reported data, the more accurate our ultimate valuations can be. We stated that given the importance of data on visits in accurate valuations for global packages, collecting data on all pre- and post-operative visits in the global period is the best way to accurately value surgical procedures with global packages.

We recognized that reporting would require submission of additional claims by those practitioners furnishing global services, but indicated that we believed the benefits of accurate data for valuation of services merited the imposition of this requirement. By using the claims system to report the data, we believed the additional burden would be minimized and referred to stakeholder reports that many practitioners are already required by their practice or health care system to report a code for each visit for internal control purposes and some of these systems already submit claims for these services, which are denied. We noted that requiring only some physicians to report this information, or requiring reporting for only some codes, could actually be more burdensome to physicians than requiring this information from all physicians on all services because of the additional steps necessary to determine whether a report is required for a particular service and adopting a mechanism to assure that data is collected and reported when required. Moreover, we stated that the challenges with implementing a limited approach at the practice level as compared to a requirement for all global services would result in less reliable data being reported.

We noted that as we analyzed the data collected and made decisions about valuations, we would reassess the data needed and what should be required from whom. Through the data collected under our proposal, we indicated that we would have the information to assess whether the post-operative care furnished varies by factors such as specialty, geography, practice setting, and practice size, and thus, the information needed for a sample selection to be representative.

While section 1848(c)(8)(B) of the Act requires us to collect data from a representative sample of physicians on the number and level of visits provided during the global period, we stated that it does not prohibit us from collecting data from a broad set of practitioners. In addition, section 1848(c)(2)(M) of the Act authorizes the collection of data from a wide range of physicians. Given the benefits of more robust data, including avoiding sample bias, obtaining more accurate data, and facilitating operational simplicity, we noted that we believed collecting data on all post-operative care initially is the best way to undertake an accurate valuation of surgical services in the future.

The following is a summary of the comments that we received on our proposal to require all practitioners furnishing 10- or 90-day global services to submit claims for the pre- and post-operative services furnished.

Comment: Commenters overwhelmingly opposed requiring all practitioners to submit claims for postoperative services. Several reasons were cited for the opposition. The most significant reason was the administrative burden and costs to physicians. Many commenters also stated that requiring all practitioners who furnish 10- or 90-day global services to report data is counter to the statute because the statute refers to collection of data from a representative sample of physicians.

One commenter stated that requiring every practitioner to report these codes will be in many ways less representative than a targeted sample, explaining that given the limited time for education, only large, technologically rich practices will have the ability to properly report these services. The commenter noted that this will leave many, smaller or rural practices without the proper education and robust billing systems in place to adequately, if at all, report these G-codes. The commenter also noted that smaller, rural practices have smaller patient populations, which can often be older and sicker than the typical patient seen in a large practice and by creating a complex system that favors one type of practice, the collected data is more likely to be biased rather than representative. Another commenter suggested that a small number of representative practices could provide us with the same level of accuracy as collected data from all physicians.

Response: In response to commenters' opposition to our proposal to require all providers of covered services to report data, we acknowledge that the stakeholders describe a much larger burden from using the G-codes than we anticipated. On the other hand, we also believe that our final policy will result in a much lower burden than the proposed policy would have. As noted above, we are not finalizing the proposed requirements to use the G-codes or the proposed requirement to report on all 10- and 90-day global procedures and thus, we believe that the overall administrative burden is significantly reduced.

We do not agree with commenters that state that we do not have the statutory authority to require reporting by all practitioners furnishing certain services. We point commenters to section 1848(c)(2)(M) of the Act, which authorizes the collection of data to use in valuing PFS services. We continue to believe that section 1848(c)(8) of the Act requires us to collect data that is representative. We also continue to believe that requiring all practitioners to report is more likely to be representative than a sample given our lack of information about what drives variation in post-operative care. However, after considering the information presented by commenters regarding the difficulties that would be placed on many physicians by the proposal, we believe that requiring reporting by all practitioners for CY 2017 may present unforeseen, alternative impediments to the sample being nationally representative of all practitioners, such as practitioners being unable to report data accurately due to constraints of time, finances or technical ability.

Comment: We did not receive any comments on the appropriate sample size. Nor did we receive data on variations in the delivery of post-operative care in response to our concern that we lacked data on how post-operative care was delivered to select a representative sample. Many commenters stated that it was possible to select a representative sample, but none provided details on how to do so.

Several commenters suggested broadly sampling using the characteristics that are frequently used for health care sampling generally, such as geographic areas, urban and rural, practice types, practice sizes, specialties and academic and non-academic. One commenter recommended that we select a sample using geographical data to identify a sample including practices of all sizes. The commenter suggested, for example, that large hospital-based practices often have practice patterns that are different from the majority of the practicing physicians in suburban and rural areas. Another commenter stated that we should not only collect data from MSAs but also from rural and less urban areas.

One commenter suggested that we consider phasing in the requirement, perhaps starting with larger groups. The commenter stated that through one of these approaches we could avoid “burdening providers with unfunded work that has not yet been tested.”

One commenter suggested that we use a geographic sampling approach similar to that one used for Comprehensive Care for Joint Replacement (CJR) model or the episode payment models proposed for cardiac and surgical hip/femur fraction and modify it to choose a geographic sampling unit of MSAs and non-MSAs.

Response: We agree with commenters that we could select a sample using an approach typically used in health care surveys or in Medicare models and other programs. To the extent that the delivery of post-operative care varies only based upon the criteria we selected, a sample based on being representative for that criteria would be likely to produce valid data.

However, instead of sampling by practice or practitioner or type of service, a geographic approach to sampling (for example, sampling all practitioners in a selected state) could help to alleviate the need to stratify the sample on a long list of criteria. By using broad geographical areas from varied areas of the country, we believe our sample will capture data from practitioners who practice in a variety of settings, single and multispecialty practices, urban and rural, a variety of medical specialties, and practitioners operating in both academic and non-academic institutions. Surgeons interviewed for the G-code development suggested that post-operative care might vary across these dimensions. A geographic approach could also mitigate some of the practical operational barriers. For example, we believe that by having all practitioners in the practice participate in reporting, we avoid concerns about incomplete data when a required reporter furnishes a procedure and another practitioner in the practice furnishes the post-operative visits. A geographic approach also makes it easier to educate practitioners on data collection requirements.

Comment: In response to operational difficulties with a representative sample, such as how to make sure participants were aware of the requirement to report and how to do so, one commenter stated that notifying a small targeted sample is a much smaller task than notifying the entire population of participating Medicare practitioners. They also stated that a targeted approach will encourage open dialogue between the participating practices and CMS, ensuring the data collected are reliable. Others suggested providing compensation for a sample of physicians to submit detailed data, would lead to capturing accurate data because they would more likely to understand and prioritize reporting because of their participation in this type of study.

Response: We disagree that it is operationally easier to notify a small segment of broadly diverse practitioners than the entire population of practitioners unless that small segment has a degree of cohesiveness, such as being in the same geographic area or specialty. We have long appreciated the stakeholder community's collaboration in broad communication efforts. In general, we have found that when something affects a small number of providers it does not receive the same response from entities that are critical for widespread adoption such as associations, who are key purveyors of information, and those developing software systems. We appreciate the suggestion that interaction among those that need to report will facilitate compliance and the quality of the data. With regard to compensation, we note that the statute provided for a 5 percent withhold to encourage compliance and we chose not to propose to implement this provision.

After consideration of the comments, we are finalizing a requirement for reporting that only applies to practitioners in selected states. In addition, those practicing only in small practices are excluded from required reporting. Those not required to report can do so voluntarily and we encourage them to do so.

Geographic Sample

As we noted in the proposed rule, we do not have adequate data on what drives variations in the delivery of pre- and post-operative care to design a sampling methodology that is certain to be representative. We also believe that submission by all practitioners would be consistent with our extensive use of claims data for other PFS services. Additionally, we understand the statute directs us to gather data from more than a select group of practitioners based on any particular attributes, such as gathering data only from “efficient” practices, consistent with longstanding recommendations from MedPAC regarding limiting data collection. We also believed that there were significant operational impairments to data reporting by a limited sample of physicians. In consideration of these factors, we proposed to require reporting by all physicians to make sure that the data we obtained reflected all services furnished. In light of the comments regarding the burden that would be created by requiring reporting by all physicians and the data that was actually needed for valuation, we think that reporting by a subset of practitioners could provide us valuable information on the number of visits typically furnished in global periods. This data could enhance the information we currently use to establish values for these services. While we acknowledge that we believe the data under this less burdensome approach will provide less information than necessary for optimal valuation for these services, we believe that the information on the number of actual visits from a subset of practitioners is preferable to the information on which we currently rely, which is the results of survey data reflecting respondents' assessment of the number of visits considered to be typical.

One commenter suggested that we could develop a geographic sample using a similar approach used by the Center for Medicare and Medicaid Innovation for the Comprehensive Care for Joint Replacement (CJR) or other proposed episode payment models, with an adjustment that would make certain we received data from rural, as well as urban areas. We reviewed these approaches and concluded that such an approach for sample selection could maximize the variability of the sample, mitigate some of our concerns, and provide a robust set of data for consideration.

Commenters suggested a sample should include geographic diversity. Studies show that health care delivery patterns often vary between geographic areas and while we have no specific information that the number of post-operative visits varies by geographic areas, it seems prudent to gather data from a variety of geographic areas to determine if there is such variation and to account for it in our data collection if it exists. In order to maximize the variability of our limited sample, we are using a methodology that requires reporting from practices in 9 states of various sizes and from various geographic areas of the country. We are using whole states for the geographic areas rather than MSAs as are used for the CJR and proposed for other models for several reasons. First, MSAs are not used for geographic adjustments under the PFS. Indeed, practitioners in most states receive state-wide geographic adjustments under the PFS. Additionally, an MSA-based approach would, by definition, not include large rural areas, something mentioned by many commenters as an important factor in variation in medical practice, and therefore, a critical criterion for sampling. Also, due to a variety of governmental and institutional requirements, the practice of medicine is primarily a state-based activity and thus the use of states will reduce the number of practitioners for whom we have only partial data based on geographic location. In contrast, we believe that practitioners often practice across county lines or in more than one MSA. We also believe that the state-wide approach will be helpful for compliance and education because there are state medical associations in every state and specialty associations in many.

To make sure that we had states of a variety of sizes, we ranked states according to the number of Medicare beneficiaries in each state. We chose the number of Medicare beneficiaries to reflect the general need for Medicare services. We divided states into four groups: The top 5 states in terms of the number of Medicare beneficiaries (group 1); 6th through 15th largest states in terms Medicare beneficiaries (group 2); the 16th through 25th largest states in terms of Medicare beneficiaries (group 3); and all remaining states (26 including the District of Columbia, group 4). The states in each group are:

• Group 1—California, Florida, New York, Pennsylvania & Texas.
• Group 2—Georgia, Illinois, Massachusetts, Michigan, New Jersey, North Carolina, Ohio, Tennessee, Virginia, and Washington.
• Group 3—Alabama, Arizona, Indiana, Kentucky, Louisiana, Maryland, Minnesota, Missouri, Wisconsin, and South Carolina.
• Group 4—Alaska, Arkansas, Colorado, Connecticut, District of Columbia, Delaware, Hawaii, Idaho, Iowa, Kansas, Maine, Mississippi, Montana, Nebraska, Nevada, New Hampshire, New Mexico, North Dakota, Oklahoma, Oregon, Rhode Island, South Dakota, Utah, Vermont, West Virginia and Wyoming.

We also recorded the Census region for each state using the Census Bureau's nine regions (New England, Middle Atlantic, South Atlantic, East South Central, West South Central, East North Central, West North Central, Mountain, and Pacific). Puerto Rico and other territories were excluded.

To ensure a mix of states in terms of size (measured by number of Medicare beneficiaries), we selected 1 state at random from group 1, followed by 2 states each at random from groups 2 and 3, and lastly 4 states from group four. After each random selection, we eliminated the remaining states in the same Census region from the remaining groups for which selection was pending to maximize geographic variation in the selection of states. In the event that this process resulted in fewer than 9 selected states (for example if none of the three Middle Atlantic states—all in Group 1 and 2—were selected in the first three picks), the last selection(s) were made randomly from states in the remaining Census region from which selections previously had not been made.

Practitioners located in the following states who meet the criteria for required reporting will be required to report the data discussed in this section of the final rule:

• Florida.
• Kentucky.
• Louisiana.
• Nevada.
• New Jersey.
• North Dakota.
• Ohio.
• Oregon.
• Rhode Island.

Exclusion for Practitioners in Small Practices

In response to comment about the burden of our proposed requirement and the concern that the burden would result in the submission of data of poor quality, we are exempting practitioners who only practice in practices with fewer than 10 practitioners from the reporting. Based upon the comments, we believe larger practices are more likely to currently require practitioners to track all visits and often use CPT code 99024 to do so. Moreover, larger practices are more likely to have coding and billing staff that can more easily adapt to this claims-based requirement. The combination of experience with reporting CPT code 99024 and the staff and resource base to devote to developing the infrastructure for such reporting will result in greater accuracy from such practitioners. By excluding practitioners who only practice in practices with fewer than 10 practitioners, we estimate that about 45 percent of practitioners will not be required to report. In defining small practices, we reviewed other programs. We chose 10 practitioners as the threshold for reporting as practices of this size are large enough to support coding and billing staff, which will make this reporting less burdensome. Also, this is the same threshold used by the value-based modifier program for its phase-in of a new requirement because of concerns about the burden of small practices.

For this purpose, we define practices as a group of practitioners whose business or financial operations, clinical facilities, records, or personnel are shared by two or more practitioners. For the purposes of this reporting requirement, such practices do not necessarily need to share the same physical address; for example, if practitioners practice in separate locations but are part of the same delivery system that shares business or financial operations, clinical facilities, records, or personnel, all practitioners in the delivery system would be included when determining if the practice includes at least 10 practitioners. Because qualified non-physician practitioners may also furnish procedures with global periods, the exception for reporting post-operative visits applies only to practices with fewer than ten physicians and qualified non-physician practitioners regardless of specialty. We are including all practitioners and specialties in the count because the exception policy uses practice size as a proxy for the likely ability of the practice to meet the reporting requirements without undue administrative burden. We recognize that physicians and qualified non-physician practitioners furnish services under a variety of practice arrangements. In determining whether a practitioner qualifies for the exception based on size of the practice, all physicians and qualified non-physician practitioners that furnish services as part of the practice should be included. This would include all practitioners, regardless of whether they are furnishing services under an employment model, a partnership model, or an independent contractor model under which they practice as a group and share facility and other resources but continue to bill Medicare independently instead of reassigning benefits. We also recognize that practice size can fluctuate over the year and anticipate that practices will determine their eligibility for the exception based on their expected staffing. Generally, practitioners in short-term locum tenens arrangements would not be included in the count of practitioners. When practitioners are also providing services in multiple settings, the count may be adjusted to reflect the estimated proportion of time spent in the group practice and other settings.

Although this policy excludes a significant number of practitioners, a majority of the global procedures furnished will be included in the reporting requirements and thus we will have data on a majority of services.

Several commenters also expressed concern that data from small practices be included to have complete information. If those practicing in small practices are motivated to report and either have the infrastructure to do so in place or the resources to develop such infrastructure, then, taken together, these attributes would minimize concerns with accuracy of data from small practices. Accordingly, we are encouraging, but not requiring, small practices to report the visits. As we collect data, we will explore mechanisms to appropriately use the voluntarily submitted claims data. Analysis of this and other data we are able to procure will allow us to assess whether the number of post-operative visits varies based upon the size of practice. To the extent that it does and that we do not have adequate data on the practice patterns in small practices from voluntarily submitted data and other sources, we will reconsider for future notice and comment rulemaking the exemption of practitioners in small practices from the reporting requirements.

The claims data received from practitioners in these states will provide more information about the number of visits typically provided in post-operative periods than is available from any other source. Through analysis of this data, we hope to learn more about what drives variations in the delivery of post-operative care. Many of the characteristics that were suggested by commenters, such as size of practice, type of practice, geographic, urban/rural, academic, hospital based, specialty, etc., will be able to be evaluated using the claims data. Moreover, we hope to be able to stratify the data received based upon comparisons to the national characteristics so that the submitted claims data can contribute to improved valuation of PFS services.

In summary, our claims-based data collection policy requires that, for procedures furnished on or after July 1, 2017, practitioners who practice in practices that includes of 10 or more practitioners in Florida, Kentucky, Louisiana, Nevada, New Jersey, North Dakota, Ohio, Oregon, and Rhode Island will be required to report on claims data on post-operative visits furnished during the global period of a specified procedure using CPT code 99024. The specified procedures are those that are furnished by more than 100 practitioners and either are nationally furnished more than 10,000 times annually or have more than $10 million in annual allowed charges. The final list of codes subject to required reporting will be available on the CMS Web site. Although required reporting begins for global procedures furnished on or after July 1, 2017, we encourage all practitioners to begin reporting for procedures furnished on or after January 1, 2017, if feasible. Similarly, we encourage those practicing in practices with fewer than 10 practitioners to report data if they can do so.

(1) Survey of Practitioners

We agreed with commenters on the CY 2016 proposed rule and at the listening session that we need more information than is currently provided on claims and that we should utilize a number of different data sources and collection approaches to collect the data needed to assess and revalue global surgery services. In addition to the claims-based reporting, we proposed to survey a large, national sample of practitioners and their clinical staff in which respondents would report information about approximately 20 discrete pre-operative and post-operative visits and other global services like care coordination and patient training. This sample would be stratified based upon specialty and geography, as well as by physician volume (procedures billed) and practice setting. The proposed survey would produce data on a large sample of pre-operative and post-operative visits and is being designed so that we could analyze the data collected in conjunction with the claims-based data that we would be collecting. We expect to obtain data from approximately 5,000 practitioners.

We noted that, if our proposal was finalized, RAND would develop and conduct this survey. RAND would also assist us in collecting and analyzing data for this survey and the claims-based data. While the primary data collection would be via a survey instrument, semi-structured interviews would be conducted and direct observations of post-operative visits would occur in a small number of pilot sites to inform survey design, validate survey results, and collect information that is not conducive to survey-based reporting.

Our proposed sampling approach would sample practitioners rather than specific procedures or visits to streamline survey data collection and minimize respondent burden. Specifically, we will use a random sample from a frame of practitioners who billed Medicare for more than a minimum threshold of surgical procedures with a 10- or 90-day global period (for example, 200 procedures) in the most recent available prior year of claims data. The sampling frame would provide responses from approximately 5,000 practitioners, stratified by specialty, geography, and practice type. Based upon preliminary analysis, we believe this number of participants will allow us to collect information on post-operative care following the full range of CPT level-2 surgical procedure code groups. For many common types of post-operative visits, we anticipate a standard deviation of the time distribution at around 9 minutes. To achieve a 95 percent confidence intervals with a width of 2 minutes, we would need 311 reported post-operative visits per procedure/procedure group. The most comprehensive approach would be to sample sufficient practitioners to observe 311 post-operative visits for each HCPCS procedure, but this approach would be cost- and time-prohibitive. Since post-operative care following similar procedures may involve similar activities and times even if there are differences in the number of visits, we proposed to sample differentially by specialty to maximize our ability to estimate attributes of post-operative care for the largest range of procedures.

Sample sizes for each specialty will be determined on the basis of number of procedures billed by the specialty and number of practitioners billing, assuming a uniform distribution of procedures across the year, an average of 2 post-operative visits by each patient and an equal distribution of procedures across practitioners within a specialty. If the procedure represented only 5 percent of total billed procedures for the specialty, we could expect only one of 20 visits sampled and reported by each practitioner would be for the particular procedure, and thus we would need to sample 311 practitioners within the specialty to achieve the target precision level on estimated post-operative visit time.

We propose targeting 311 reporting practitioners from each specialty which is the only specialty contributing at least 5 percent of billings for any one procedure group code, defined as procedures sharing a CPT level 2 heading. For other specialties, the target will be defined by the maximum value of 311 divided by the number of specialties contributing at least 5 percent for any procedure group code for which that specialty contributes. The target sample size for a specialty will be capped at 25 percent of the eligible practitioners within the specialty. For example, if a specialty contributed to two procedure group codes, one of which had four contributing specialties and the other had three contributing specialties, the specialty of interest would have a target of 104 reporting practitioners (which is driven by the procedure group code that is tied to three specialties). These guidelines will target at least 311 reporting practitioners for each procedure group code, and result in a total target sample size of 4,872 providers. A smaller sample size would reduce the precision of estimates from the survey and more importantly risk missing important differences in post-operative care for specific specialties or following different types of surgical procedures. We expect a response rate in excess of 50 percent. Given this response rate (and some uncertainty in this response rate estimate), we will need to approach at least 9,722 practitioners for our target of 4,872 practitioners. Should the response rate be lower than expected, we will continue to sample in waves until we reach the target of approximately 4,872 practitioners. Non-response bias will be assessed by comparing available characteristics of non-respondents (for example, practice type, geography, procedure volume etc.) to those of respondents.

We did not propose that respondents report on the entire period of post-operative care for individual patients, as a 90-day follow-up window (for surgeries currently with a 90-day global period) is too long to implement practically in this study setting and would be more burdensome to practitioners. Instead, we proposed to collect information on a range of different post-operative services resulting from surgeries furnished by the in-sample practitioner prior to or during a fixed reporting period.

Practitioners will be asked to describe 20 post-operative visits furnished to Medicare beneficiaries or other patients during the reporting period. The information collected through the survey instrument, which will be developed based upon direct observation and discussions in a small number of pilot sites, will include contextual information to describe the background for the post-operative care, including, for example:

• Procedure codes(s) and date of service for procedure upon which the global period is based.
• Procedure place of service.
• Whether or not there were complications during or after the procedure.
• The number in sequence of the follow-up visit (for example, the first visit after the procedure).

The survey instrument will also collect information on the visit in question including, for example:

• Which level of visit using existing billing codes.
• Specific face-to-face and non-face-to-face activities furnished on the day of the visit.
• The total time spent on face-to-face and non-face-to-face activities on the day of the visit.
• Direct practice expense items used during the visit, for example supplies like surgical dressings and clinical staff time.

Finally, the instrument will ask respondents to report other prior or anticipated care furnished to the patient by the practice outside of the context of a post-operative visit, for example non-face-to-face services.

The survey approach will complement the claims data collection by collecting detailed information on the activities, time, intensity, and resources involved in delivering global services. The resulting visit-level survey data would allow us to explore in detail the variation in activities, time, intensity, and resources associated with global services within and between physicians and procedures, and would help to validate the information gathered through claims. A summary of the work that RAND would be doing is available on the CMS Web site under downloads for the CY 2017 PFS proposed rule with comment period at http://www.cms.gov/physicianfeesched/downloads/.

The following is a summary of the comments that we received on our proposal to conduct a survey of practitioners furnishing 10- and 90-day global services to obtain information about the face-to-face activities and other activities included in post-operative care.

Comment: Most commenters were generally supportive of the survey effort and noted that the provider survey will collect useful information on the level of visits, as well as important contextual detail that will not be available from the claims-based reporting. One commenter stated that a limited approach through surveys of physicians and practices looking at a targeted selection of services, and using CPT code 99024 for the claims based component would yield meaningful and actionable data for the agency and stakeholders.

Response: We agree that the survey portion of the data collection approach will provide useful information on level and context. The survey will complement claims-based reporting and will provide us with important information on non-face-to-face activities and other activities that are not reported with CPT code 99024.

Comment: One commenter pointed out challenges in survey response and in estimating time for visits by aggregating practitioner time estimates for specific activities.

Response: While we have not finalized the design of the survey instrument, we are aware of challenges in collecting detailed time estimates for specific activities. We do not intend to sum estimated times for specific activities to arrive at a total duration for the visit. We also recognize the challenges related to survey response rates and are working with our contractor accordingly.

Comment: Several commenters suggested that the survey effort should not target all 4,200 procedure codes.

Response: The survey component of the data collection effort is not designed to collect information on visits following all global procedure codes. Rather, we expect the sample to be stratified by specialty and to result in a sufficient qualitative data to address key procedures in each specialty furnishing procedures with global periods.

Comment: Some commenters believed that the purpose of the direct observation component of the data collection effort was unclear.

Response: The direct observation component will consist of external observers capturing the activities conducted in a sample of post-operative visits at a small number of practices. It is designed to provide additional context to inform future data collection efforts and to gauge where the practitioner survey does or does not capture the full range of activities. It is not a data collection activity per se.

After consideration of the comments, we are finalizing our proposal to conduct a survey of practitioners to gain information on post-operative activities to supplement our claims-based data collection as proposed. We expect that the survey will be in the field mid-2017.

(2) Required Participation in Data Collection

Using the authority we are provided under sections 1848(c)(8) and 1848(c)(2)(M) of the Act, we proposed to require all practitioners who furnish a 10- or 90-day global service to submit a claim(s) providing information on all services furnished within the relevant global service period in the form and manner described in this section of the final rule, beginning with surgical or procedural services furnished on or after January 1, 2017. We also proposed to require participation by practitioners selected for the broad-based survey through which we proposed to gather additional data needed to value surgical services, such as the clinical labor and equipment involved that cannot be efficiently collected on claim (see below).

Given the importance of the proposed survey effort, making sure that we get valid data is critical. By eliminating the bias that would be associated with using only data reported voluntarily, we stated that we expected to get more accurate and representative data. In addition to the potential bias inherent in voluntary surveys, we expressed concern that relying on voluntary data reporting would limit the adequacy of the volume of data we obtain, would require more effort to recruit participants, and may make it impossible to obtain data for valuation for CY 2019 as required by the statute.

Based on our previous experience with requesting voluntary cooperation in data collection activity, voluntary participation poses a significant challenge in collection and use of data. Specifically, the Urban Institute's work (under contract with us) to validate work RVUs by conducting direct observation of the time it took to furnish certain elements of services paid under the physician fee schedule provides evidence of this challenge. (See https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/Downloads/RVUs-Validation-Urban-Interim-Report.pdf for an interim report that describes challenges in securing participation in voluntary data collection.) Similarly, we routinely request invoices on equipment and supplies that are used in furnishing PFS services and often receive no more than one invoice. These experiences support the idea that mandatory participation in data collection activities is essential if we are to collect valid and unbiased data.

Section 1848(a)(9) of the Act authorizes us, through rulemaking, to withhold payment of up to 5 percent of the payment for services on which the practitioner is required to report under section 1848(c)(8)(B)(i) of the Act until the practitioner has completed the required reporting. Some commenters opposed the imposition of this payment consequence for failure to report, and others stated that it was too large a penalty. While withholding a portion of payment would encourage practitioners to report the required information, we did not propose to implement this option for CY 2017. We stated that requiring physicians to report the information on claims, combined with the incentive to report complete information so that revaluations of payment rates for global services are based on accurate data, would result in compliance with the reporting requirements. However, we noted that if we find that compliance with required claims-based reporting is not acceptable, we would consider in future rulemaking imposing up to a 5 percent payment withhold as authorized by the statute.

Consistent with the requirements of section 1848(c)(2)(M) of the Act, should the data collected under this requirement be used to determine RVUs, we will disclose the information source and discuss the use of such information in such determination of relative values through future notice and comment rulemaking.

The following is a summary of the comments we received on our proposal to require reporting in the claims-based survey and participation in the survey.

Comment: Many commenters objected overall to the administrative burden of our proposal and questioned the need for some of the data we were proposing to collect, primarily through the claims-based reporting, and made many recommendations for less burdensome data collection to achieve our goals. Some objected to any claims-based reporting at this time. A few recommended a different approach that involved collecting information from a small number of practices that agree to participate and that we pay such practices for participation. However, none recommended that we go forward with data collection on a totally voluntary basis. Some indicated concern that practitioners would not provide required information.

Response: We appreciate the many ideas for how to improve our data collection effort, particularly those that provided information on how to collect the information that we need while imposing a lower administrative burden on practitioners.

Comment: A few commenters supported our not proposing to implement the 5 percent withhold until claims on the post-operative care were submitted.

Response: We appreciate the support of commenters.

After considering the comments, we are finalizing our proposal to require participation in the claims-based reporting. It should be noted, however, due to our modifying the requirement to apply only to those identified as part of the geographic sample, on selected procedures, using one code, and exempting those practicing in groups with fewer than 10 practitioners, as discussed above, the impact of the requirement is significantly reduced overall, including for the subset of practitioners who will have to report under the finalized requirements.

We are not implementing the statutory provision that authorizes a 5 percent withhold of payment for the global services until claims are filed for the post-operative care, if required. We reiterate that should we find that compliance with required claims-based reporting limits confidence in the use of the information for improving the accuracy of payments for the global codes, we would consider in future rulemaking imposing up to a 5 percent payment withhold as authorized by the statute.

(3) Data Collection From Accountable Care Organizations (ACOs)

We are particularly interested in knowing whether physicians and practices affiliated with ACOs expend greater time and effort in providing post-operative global services in keeping with their goal of improving care coordination for their assigned beneficiaries. ACOs are organizations in which practitioners and hospitals voluntarily come together to provide high-quality and coordinated care for their patients. Because such organizations share in the savings realized by Medicare, their incentive is to minimize post-operative visits while maintaining high quality post-operative care for patients. In addition, we believe that such organizations offer us the opportunity to gain more in-depth information about delivery of surgical services.

We proposed to collect data on the activities and resources involved in delivering services in and around surgical events in the ACO context by surveying a small number of ACOs (Pioneer and Next Generation ACOs). Similar to the approach of the more general practitioner survey, this effort would begin with an initial phase of primary data collection using a range of methodologies in a small number of ACOs; development, piloting, and validation of an additional survey module specific to ACOs. A survey of practitioners participating in approximately 4 to 6 ACOs using the survey instrument along with the additional ACO-specific module will be used to collect data from on pre- and post-operative visits.

The following is summary of the comments we received about our proposal for data ACO data collection.

Comment: Several commenters supported a separate survey of practitioners participating in ACOs. One commenter agreed with CMS that this data collection effort may provide a unique and useful perspective on the matter at hand. Several commenters indicated that there are likely differences in pre- and post-operative care between practitioners who do participate in ACOs and those that do not. One commenter cautioned against extrapolating information gathered from ACOs to value global surgery services that are provided outside of the ACO setting because ACOs are structured differently than other practice settings and data from ACOs may, therefore, be skewed [and] that ACO participants typically are larger practices and thus would underrepresent smaller or solo practitioners.

Response: We agree that ACOs may be structured differently than other practice settings and that these differences may contribute to variations in the provision of outpatient care. By separately surveying ACOs we will be able to investigate whether there are differences in pre- and post-operative care in ACO settings compared to non-ACO settings.

After consideration of the comments received, we are finalizing our proposal for data collection in ACOs. We recognize and will continue to consider the concerns raised by commenters as we implement this project.

(6) Re-Valuation Based Upon Collected Data

We recognize that the some of the data collection activities being undertaken vary from how information is currently gathered to support PFS valuations for global surgery services. However, we believe the proposed claims-based data collection is generally consistent with how claims data is reported for other kinds of services paid under the PFS. We believe that the authority and requirements included in the statute through the MACRA and PAMA were intended to expand and enhance data that might be available to enhance the accuracy of PFS payments. In the proposed rule, we indicated that because these are new approaches to collecting data and in an area—global surgery—where very little data has previously been collected, we cannot describe exactly how this information would be used in valuing services. What is clear is that the claims-based data would provide information parallel to the kinds of claims-data used in developing RVUs for other PFS services and that by collecting these data, we would know far more than we do now about how post-operative care is delivered and gain insight to support appropriate packaging and valuation. We would include any revaluation proposals based on these data in subsequent notice and comment rulemaking.

Even though we did not make a proposal regarding how future re-valuations would use the data collected under these proposals, we received several comments on such revaluations. The following is summary of the comments we received regarding use of the data we obtain through this three-pronged data collection activity in future re-valuations.

Comment: Some commenters stated that the RUC process worked well to value services and should continue to be used to value these and other services. Some of these objected to any claims-based data collection for a variety of reasons including that it was unlikely to provide valid and reliable data, that the RUC process worked well and should continue to be used, and the that since other codes would not be valued on the basis of similar data use of this data would harm the fee schedule's relativity. Some suggested that we use the data obtained here to identify misvalued codes and refer them to the RUC for further evaluation under the usual process. Some commenters suggested that we not collect any data until we could describe how it would be used.

Response: We believe that the Congress enacted the two data collection provisions included in the Act to further the accuracy of PFS rates by having additional data available to the RUC as it makes recommendations to us and to us to inform our evaluation of those recommendations. We do not believe this data collection was intended to replace the RUC or the processes that have been established over the last two decades for valuing physician services. We agree with commenters that one way the data might be used is to identify potentially misvalued codes for the RUC to evaluate. However, we also stress that we do not agree that the use of claims data to value services within global surgery packages would be inconsistent with the valuation of other PFS services. On the contrary, very few other PFS services include estimated work RVUs based on face-to-face patient encounters over multiple days or months. Outside of these services, work RVUs are estimated per patient encounter (or in other cases over longer periods of time for non-face-to-face work). Therefore, the outer limit of any misvaluation between the estimated typical and the actual is the overall value for a single face-to-face service. Under the global packages, potential misvaluations can range from the difference between the estimated typical services for a full global period and the actual services furnished for a full global period for a given patient. We are not finalizing any provisions regarding valuation of global surgical services. Instead, such issues will be addressed in future rulemaking after we collect data and analyze data.

E. Improving Payment Accuracy for Primary Care, Care Management and Patient-Centered Services

1. Overview

In recent years, we have undertaken ongoing efforts to support primary care and patient-centered care management within the PFS as part of HHS' broader efforts to achieve better care, smarter spending and healthier people through delivery system reform. We have recognized the need to improve payment accuracy for these services over several years, especially beginning in the CY 2012 PFS proposed rule (76 FR 42793) and continuing in each subsequent year of rulemaking. In the CY 2012 proposed rule, we acknowledged the limitations of the current code set that describes evaluation and management (E/M) services within the PFS. For example, E/M services represent a high proportion of PFS expenditures, but have not been recently revalued to account for significant changes in the disease burden of the Medicare patient population and changes in health care practice that are underway to meet the current population's health care needs. These trends in the Medicare population and health care practice have been widely recognized in the provider community and by health services researchers and policymakers alike. We believe the focus of the health care system has shifted to delivery system reforms, such as patient-centered medical homes, clinical practice improvement, and increased investment in primary and comprehensive care management/coordination services for chronic and other conditions. This shift requires more centralized management of patient needs and extensive care coordination among practitioners and providers, often on a non-face-to-face basis across an extended period of time. In contrast, the current CPT code set is designed with an overall orientation to pay for discrete services and procedural care as opposed to ongoing primary care, care management and coordination, and cognitive services. It includes thousands of separately paid, individual codes, most of which describe highly specialized procedures and diagnostic tests, while there are relatively few codes that describe care management and cognitive services. The term “cognitive services” refers to the type of work that is usually classified and described under the current code set for E/M services, such as the critical thinking involved in data gathering and analysis, planning, management, decision-making, and exercising judgment in ambiguous or uncertain situations. It is often used to describe PFS services that are not procedural or strictly diagnostic in nature. Further, in the past, we have not recognized as separately payable many existing CPT codes that describe care management and cognitive services, viewing them as bundled and paid as part of other services including the broadly drawn E/M codes that describe face-to-face visits billed by physicians and practitioners in all specialties.

This has resulted in minimal service variation for ongoing primary care, care management and coordination, and cognitive services relative to other PFS services, and in potential misvaluation of E/M services under the PFS (76 FR 42793). Some stakeholders believe that there is substantial misvaluation of physician work within the PFS, and that the current service codes fail to capture the range and intensity of nonprocedural physician activities (E/M services) and the “cognitive” work of certain specialties (http://www.nejm.org/doi/full/10.1056/NEJMp1600999#t=article).

Recognizing the inverse for specialties that furnish other kinds of services, MedPAC has noted that the PFS allows some specialties to more easily increase the volume of services they provide, and therefore, their revenue from Medicare relative to other specialties, particularly those that spend most of their time providing E/M services. (MedPAC March 2015 Report to the Congress, available at http://www.medpac.gov/-documents-/reports). We agree with this analysis, and we recognize that the current set of E/M codes limits Medicare's ability under the PFS to appropriately recognize the relative resource costs of primary care, care management/coordination and cognitive services relative to specialized procedures and diagnostic tests.

In recent years, we have been engaged in an ongoing incremental effort to update and improve the relative value of primary care, care management/coordination, and cognitive services within the PFS by identifying gaps in appropriate payment and coding. These efforts include changes in payment and coding for a broad range of PFS services. This effort is particularly vital in the context of the forthcoming transition to the Quality Payment Program that includes the Merit-Based Incentive Payment System (MIPS) and Alternative Payment Models (APMs) incentives under the Medicare Access and CHIP Reauthorization Act of 2015 (MACRA) (Pub. L. 114-10, enacted April 16, 2015), since MIPS and many APMs will adopt and build on PFS coding, RVUs and PFS payment as their foundation.

In CY 2013, we began by focusing on post-discharge care management and transition of beneficiaries back into the community, establishing new codes to pay separately for transitional care management (TCM) services. Next we finalized new coding and separate payment beginning in CY 2015 for chronic care management (CCM) services provided by clinical staff. In the CY 2016 PFS proposed rule (80 FR 41708 through 41711), we solicited public comments on three additional policy areas of consideration: (1) Improving payment for the professional work of care management services through coding that would more accurately describe and value the work of primary care and other cognitive specialties for complex patients (for example, monthly timed services including care coordination, patient/caregiver education, medication management, assessment and integration of data, care planning); (2) establishing separate payment for collaborative care, particularly, how we might better value and pay for robust inter-professional consultation between primary care physicians and psychiatrists (developing codes to describe and provide payment for the evidence-based psychiatric collaborative care model (CoCM)), and between primary care physicians and other (non-mental health) specialists; and (3) assessing whether current PFS payment for CCM services is adequate and whether we should reduce the administrative burden associated with furnishing and billing these services.

We received substantial feedback on this comment solicitation, which we summarized in the CY 2017 PFS proposed rule and used to develop the following coding and payment proposals for CY 2017 (81 FR 46200 through 46215, and 46263 through 46265):

• Separate payment for existing codes describing prolonged E/M services without direct patient contact by the physician (or other billing practitioner), and increased payment for prolonged E/M services with direct patient contact by the physician (or other billing practitioner) adopting the RUC-recommended values.

• New coding and payment mechanisms for behavioral health integration (BHI) services including substance use disorder treatment, specifically three codes to describe services furnished as part of the psychiatric CoCM and one code to address other BHI care models.
• Separate payment for complex CCM services, reduced administrative burden for CCM, and an add-on code to the visit during which CCM is initiated (the CCM initiating visit) to reflect the work of the billing practitioner in assessing the beneficiary and establishing the CCM care plan.
• A new code for cognition and functional assessment and care planning, for treatment of cognitive impairment.
• An adjustment to payment for routine visits furnished to beneficiaries for whom the use of specialized mobility-assistive technology (such as adjustable height chairs or tables, patient lifts, and adjustable padded leg supports) is medically necessary.

We noted that the development of coding for these and other kinds of services across the PFS is typically an iterative process that responds to changes in medical practice and may be best refined over several years, with PFS rulemaking and the development of CPT codes as important parts of that process. We noted with interest that the CPT Editorial Panel and AMA/RUC restructured the former Chronic Care Coordination Workgroup to establish a new Emerging CPT and RUC Issues Workgroup that we hope will continue to consider the issues raised in this section of our CY 2017 proposed rule. At the time of publication of the proposed rule, we were aware that CPT had approved a code to describe assessment and care planning for treatment of cognitive impairment; however, it would not be ready in time for valuation in CY 2017. Therefore, we proposed to make payment using a G-code (G0505 ) for this service in CY 2017. We were also aware that CPT had approved three codes that describe services furnished consistent with the psychiatric CoCM, but that they would also not be ready in time for valuation in CY 2017. We discuss these services in more detail in the next section of this final rule.

To facilitate separate payment for these services furnished to Medicare beneficiaries during CY 2017, we proposed to make payment through the use of three G-codes (G0502, G0503, and G0504—see below) that parallel the new CPT codes, as well as a fourth G-code (G0507—see below) to describe services furnished using other models of BHI in the primary care setting. We intended for these to be temporary codes and would consider whether to adopt and establish values for the new CPT codes under our standard process, potentially for CY 2018. We anticipated continuing the multi-year process of implementing initiatives designed to improve payment for, and recognize long-term investment in, primary care, care management and cognitive services, and patient-centered services. While we recognized that there may be some overlap in the patient populations for the proposed new codes, we noted that time spent by a practitioner or clinical staff could not be counted more than once for any code (or assigned to more than one patient), consistent with PFS coding conventions. We expressed continued consideration of additional codes for CCM services that would describe the time of the physician or other billing practitioner. We also expressed interest in whether there should be changes under the PFS to reflect additional models of inter-professional collaboration for health conditions, in addition to those we proposed for BHI.

We proposed to pay under the PFS for services described by new coding as follows (please note that the descriptions included for G0502, G0503, and G0504 are from Current Procedural Terminology (CPT®) Copyright 2016 American Medical Association (and we understand from CPT that they will be effective as part of CPT codes January 1, 2018). All rights reserved):

• G0502: Initial psychiatric collaborative care management, first 70 minutes in the first calendar month of behavioral health care manager activities, in consultation with a psychiatric consultant, and directed by the treating physician or other qualified health care professional, with the following required elements:

++ Outreach to and engagement in treatment of a patient directed by the treating physician or other qualified health care professional;

++ Initial assessment of the patient, including administration of validated rating scales, with the development of an individualized treatment plan;

++ Review by the psychiatric consultant with modifications of the plan if recommended;

++ Entering patient in a registry and tracking patient follow-up and progress using the registry, with appropriate documentation, and participation in weekly caseload consultation with the psychiatric consultant; and

++ Provision of brief interventions using evidence-based techniques such as behavioral activation, motivational interviewing, and other focused treatment strategies.

• G0503: Subsequent psychiatric collaborative care management, first 60 minutes in a subsequent month of behavioral health care manager activities, in consultation with a psychiatric consultant, and directed by the treating physician or other qualified health care professional, with the following required elements:

++ Tracking patient follow-up and progress using the registry, with appropriate documentation;

++ Participation in weekly caseload consultation with the psychiatric consultant;

++ Ongoing collaboration with and coordination of the patient's mental health care with the treating physician or other qualified health care professional and any other treating mental health providers;

++ Additional review of progress and recommendations for changes in treatment, as indicated, including medications, based on recommendations provided by the psychiatric consultant;

++ Provision of brief interventions using evidence-based techniques such as behavioral activation, motivational interviewing, and other focused treatment strategies;

++ Monitoring of patient outcomes using validated rating scales; and relapse prevention planning with patients as they achieve remission of symptoms and/or other treatment goals and are prepared for discharge from active treatment.

• G0504: Initial or subsequent psychiatric collaborative care management, each additional 30 minutes in a calendar month of behavioral health care manager activities, in consultation with a psychiatric consultant, and directed by the treating physician or other qualified health care professional (List separately in addition to code for primary procedure) (Use G0504 in conjunction with G0502, G0503).
• G0507: Care management services for behavioral health conditions, at least 20 minutes of clinical staff time, directed by a physician or other qualified health care professional time, per calendar month.
• G0505: Cognition and functional assessment using standardized instruments with development of recorded care plan for the patient with cognitive impairment, history obtained from patient and/or caregiver, by the physician or other qualified health care professional in office or other outpatient setting or home or domiciliary or rest home.
• G0506: Comprehensive assessment of and care planning by the physician or other qualified health care professional for patients requiring chronic care management services, including assessment during the provision of a face-to-face service (billed separately from monthly care management services) (Add-on code, list separately in addition to primary service).
• G0501: Resource-intensive services for patients for whom the use of specialized mobility-assistive technology (such as adjustable height chairs or tables, patient lifts, and adjustable padded leg supports) is medically necessary and used during the provision of an office/outpatient evaluation and management visit (Add-on code, list separately in addition to primary procedure).

Regarding the majority of these proposals, the public comments were broadly supportive, some viewing our proposals as a temporary solution to an underlying need to revalue E/M services, especially outpatient E/M. Several commenters recommended that CMS utilize the global surgery data collection effort or another major research initiative to distinguish and revalue different kinds of E/M work. The commenters made recommendations about the scope and definition of the proposed services, what types of individuals should be able to provide them, and potential alignment and overlap. The commenters agreed with the need to increase the relative value of primary care, care management and other cognitive care under the PFS and minimize administrative burden for such services, while ensuring value to the program and beneficiaries. The public comments raise or inform a number of issues around how to define and pay for care that is collaborative, integrative or continuous, and we discuss the comments in greater detail below.

2. Non-Face-to-Face Prolonged Evaluation Management (E/M) Services

In public comments on the CY 2016 PFS proposed rule, many commenters recommended that CMS should establish separate payment for non-face-to-face prolonged E/M service codes that we currently consider to be “bundled” under the PFS (CPT codes 99358, 99359). The CPT descriptors are:

• CPT code 99358 (Prolonged evaluation and management service before and/or after direct patient care, first hour); and
• CPT code 99359 (Prolonged evaluation and management service before and/or after direct patient care, each additional 30 minutes (List separately in addition to code for prolonged service).

Commenters believed that separate payment for these existing CPT codes would provide a means for physicians and other billing practitioners to receive payment that more appropriately accounts for time that they spend providing non-face-to-face care. We agreed that these codes would provide a means to recognize the additional resource costs of physicians and other billing practitioners, when they spend an extraordinary amount of time outside of an E/M visit performing work that is related to that visit and does not involve direct patient contact (such as extensive medical record review, review of diagnostic test results or other ongoing care management work). We also believed that doing so in the context of the ongoing changes in health care practice to meet the current population's health care needs would be beneficial for Medicare beneficiaries and consistent with our overarching goals related to patient-centered care.

These non-face-to-face prolonged service codes are broadly described (although they include only time spent personally by the physician or other billing practitioner) and have a relatively high time threshold (the time counted must be an hour or more beyond the usual service time for the primary or “companion” E/M code that is also billed). They are not reported for time spent in care plan oversight services or other non-face-to-face services that have more specific codes and no upper time limit in the CPT code set. We believed this made these codes sufficiently distinct from the other codes we proposed for CY 2017 as part of our primary care/cognitive care/care management initiative described in this section of our final rule. Accordingly, we proposed to recognize CPT codes 99358 and 99359 for separate payment under the PFS beginning in CY 2017. We noted that time could not be counted more than once towards the provision of CPT codes 99358 or 99359 and any other PFS service. We addressed their valuation in the valuation section of the CY 2017 proposed rule.

Through a drafting error, we stated in the proposed rule that we would require these services to be furnished on the same day by the same physician or other billing practitioner as the companion E/M code. We intended to propose conformity with CPT guidance that requires that time counted towards the codes describe services furnished during a single day directly related to a discrete face-to-face service that may be provided on a different day, provided that the services are directly related to those furnished in a face-to-face visit.

We also solicited public comment on our interpretation of existing CPT guidance governing concurrent billing or overlap of CPT codes 99358 and 99359 with complex CCM services (CPT codes 99487 and 99489) and TCM services (CPT codes 99495 and 99496). Specifically CPT provides, “Do not report 99358, 99359 during the same month with 99487-99489. Do not report 99358, 99359 when performed during the service time of codes 99495 or 99496.” Complex CCM services and TCM services are similar to the non-face-to-face prolonged services in that they include substantial non-face-to-face work by the billing physician or other practitioner. The TCM and CCM codes similarly focus on a broader episode of patient care that extends beyond a single day, although they have a monthly service period and the prolonged service codes do not. We sought public input on the intersection of the non-face-to-face prolonged service codes with CCM and TCM services, and with the proposed add-on code to the CCM initiating visit G0506 (Comprehensive assessment of and care planning for patients requiring CCM services). We also solicited comment regarding how distinctions could be made between time associated with prolonged services and the time bundled into other E/M services, particularly pre- and post-service times, which would continue to be bundled with the other E/M service codes. For all of these services, we expressed concern that there would potentially be program integrity risks as the same or similar non-face-to-face activities could be undertaken to meet the billing requirements for a number of codes. We solicited public comment to help us identify the full extent of program integrity considerations, as well as options for mitigating program integrity risks.

Comment: Many commenters recommended that we adopt the CPT coding provision for CPT codes 99358 and 99359 that allows the prolonged services to be provided on a different day than the companion E/M code. At the same time, several commenters indicated that they request changes to the codes through the established processes of the CPT Editorial Panel. For example, some commenters suggested that CPT codes 99358 and 99359 should be revised so that they have a limited (calendar month) service period or measure shorter time increments (15 minutes). Some commenters recommended that a given physician should not be allowed to report CPT codes 99358 and 99359 for the same beneficiary during the same time he or she reported CCM, TCM, or G0506. These commenters stated that CCM, TCM, and proposed G0506 encompass non-face-to-face care provided to the beneficiary during a given period of time that would be duplicated if the physician is also allowed to report CPT codes 99358 and 99359 during the same time period. Other commenters stated that it would be unusual for G0506 and non-face-to-face prolonged services (CPT codes 99358 and 99359) to be reported for services on the same day, but that both should be allowed if time thresholds are met. To facilitate determination of whether time thresholds are met for various potential code combinations, some commenters recommended that CMS establish a time for G0506 and publish typical times for the companion codes to the prolonged service codes. This would enable practitioners to determine when they have exceeded “usual” or average times for E/M services and may bill prolonged services. Some commenters recommended that CMS provide tables showing times for E/M visits, CCM, G0506 and prolonged services with specific clinical examples for concurrent billing.

Some commenters believed there might be some overlap between the proposed non-face-to-face prolonged service codes and the post-service work of G0505 (Cognition and functional assessment by the physician or other qualified health care professional in office or other outpatient). Some commenters believed there is a discrepancy between our proposal to allow G0505 to be a companion code to prolonged services, and CPT's intent that G0505 should only be billed on the same day as another E/M visit if they are unrelated.

MedPAC commented that the companion E/M codes should be revalued instead of providing separate payment for prolonged services associated with the companion codes. However, if we finalize as proposed, MedPAC recommended that we clarify what situations the prolonged codes are appropriate for, beyond average times. Another commenter recommended an alternative policy instead of the non-face-to-face prolonged service codes, namely several modifiers and add-on codes to E/M services, associated with increased work RVUs. A typical time for the primary service would not need to be established. This coding schema would focus on visits actively treating patients with four or more chronic conditions; patients with three or more chronic problems introducing an acute problem during their visit; unexpected abnormal studies; and electronic communication after visits with the patient, lab, and other clinicians. One commenter drew a distinction between prolonged service work and care management services, where care management does not include extensive review of medical records, review of diagnostic tests and further discussion with a caregiver.

Response: We appreciate the comments. First, we had intended to propose to adopt the CPT coding provision for CPT codes 99358 and 99359 that allows the prolonged time to be provided on a different day than the companion E/M code, along with the rest of the CPT prefatory language for these codes. Our final policy will adopt the CPT guidance that allows the prolonged time to be reported for time on a different day than the companion E/M code, along with the rest of the CPT prefatory language for CPT codes 99358 and 99359.

Second, the public comments elucidate that it is difficult to assess potential overlap between prolonged services and many other codes because the included services, service periods and timeframes are not aligned. For example, most services paid under the PFS are valued based on assumptions regarding the typical pre-service, intra-service and post-service time, but do not have required thresholds for time spent. It is difficult to distinguish the times associated with these services from the times for codes that include time requirements in their descriptor. It is also difficult to distinguish the time and other work included in codes that generally describe services furnished during one day (prolonged services and E/M visits) with codes that describe time and work over substantially different service periods (such as the calendar month services like CCM or BHI services) or add-on codes with no pre or post-service time (such as G0506). In addition, because portions of many services are likely describing work that is furnished “incident to” a physician's or practitioner's services, the time and effort of the billing practitioner may not be the only relevant time and effort to consider. Moreover, the comments reflect a desire and intent on the part of stakeholders to alter the prolonged service codes in the near future, which would, in turn, alter their intersection with the codes proposed in this section of our 2017 rule and many other codes. The public comments also reflect a lack of consensus regarding appropriate medical practice and reporting patterns for prolonged services in relation to the services described by the CCM, TCM, proposed G0505 and proposed G0506 codes.

Having considered this feedback, we have decided to finalize our proposal for separate payment of the non-face-to-face prolonged service codes (CPT 99358, 99359) and adopt the CPT code descriptors and prefatory language for reporting these services. We stress that we intend these codes to be used to report extended non-face-to-face time that is spent by the billing physician or other practitioner (not clinical staff) that is not within the scope of practice of clinical staff, and that is not adequately identified or valued under existing codes or the 2017 finalized new codes. We appreciate the commenters' suggestion to display the typical times associated with relevant services. We have posted a file that notes the times assumed to be typical for purposes of PFS rate-setting. That file is available on our Web site under downloads for the CY 2017 PFS final rule at http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/PFS-Federal-Regulation-Notices.html. We note that while these typical times are not required to bill the displayed codes, we would expect that only time spent in excess of these times would be reported under a non-face-to-face prolonged service code.

Based on our analysis of comments, we do not believe there is significant overlap between CPT codes 99358 and 99359 and the CCM codes (CPT 99487, 99489, 99490) or our finalized BHI service codes (G0502, G0503, G0504, G0507 discussed below). The work of the billing practitioner in the provision of non-complex CCM and the BHI services is related to the direction of ongoing care management and coordination activities of other individuals, compared to the work of 99358 and 99359 which is described as personally performed and directly related to a face-to-face service. On that basis, we do not believe that there is significant overlap in the description of services or the valuation.

The potential intersection of CPT codes 99358 and 99359 with the complex CCM codes is harder to assess because complex CCM explicitly includes medical decision-making of moderate to high complexity by the billing practitioner, which is not performed by clinical staff. The complex CCM codes, however, only measure or count the time of clinical staff. Similarly, TCM includes moderate to high complexity medical decision-making during the service period as well as a level 4 or 5 face-to-face visit, even though clinical staff may perform a number of other aspects of the service. For CY 2017, for administrative simplicity, we are adopting the CPT provision (and finalizing as proposed) that complex CCM cannot be reported during the same month as non-face-to-face prolonged services, CPT codes 99358 and 99359 (by a single practitioner). Similarly, we are adopting the CPT provision that non-face-to-face prolonged services, CPT codes 99358 and 99359 may not be reported when performed during the service time of TCM (CPT codes 99495 and 99496) (by a single practitioner). We interpret the CPT provision to mean that CPT codes 99358 and 99359 cannot be reported during the TCM 30-day service period, by the same practitioner who is reporting the TCM.

Regarding potential intersection of CPT codes 99358 and 99359 with proposed G0505 (Cognition and functional assessment by the physician or other qualified health care professional in office or other outpatient), we are finalizing our proposal that G0505 be designated as a companion or “base” E/M code to non-face-to-face prolonged services (CPT codes 99358 and 99359) (see section II.E.5 for a detailed discussion of G0505). That is, for CY 2017 CPT codes 99358 and 99359 may be reported with G0505 as the associated companion code, whether furnished on the same day or a different day. We believe CPT intended the code on which G0505 is modeled to function like a specific E/M service, and that while the specificity of the service explicitly includes care planning unique to the needs of patients with particular conditions, there may well be circumstances where the pre- or post-time for a particular beneficiary may be prolonged. In their current form, the non-face-to-face prolonged service codes exist for the purpose of providing additional payment to account for the biller's additional time related to E/M visits. Therefore, we believe the non-face-to-face prolonged service codes should be reportable when related to E/M services, including those such as G0505 that describe more specific E/M work. We look forward to continued feedback on this issue, including through potential revisions to CPT guidance.

Regarding intersection of CPT codes 99358 and 99359 with G0506, we note that G0506 is already an add-on code to another E/M service (the CCM initiating visit, which can be the AWV/IPPE or a qualifying face-to-face E/M visit). We are providing in section II.E.4.a that at this time (beginning in CY 2017), G0506 will be a code that is only billable one time, at the outset of CCM services. We agree with commenters that it would be unusual for physicians to spend enough time with a given beneficiary on a given day to warrant reporting all three codes (the initiating visit code, G0506, and a prolonged service code). We also believe that a simpler approach is preferable at this time (two related codes for CCM initiation, instead of possibly three). Therefore our final policy for CY 2017 is that prolonged services (whether face-to-face or non-face-to-face) cannot be reported in addition to G0506 in association with a companion E/M code that also qualifies as the CCM initiating visit. In association with the CCM initiating visit, a billing practitioner may choose to report either prolonged services or G0506 (if requirements to bill both prolonged services and G0506 are met), but cannot report both a prolonged service code and G0506.

3. Establishing Separate Payment for Behavioral Health Integration (BHI)

In the CY 2016 PFS final rule with comment period (80 FR 70920), we stated that we believe the care and management for Medicare beneficiaries with behavioral health conditions often requires extensive discussion, information-sharing and planning between a primary care physician and a specialist. In CY 2016 rulemaking, we described that in recent years, many randomized controlled trials have established an evidence base for an approach to caring for patients with behavioral health conditions called the psychiatric Collaborative Care Model (CoCM). We sought information to assist us in considering refinements to coding and payment to address this model in particular. The psychiatric CoCM is one of many models for behavioral health integration or BHI, a term that refers broadly to collaborative care that integrates behavioral health services principally with primary care, but that may also integrate behavioral health care with inpatient and other clinical care. BHI is a team-based approach to care that focuses on integrative treatment of patients with medical and mental or behavioral health conditions. In the CY 2017 proposed rule (81 FR 46203 through 46205), we proposed four new G-codes for BHI services: Three describing the psychiatric CoCM specifically, and one generally describing related models of care.

a. Psychiatric Collaborative Care Model (CoCM)

A specific model for BHI, psychiatric CoCM typically is provided by a primary care team consisting of a primary care provider and a care manager who works in collaboration with a psychiatric consultant, such as a psychiatrist. Care is directed by the primary care team and includes structured care management with regular assessments of clinical status using validated tools and modification of treatment as appropriate. The psychiatric consultant provides regular consultations to the primary care team to review the clinical status and care of patients and to make recommendations. As we previously noted, several resources have been published that describe the psychiatric CoCM in greater detail and assess the impact of the model, including pieces from the University of Washington (http://aims.uw.edu/), the Institute for Clinical and Economic Review (http://icer-review.org/announcements/icer-report-presents-evidence-based-guidance-to-support-integration-of-behavioral-health-into-primary-care/), and the Cochrane Collaboration (http://www.cochrane.org/CD006525/DEPRESSN_collaborative-care-forpeople-with-depression-and-anxiety). Because this particular kind of collaborative care model has been tested and documented in medical literature, in the CY 2016 proposed rule we expressed particular interest in how coding used to describe PFS services might facilitate appropriate valuation of the services furnished under this model. We solicited public comments to assist us in considering refinements to coding and payment to address this model in particular relative to current coding and payment policies, as well as information related to various requirements and aspects of these services.

After consideration of the comments, we proposed in the CY 2017 PFS proposed rule to begin making separate payment for services furnished using the psychiatric CoCM, beginning January 1, 2017. We were aware that the CPT Editorial Panel, recognizing the need for new coding for services under this model of care, had approved three codes to describe the psychiatric collaborative care that is consistent with this model, but the codes would not be ready in time for valuation in CY 2017. Current CPT coding does not accurately describe or facilitate appropriate payment for the treatment of Medicare beneficiaries under this model of care. For example, under current Medicare payment policy, there is no payment made specifically for regular monitoring of patients using validated clinical rating scales or for regular psychiatric caseload review and consultation that does not involve face-to-face contact with the patient. We believed that these resources are directly involved in furnishing ongoing care management services to specific patients with specific needs, but they are not appropriately recognized under current coding and payment mechanisms. Because PFS valuation is based on the relative resource costs of the PFS services furnished to Medicare beneficiaries, we believed that appropriate coding for these services for CY 2017 will facilitate accurate payment for these and other PFS services. Therefore, we proposed separate payment for services under the psychiatric CoCM using three new G-codes, as detailed below: G0502, G0503, and G0504, which would parallel the CPT codes that are being created to report these services.

The proposed code descriptors were as follows (from Current Procedural Terminology (CPT®) Copyright 2016 American Medical Association (and we understand from CPT that they will be effective as part of CPT codes January 1, 2018). All rights reserved):

• G0502: Initial psychiatric collaborative care management, first 70 minutes in the first calendar month of behavioral health care manager activities, in consultation with a psychiatric consultant, and directed by the treating physician or other qualified health care professional, with the following required elements:

++ Outreach to and engagement in treatment of a patient directed by the treating physician or other qualified health care professional;

++ Initial assessment of the patient, including administration of validated rating scales, with the development of an individualized treatment plan;

++ Review by the psychiatric consultant with modifications of the plan if recommended;

++ Entering patient in a registry and tracking patient follow-up and progress using the registry, with appropriate documentation, and participation in weekly caseload consultation with the psychiatric consultant; and

++ Provision of brief interventions using evidence-based techniques such as behavioral activation, motivational interviewing, and other focused treatment strategies.

• G0503: Subsequent psychiatric collaborative care management, first 60 minutes in a subsequent month of behavioral health care manager activities, in consultation with a psychiatric consultant, and directed by the treating physician or other qualified health care professional, with the following required elements:

++ Tracking patient follow-up and progress using the registry, with appropriate documentation;

++ Participation in weekly caseload consultation with the psychiatric consultant;

++ Ongoing collaboration with and coordination of the patient's mental health care with the treating physician or other qualified health care professional and any other treating mental health providers;

++ Additional review of progress and recommendations for changes in treatment, as indicated, including medications, based on recommendations provided by the psychiatric consultant;

++ Provision of brief interventions using evidence-based techniques such as behavioral activation, motivational interviewing, and other focused treatment strategies;

++ Monitoring of patient outcomes using validated rating scales; and relapse prevention planning with patients as they achieve remission of symptoms and/or other treatment goals and are prepared for discharge from active treatment.

• G0504: Initial or subsequent psychiatric collaborative care management, each additional 30 minutes in a calendar month of behavioral health care manager activities, in consultation with a psychiatric consultant, and directed by the treating physician or other qualified health care professional (List separately in addition to code for primary procedure) (Use G0504 in conjunction with G0502, G0503).

We stated that we intend these to be temporary codes and would consider whether to adopt and establish values for the associated new CPT codes under our standard process once those codes are active.

We proposed that these services would be furnished under the direction of a treating physician or other qualified health care professional during a calendar month. These services would be furnished when a patient has a diagnosed psychiatric disorder that requires a behavioral health care assessment; establishing, implementing, revising, or monitoring a care plan; and provision of brief interventions. The diagnosis could be either pre-existing or made by the billing practitioner. These services would be reported by the treating physician or other qualified health care professional and include the services of the treating physician or other qualified health care professional, the behavioral health care manager (see description below) who would furnish services incident to services of the treating physician or other qualified health care professional, and the psychiatric consultant (see description below) whose consultative services would be furnished incident to services of the treating physician or other qualified health care professional. We proposed that beneficiaries who are appropriate candidates for care reported using the psychiatric CoCM codes could have newly diagnosed conditions, need help in engaging in treatment, have not responded to standard care delivered in a non-psychiatric setting, or require further assessment and engagement prior to consideration of referral to a psychiatric care setting. Beneficiaries would be treated for an episode of care, defined as beginning when the behavioral health care manager engages in care of the beneficiary under the appropriate supervision of the billing practitioner and ending with:

• The attainment of targeted treatment goals, which typically results in the discontinuation of care management services and continuation of usual follow-up with the treating physician or other qualified healthcare professional; or
• Failure to attain targeted treatment goals culminating in referral to a psychiatric care provider for ongoing treatment; or
• Lack of continued engagement with no psychiatric collaborative care management services provided over a consecutive 6-month calendar period (break in episode).

A new episode of care would start after a break in episode of 6 calendar months or more.

The treating physician or other qualified health care professional would direct the behavioral health care manager and continue to oversee the beneficiary's care, including prescribing medications, providing treatments for medical conditions, and making referrals to specialty care when needed. Medically necessary E/M and other services could be reported separately by the treating physician or other qualified health care professional, or other physicians or practitioners, during the same calendar month. Time spent by the treating physician or other qualified health care professional on activities for services reported separately could not be included in the services reported using G0502, G0503, and G0504. We proposed that the behavioral health care manager would be a member of the treating physician or other qualified health care professional's clinical staff with formal education or specialized training in behavioral health (which could include a range of disciplines, for example, social work, nursing, and psychology) who provides care management services, as well as an assessment of needs, including the administration of validated rating scales, the development of a care plan, provision of brief interventions, ongoing collaboration with the treating physician or other qualified health care professional, maintenance of a registry, all in consultation with a psychiatric consultant. The behavioral health care manager would furnish these services both face-to-face and non-face-to-face, and consult with the psychiatric consultant minimally on a weekly basis. We proposed that the behavioral health care manager would be on-site at the location where the treating physician or other qualified health care professional furnishes services to the beneficiary.

We proposed that the behavioral health care manager may or may not be a professional who meets all the requirements to independently furnish and report services to Medicare. If otherwise eligible, then that individual could report separate services furnished to a beneficiary receiving the services described by G0502, G0503, G0504, and G0507 in the same calendar month. These could include: Psychiatric evaluation (90791, 90792), psychotherapy (90832, 90833, 90834, 90836, 90837, 90838), psychotherapy for crisis (90839, 90840), family psychotherapy (90846, 90847), multiple family group psychotherapy (90849), group psychotherapy (90853), smoking and tobacco use cessation counseling (99406, 90407), and alcohol or substance abuse intervention services (G0396, G0397). Time spent by the behavioral health care manager on activities for services reported separately could not be included in the services reported using time applied to G0502, G0503, and G0504.

The psychiatric consultant involved in the “incident to” care furnished under this model would be a medical professional trained in psychiatry and qualified to prescribe the full range of medications. The psychiatric consultant would advise and make recommendations, as needed, for psychiatric and other medical care, including psychiatric and other medical diagnoses, treatment strategies including appropriate therapies, medication management, medical management of complications associated with treatment of psychiatric disorders, and referral for specialty services, that are communicated to the treating physician or other qualified health care professional, typically through the behavioral health care manager. The psychiatric consultant would not typically see the patient or prescribe medications, except in rare circumstances, but could and should facilitate a referral to a psychiatric care provider when clinically indicated.

In the event that the psychiatric consultant furnished services to the beneficiary directly in the calendar month described by other codes, such as E/M services or psychiatric evaluation (CPT codes 90791 and 90792), those services could be reported separately by the psychiatric consultant. Time spent by the psychiatric consultant on activities for services reported separately could not be included in the services reported using G0502, G0503, and G0504.

We also noted that, although the psychiatric CoCM has been studied extensively in the setting of specific behavioral health conditions (for example, depression), we received persuasive comments in response to the CY 2016 proposed rule recommending that we not specify particular diagnoses required for use of the codes for several reasons, including that: There may be overlap in behavioral health conditions; there are concerns that there could be modification of diagnoses to fit within payment rules which could skew the accuracy of submitted diagnosis code data; and for many patients for whom specialty care is not available, or who choose for other reasons to remain in primary care, primary care treatment will be more effective if it is provided within a model of integrated care that includes care management and psychiatric consultation.

Comment: The public comments were very supportive of our creation of the three G-codes for CY 2017 to pay for services furnished using the psychiatric CoCM. The commenters offered a number of recommendations regarding valuation of the codes. Some commenters requested additional codes, sought clarification, or presented statements in favor of including the services of practitioners other than psychiatrists, especially psychologists and social workers, within the proposed codes.

Response: We thank the commenters for their support of coding and valuation for services furnished using the psychiatric CoCM, and for their recommendations regarding appropriate valuation. We address the comments on valuation in section II.L of this final rule. We address the comments regarding payment for services of psychologists and social workers below.

Comment: Several commenters expressed concern that making separate payment for psychiatric CoCM for the treatment of mood disorders might result in neglecting treatment for other mental health conditions. Other commenters expressed support for not designating a limited set of eligible behavioral health diagnoses. One commenter stated that requiring a diagnosed behavioral health condition might mean that subclinical issues or undiagnosed behavioral health conditions would be neglected.

Response: We continue to believe that we should not limit billing and payment for the psychiatric CoCM codes to a limited set of behavioral health conditions. As we understand it, the psychiatric CoCM model of care may be used to treat patients with any behavioral health condition that is being treated by the billing practitioner, including substance use disorders. In the Collaborative Care literature reviewed by the Cochrane Collaboration and others, there is stronger evidence of effectiveness and cost-effectiveness for certain behavioral disorders, particularly mood and anxiety disorders, than for others. However, we continue to receive persuasive comments indicating that the psychiatric CoCM is recommended for broader incorporation into clinical practice, and recommending that we not specify the use of the psychiatric CoCM codes for only particular behavioral health diagnoses. Therefore we are not limiting billing and payment for the psychiatric CoCM codes to a specified set of behavioral health conditions.

In response to the public comment regarding whether we should require a diagnosed psychiatric disorder (as opposed to a subclinical or undiagnosed condition), we are clarifying that as described, the services require that there must be a presenting psychiatric or behavioral health condition(s) that, in the clinical judgment of the treating physician or other qualified health professional, warrants “referral” to the behavioral health care manager for further assessment and treatment through provision of psychiatric CoCM services. “Referral” is placed in quotes because the behavioral health care manager may be located in the same practice as the treating physician or other qualified health professional, who in any event provides ongoing oversight and continues to treat the beneficiary. However, the referring diagnosis (or diagnoses) may be either pre-existing or made by the treating physician or other qualified health professional, and we are not establishing any specific list of eligible or included diagnoses or conditions. The treating physician or other qualified health professional may not be qualified or able to fully diagnose all relevant psychiatric or behavioral health condition(s) prior to referring the beneficiary for psychiatric CoCM services. If in the course of providing psychiatric CoCM services, it becomes clear that the referring condition(s) or other diagnoses cannot be addressed by psychiatric CoCM services, then we understand that the beneficiary should be referred for other psychiatric treatment or should continue usual follow-up care with the treating practitioner, because the episode of psychiatric CoCM services ends if there is failure to attain targeted treatment goals after or despite changes in treatment, as indicated. Beneficiaries receiving care reported using the psychiatric CoCM codes may, but are not required to have comorbid chronic or other medical condition(s) that are being managed by the treating practitioner.

Comment: Several commenters who supported payment for the proposed codes for psychiatric CoCM services in primary care settings, raised questions about whether these codes could be used to bill for services furnished in other settings that are not traditional primary care settings, such as inpatient or long-term care, oncology practices, or emergency departments. Some of these commenters recommended additional new codes to pay for services furnished in these other settings.

Response: The psychiatric CoCM trials and real world implementation have mainly included primary care practice that broadly includes pediatrics, obstetrics/gynecology, and geriatrics as well as family practice and general internal medicine. The psychiatric CoCM has also been used in cardiology and oncology practice, and we believe it could be used in various medical specialty settings, as long as the specialist physician or practitioner is managing the beneficiary's behavioral health condition(s) as well as other medical conditions (for example, cancer, status-post acute myocardial infarction and other conditions where co-morbid depression is common). Accordingly, we are not limiting the code to reporting by only “traditional” primary care specialties. We believe primary care practitioners will most frequently perform the services described by the new psychiatric CoCM codes, but if other specialist practitioners perform these services and meet all of the requirements to bill the code(s), then they may report the psychiatric CoCM codes. We are interested in receiving additional, more specific information from stakeholders regarding which specialties furnish psychiatric CoCM services. We note that we would generally not expect psychiatrists to bill the psychiatric CoCM codes, because psychiatric work is defined as a sub-component of the psychiatric CoCM codes.

Regarding psychiatric CoCM services furnished to inpatients or beneficiaries in long-term care settings such as nursing or custodial care facilities, we note that the forthcoming CPT codes are not limited to office or other outpatient or domiciliary services. Moreover, our goal is to separately identify and pay for psychiatric CoCM services furnished to beneficiaries in any appropriate setting of care, whether inpatient or outpatient, in recognition of the associated time and service complexity. Care of beneficiaries who are admitted to a facility, are in long-term care, or are transitioning among settings during the month may be more complex than the care of other types of patients. While there is some overlap between psychiatric CoCM and CCM services, they are distinct services with differing patient populations, as discussed elsewhere in this section of our final rule. Therefore, we have valued the psychiatric CoCM services in both facility and non-facility settings (see section II.L on valuation). We are not limiting the time that can be counted towards the monthly time requirement to bill the psychiatric CoCM code(s) to time that is spent in the care of an outpatient or a beneficiary residing in the community. However, we also stress that G0502, G0503 and G0504 can only be reported by a treating physician or other qualified health care professional when he or she has directed the psychiatric CoCM service for the duration of time that he or she is reporting it, and has a qualifying relationship with individuals providing the service under his or her direction and control. Also, time and effort that is spent managing care transitions for CCM or TCM patients and that is counted towards reporting TCM or CCM services, cannot also be counted towards reporting any transitional care management activities reported under a BHI service code(s), either the psychiatric CoCM codes or the code describing other BHI services. We welcome additional input from stakeholders regarding appropriate (or inappropriate) sites of service for G0502, G0503 and G0504.

We note that for CY 2017, the facility PE RVU for psychiatric CoCM services will include the indirect PE allocated based on the work RVUs, but no direct PE (which is explicitly comprised of other labor, equipment and supplies). This is because historically, the PFS facility rate for a given professional service assumes that the billing practitioner is not bearing a significant resource cost in labor by other individuals, equipment or supplies. We generally assume that those costs are instead borne by the facility, and are adequately accounted for in a separate payment made to the facility to account for these costs and other costs incurred by the facility for the beneficiary's facility stay. For BHI services and similar care management services such as CCM, we have been considering whether this approach to PFS valuation is optimal because the PFS service, in significant part, may be provided by the behavioral health care manager, clinical staff, or even other physicians under the employment of the billing practitioner or under contract to the billing practitioner. These individuals may provide much of the PFS service remotely, and are not necessarily employees or staff of the facility. Indeed, the BHI services are defined in terms of activities performed by individual(s) other than the billing practitioner and who may not be affiliated with or located within the facility, even though as we discuss below the billing practitioner must also perform certain work. For this type of PFS service, there may be more direct practice expense borne by the billing practitioner even though the beneficiary is located, for part or all of the month, in a facility receiving institutional payment. We plan to consider these issues further in the future.

Comment: One specialty association supported the proposed psychiatric CoCM codes, noting that although few of their members would use these codes, they set an important precedent to recognize interdisciplinary care that requires significant non-face-to-face work. This commenter anticipated that similar code series may be developed in the future to describe complex management in other specialties including neurology, and supported the adoption of language approved at CPT that carefully defined the roles of multiple professionals. Other commenters similarly expressed support for separate payment for additional collaborative care services, including inter-professional consultation in the treatment of other illnesses such as cancer or multiple sclerosis.

Response: We continue to be interested in new coding that describes integrative, collaborative or consultative care among specialties other than primary care and behavioral health/psychiatry. We are especially interested in new coding that describes such care in sufficient detail that distinguishes it from existing service codes, and that would further the appropriate valuation of cognitive services. We will continue to follow any new coding proposals at CPT relevant for the Medicare population. We note that we have followed CPT's lead in finalizing proposed code G0505 for cognitive impairment assessment and care planning (see section II.E.5) as well as for psychiatric CoCM services. BHI is a unique type of service that we believe until now has not been well identified nor appropriately valued under existing codes. BHI is not comprised of mere consultation among professionals and has a unique evidence base, in addition to being recently addressed by forthcoming CPT coding. In addition, given the shortage of available psychiatric and other mental health professionals in many parts of the country, we believe it is important to identify and make accurate payment for models of care that facilitate access to psychiatric and other behavioral health specialty care through innovations in medical practice, like the ones described by these codes.

Comment: One commenter asked CMS to clarify inclusion of nurse practitioners who are primary care practitioners and, in the specialty of psychiatry, psychiatric nurse practitioners who can perform psychiatric evaluations and treat psychiatric problems.

Response: Nurse practitioners are authorized to independently bill Medicare for their services, and can also bill Medicare for services furnished incident to their services. Therefore, nurse practitioners who furnish the psychiatric CoCM services as described may bill for the psychiatric CoCM codes. Nurse practitioners who meet our final qualifications to serve as the behavioral health care manager may provide the behavioral health care manager services incident to the services of another (billing) practitioner. Nurse practitioners who meet all of our final requirements to serve as the psychiatric consultant may provide the psychiatric consultant services incident to the services of the billing practitioner.

Comment: Regarding the care planning requirements for psychiatric CoCM services, some commenters noted that there is not necessarily value in accumulating or enumerating a number of different types of care plans addressing different aspects of the beneficiary's problems, such as a behavioral or psychiatric care plan, a CCM care plan, and a cognitive impairment care plan (see G0505 in section II.E.5).

Response: While the proposed descriptors for the psychiatric CoCM services referred to an “individualized treatment plan,” not a “care plan,” we proposed in addition that the behavioral health care manager would “develop a care plan.” While any care planning should take into account the whole patient, our intent is that the care planning included in the CCM coding (and G0506, the CCM initiating visit add-on code) will be the most comprehensive in nature, addressing all health issues with particular focus on the multiple chronic conditions being managed by the billing practitioner. In that sense, the CCM care plan is an integrative care plan incorporating more comprehensive health information on all of the beneficiary's health issues, or reconciling care plans of other practitioners. In contrast, the BHI care planning will focus on behavioral health or psychiatric issues, in particular, just as cognitive impairment care planning will focus on cognitive impairment issues, in particular (see section II.E.5). We are not requiring the psychiatric CoCM treating practitioner or behavioral health consultant to perform care planning that incorporates comprehensive health information on all of the beneficiary's health issues or reconciles the care plans of other practitioners, as would be expected for CCM care planning.

We understand that adoption of EHRs may be lower among behavioral health practitioners and note that resources are available to help inform how care plans can support team-based care and BHI. Our understanding from the public comments last year and subsequent discussions with experts on the psychiatric CoCM model of care, is that no specific electronic technology or format is necessary or indispensable to carry out the psychiatric CoCM model of care, or perform the services included in the codes we are creating to describe the services furnished using that model. We believe the format of the behavioral health care plan (or any care plan) is less important than having a process whereby feedback and expertise from all relevant practitioners and providers, whether internal or external to the billing practice, are integrated into the beneficiary's treatment plan and goals; that this plan be regularly assessed and revisited by the practitioner who is assuming an overall care management role for the beneficiary in a given month; that the patient is engaged in the care planning process; and that the care planning be documented in the medical record (as with any required element of any PFS service). We are revising the requirement for care planning by the behavioral health care manager accordingly, that he or she will perform “behavioral health care planning in relation to behavioral/psychiatric health problems, including revision for patients who are not progressing or whose status changes.”

Comment: A number of commenters recommended that we should not require the behavioral health care manager for the psychiatric CoCM services to be located on site within the primary care practice. The commenters noted that in some settings, particularly rural areas or smaller practices, this may be especially important. Some commenters assumed that there is also a behavioral health care manager for G0507 (discussed below). These commenters compared BHI services (the psychiatric CoCM services and G0507) to CCM and recommended that CMS adopt the same requirements for all the BHI codes as for CCM, regarding supervision, location of a behavioral health care manager, and third party outsourcing.

Response: For the psychiatric CoCM services, we proposed that the behavioral health care manager would be a member of the treating physician or other qualified health care professional's clinical staff, and would be required to be located on site but able to work under general supervision. In addition, we proposed that the behavioral health care manager provides his or her services both face-to-face and non-face-to-face. We believed that services provided using the psychiatric CoCM model of care commonly involve face-to-face interaction between the behavioral health care manager and the beneficiary on appropriate occasions, such as the outset of services (a “warm hand-off” from the treating physician or other qualified health care professional). In addition, whether face-to-face or non-face-to-face, many of the included behavioral health care manager duties could be performed while the treating practitioner is not in the office and could be performed after hours. We note that the behavioral health care manager duties are listed in full above, and include care management services, as well as an assessment of needs, including the administration of validated rating scales, behavioral health care planning, provision of brief interventions, ongoing collaboration with the treating physician or other qualified health care professional, and maintenance of a registry, all in consultation with a psychiatric consultant.

The delivery of the psychiatric CoCM depends, in part, on continuity of care between a given patient and the assigned behavioral health care manager. Also it requires collaboration, integration and ongoing data flow between the behavioral health care manager and the treating practitioner the behavioral health care manager is supporting, as well as with the psychiatric consultant who is usually remotely located under the psychiatric CoCM model of care. As previously discussed, the psychiatric CoCM is an integrative model of care, and in considering our proposal we were concerned that allowing the behavioral health care manager to be located remotely would compromise their ability to collaborate, communicate, and timely treat and share information with the beneficiary and the rest of the care team. We are aware of many care management companies and health information technology companies that may seek to provide remote care management and related services under all of the new BHI codes, as they have for CCM and similar services recently adopted under the PFS. We received public comments from several such stakeholders that indicated an interest in the provision of BHI services and related health information technology. We understand that there have been successful implementations (positive randomized controlled trials) of the psychiatric CoCM using remote call centers; however, in these implementations, call center staff were not randomly rotated among patients and there was ongoing data flow and connectivity between the behavioral health care manager and the other members of the care team, as well as the patient. Moreover, the behavioral health care manager would presumably have to be on site at least some of the time (even if under general supervision), in order to provide some of their services in-person with the beneficiary.

The fact that we proposed and are finalizing general supervision for the psychiatric CoCM codes as we did for CCM services (see section II.E.3.b) does not mean that general supervision alone suffices to meet the requirements of the psychiatric CoCM for continuity, collaboration and integration among the care team members, including the beneficiary. General supervision means that the service is furnished under the overall direction and control of the practitioner billing the service, but without the presence of the practitioner being required during the performance of the service. This definition does not directly govern where individual(s) providing the service on an incident to basis are located, whether on site or remote. Rather, it governs the location and informs the involvement of the billing practitioner.

For payment purposes, we are assigning general supervision to the psychiatric CoCM codes because we do not believe it is clinically necessary that the professionals on the team who provide services other than the treating practitioner (namely, the behavioral health care manager and the psychiatric consultant) must have the billing practitioner immediately available to them at all times, as would be required under a higher level of supervision. However, general supervision sets the minimum standard for supervision and does not, by itself, meet the requirements we are setting for billing new codes G0502, G0503 and G0504. While certain aspects of psychiatric CoCM services might be furnished under general supervision, we do not believe the general supervision requirement adequately describes the nature of the relationship and interactions of the respective team members for services furnished using the psychiatric CoCM or the codes we are creating to describe those services. Moreover it only directly addresses the physical location of the billing practitioner, not the behavioral health care manager, necessarily.

After considering the public comments, we are not finalizing our proposal that the behavioral health care manager must be a member of the treating physician or other qualified health care professional's clinical staff. As some of the psychiatric CoCM services can be contracted out to a third party (subject to rules discussed below), the contracted individuals are not necessarily employees of the treating practitioner.

Regarding the face-to-face provision of services by the behavioral health care manager, we are requiring that the behavioral health care manager must be available to provide services on a face-to-face basis, but not that face-to-face services must be provided. We are not finalizing the proposed requirement that the behavioral health care manager must be located on site, in order to allow for after-hours or appropriate remote provision of services. However, to ensure clinical integration with the treating practitioner and familiarity and continuity with the beneficiary, which are characteristic of services furnished under the psychiatric CoCM model of care, we are requiring that the behavioral health care manager must have a collaborative, integrated relationship with the rest of the care team members, and be able to perform all of the required elements of the psychiatric CoCM services delineated for the behavioral health care manager. The behavioral health care manager must have the ability to engage the beneficiary outside of regular clinic hours as necessary to perform their duties under the CoCM model, and have a continuous relationship with the beneficiary. This does not mean the behavioral health care manager is necessarily an employee of or always physically located within the practice, nor does it require provision of behavioral health care manager services to the beneficiary on site. The behavioral health care manager may provide his or her services from a remote location that is remote from the billing practitioner or remote from the beneficiary, subject to incident to rules and regulations in 42 CFR 410.26, if he or she has a qualifying relationship with the rest of the care team including the beneficiary, and is available to provide services face-to-face.

We will monitor this issue going forward, not just for the psychiatric CoCM but also for the general BHI service code (G0507) we are finalizing, as well as for TCM and CCM services. As we discuss in the final rule section on CCM below, we are continuing to consider whether outsourcing certain aspects of these services to a third party fragments care, leads to insufficient involvement and oversight of the billing practitioner or results in services that do not actually represent or facilitate continuous, seamless transitional care and other required aspects of these services. We will continue to consider how to best define the continuity of care that is required for services furnished and billed under all of these codes, and whether arrangements for remote provision of services whether by a case management company or another entity increases rather than reduces service fragmentation. Advances in health information technology provide opportunities for remote connectivity and interoperability that may assist and be useful, if not necessary, for reducing care fragmentation. However, remote provision of services by entities having only a loose association with the treating practitioner can detract from continuous, patient-centered care, whether or not those entities employ certified or other electronic technology.

We note that while time spent by the treating practitioner is not explicitly counted for in codes G0502, G0503 and G0504, these codes are valued to include work performed directly by the treating practitioner. The treating practitioner directs the behavioral health care manager and continues to oversee the patient's care, including prescribing medications, providing treatments for medical conditions, and making referrals to specialty care when needed. We are finalizing as proposed that some of these services may be separately billable. However, we wish to emphasize that the treating practitioner must remain involved in ongoing oversight, management, collaboration and reassessment as appropriate to bill the psychiatric CoCM codes.

Comment: We received a number of comments requesting that we allow or recognize pharmacists, especially neurologic or psychiatric pharmacists, or doctoral-level clinical psychologists to serve as the psychiatric consultant. Some commenters were concerned that CMS is advocating pharmacotherapy over psychotherapy by requiring a psychiatric consultant who can prescribe medication.

Response: We agree with the commenters that there are multiple types of indicated treatment for behavioral health conditions, including psychotherapy and other psychosocial interventions as well as pharmacotherapy that are available and should be offered to beneficiaries receiving psychiatric CoCM services. Our intent is not to inappropriately steer beneficiaries into medication-based treatment, but rather that the psychiatric consultant be able to present and recommend the full range of treatment options including but not limited to medications, and to advise regarding any medications the beneficiary chooses to take. Under the psychiatric CoCM, the psychiatric consultant must be able to prescribe medication. As we discuss in section II.L on valuation of G0502, G0503 and G0504, we agree with the commenters who stated that the role of the psychiatric consultant under these codes is primarily evaluation and management, which is not within the scope of pharmacists or clinical psychologists under Medicare rules. Therefore, we are finalizing the role and qualifications of the psychiatric consultant as proposed. The general BHI code (G0507), which we are finalizing, was intended and may be used to report other models of care, where the beneficiary may not receive E/M services from the consultant and the consultant may only be authorized to provide psychotherapy or consultation regarding medications (see section II.E.3.b).

Comment: We received a number of comments recommending various types of professionals as qualified to serve as the behavioral health care manager, such as licensed clinical social workers (LCSWs) and psychologists.

Response: Unlike CCM and the general BHI service (code G0507), the psychiatric CoCM codes are used to report time that is spent in specified activities performed by a behavioral health care manager having formal education or specialized training in those activities, whether or not the behavioral health care manager is eligible to directly bill Medicare for other services. The behavioral health care manager may or may not be a professional who meets all the requirements to independently furnish and report services to Medicare. The behavioral health care manager must also meet any applicable licensure and state law requirements, which is required under 42 CFR 410.26 for all services provided under the PFS. LCSWs would meet these requirements, as would qualified registered nurses, clinical psychologists and other qualified clinical staff. Time spent by administrative or clerical staff cannot be counted towards the time required to bill G0502, G0503 or G0504.

Evaluation and management services (such as face-to-face E/M visits) may be separately billed during the service period or on the same day as the psychiatric CoCM services, provided time is not counted twice towards the same code.

b. General Behavioral Health Integration (BHI)

We recognize that the psychiatric CoCM is prescriptive and that much of its demonstrated success may be attributable to adherence to a set of elements and guidelines of care. We are finalizing the code set discussed above to pay accurately for care furnished using this specific model of care, given its widespread adoption and recognized effectiveness. However, we note that PFS coding, in general, does not dictate how physicians practice medicine and believe that it should, instead, reflect the practice of medicine. We also recognize that there are primary care practices that are incurring, or may incur, resource costs inherent to treatment of patients with similar conditions based on BHI models of care other than the psychiatric CoCM that may benefit beneficiaries with behavioral health conditions (see, for example, the approaches described at http://www.integration.samhsa.gov/integrated-care-models). There are a variety of care models ranging from behavioral health professionals embedded within a primary care office for same-day treatment, to remote consultation, to assessment-and-referral (see, for example, http://www.commonwealthfund.org/publications/newsletters/quality-matters/2014/august-september/profiles; and http://www.integration.samhsa.gov/integrated-care-models). These models of care have tended to arise from clinical practice as opposed to the research environment (http://psychnews.psychiatryonline.org/doi/full/10.1176/appi.pn.2014.10b25), and include resource costs that differ in various respects from those associated with the psychiatric CoCM.

To recognize the resource costs associated with furnishing such BHI services to Medicare beneficiaries, we also proposed to make payment using a new G-code that describes care management for beneficiaries with behavioral health conditions under other models of care. We believe that the resources associated with such care are not currently adequately recognized under the PFS. The proposed code was G0507 (Care management services for behavioral health conditions, at least 20 minutes of clinical staff time, directed by a physician or other qualified health care professional, per calendar month). We noted that we would expect this code to be refined over time as we receive more information about other BHI models being used and how they are implemented.

We sought stakeholder input on whether we should consider different increments of time for this code, such as a base code plus an add-on code comprised of additional 20 minute increments. We recognized that BHI services furnished under the proposed code may range in resource costs. We believed that appropriate payment for these services would further the refinement and implementation of BHI models of care, and that having utilization data would inform future refinement of the proposed code's valuation.

Comment: The commenters were supportive of new coding to support payment for other BHI models of care. They believed G0507 could be used by some smaller or medium sized practices who could not conform to the strict parameters of the psychiatric CoCM but provide very similar services. They also stated that G0507 would be appropriate to report services furnished under other BHI models of care that may not require psychiatric services. We received a few comments describing particular models of care in great detail; a few commenters referenced the Veterans' Administration BHI care models, the Primary Care Behavioral Health/Behavioral Health Consultation (PCBH/BHC) Model, or general models in place within other health care systems. However, there was consensus among the commenters that another code(s) in addition to the psychiatric CoCM codes would be useful to collect information on how other behavioral health care models are being used and implemented.

Many commenters recommended that CMS provide more of a framework or description of included services and provider types without being unduly burdensome. Some commenters recommended service elements similar to the CCM service elements (continuity of care with a designated member of the care team; a written care plan; a comprehensive assessment of behavioral health or psychiatric and other medical conditions as well as any functional and psychosocial needs, updated as necessary; routine evaluation of patient progress using a tracking system; services should be documented in the medical record and available to other treating professionals). These commenters recommended that eligible patients should have a diagnosed psychiatric or substance use disorder that requires care management services. Several commenters recommended that BHI payments be tied to the use of behavioral health assessment tools for screening and collection of treatment outcomes throughout the sessions of care in primary care. These commenters believed this would better position behavioral health to benefit from the movement toward value-based payment in the future. Some commenters assumed there is a designated behavioral health care manager for the service described by G0507, and recommended that we adopt similar rules for this care manager as apply for clinical staff providing CCM services.

Response: We continue to believe that another code, or set of BHI codes, in addition to the psychiatric CoCM code set would be useful to pay appropriately for BHI services furnished to Medicare beneficiaries. We also believe that such payment could facilitate our ability to identify and collect data regarding similar or related BHI service models. We agree with the commenters that we should provide more specificity around the services eligible for reporting under this other code(s). One way to do this would be to create codes with tiered times. Some commenters supported such an approach, while others believed it would be premature. At this time, we are not creating multiple levels of codes distinguishing levels of general BHI services using time or any other metric, but we may reconsider this in the future (also see section II.L on G0507 valuation).

Regarding included elements of the general BHI service (G0507), we agree with the commenters that we should be more specific in our definition of this service. We wish to provide greater specificity without being overly prescriptive, since a range of activities may be included in BHI models of care other than the psychiatric CoCM. We believe we should include a core set of service elements that are similar to core elements of the psychiatric CoCM, especially a systematic process for initial assessment and routine follow up evaluation, revising the treatment approach or methods for patients who are not progressing or whose status changes; facilitating and coordinating behavioral health expertise and treatment; and designating a member of the care team with whom the beneficiary has a continuous relationship. We may revisit the included services in future years, but for CY 2017 the required service elements for the general BHI service (G0507) will be:

• Initial assessment or follow-up monitoring, including the use of applicable validated rating scales;
• Behavioral health care planning in relation to behavioral/psychiatric health problems, including revision for patients who are not progressing or whose status changes;
• Facilitating and coordinating treatment such as psychotherapy, pharmacotherapy, counseling and/or psychiatric consultation; and
• Continuity of care with a designated member of the care team.

Accordingly, the final code descriptor will be, G0507: Care management services for behavioral health conditions, at least 20 minutes of clinical staff time, directed by a physician or other qualified health care professional, per calendar month, with the following required elements:

• Initial assessment or follow-up monitoring, including the use of applicable validated rating scales;
• Behavioral health care planning in relation to behavioral/psychiatric health problems, including revision for patients who are not progressing or whose status changes;
• Facilitating and coordinating treatment such as psychotherapy, pharmacotherapy, counseling and/or psychiatric consultation; and
• Continuity of care with a designated member of the care team.

We are aware of a number of validated rating scales that are available for use for a number of conditions addressed by BHI models of care, such as those described by the Kennedy Forum (see http://thekennedyforum-dot-org.s3.amazonaws.com/documents/MBC_supplement.pdf). We are requiring the use of such scales when applicable to the condition(s) that are being treated. Medication Assisted Treatment (MAT) may be a treatment that is facilitated under the facilitating treatment service element.

Regarding diagnosis, we believe we should specify similar diagnostic criteria for G0507 and the psychiatric CoCM services (G0502, G0503 and G0504). Accordingly we are providing that beneficiaries who are appropriate candidates for services billed under G0507 will have an identified psychiatric or behavioral health condition(s) that requires a behavioral health care assessment, behavioral health care planning, and provision of interventions. Eligible beneficiaries must present with a condition(s) that in the treating practitioner's clinical judgment, warrants the services included in G0507. The presenting condition(s) may be pre-existing or newly diagnosed by the treating practitioner, and may be refined as treatment progresses. Beneficiaries receiving services reported under G0507 may, but are not required to have comorbid chronic or other medical condition(s) that are being managed by the treating practitioner. We are not limiting billing and payment for G0507 to a specified set of behavioral health conditions, because there may be overlap in behavioral health conditions; if we specified only certain diagnoses, practitioners might modify diagnoses to fit within payment rules; and for many beneficiaries for whom specialty care is not available, or who choose for other reasons to remain within primary care, their behavioral health condition(s) can be addressed using a model of integrated care.

Regarding rules for clinical staff, we are clarifying that services included in the code G0507 may be provided directly by the treating practitioner or provided by other qualifying individuals (whom we term “clinical staff”) under his or her direction, during the calendar month service period. Unlike the psychiatric CoCM codes, for G0507 there is not necessarily a specific individual designated as a “behavioral health care manager” with formal or specialized education in providing the services (although there could be). Similarly, there is not necessarily a psychiatric or other behavioral health specialist consultant (although there could be), and we note that G0507 is not valued to explicitly account for such a consultant. We will apply the same definition of the term “clinical staff” that we have applied for CCM to G0507, namely, the CPT definition of this term, subject to the incident to rules and regulations and applicable state law, licensure and scope of practice at 42 CFR 410.26. For G0507, then, we note that the term “clinical staff” will encompass or include a psychiatric or other behavioral health specialist consultant, if the treating practitioner obtains consultative expertise. Clinical staff that provide included services do not have to be employed by the treating practitioner or located on site, necessarily, and may or may not be a professional who is permitted to independently furnish and report services to Medicare. Time spent by administrative or clerical staff cannot be counted towards the time required to bill G0507.

G0507 is valued to include minimal work by the treating practitioner; the bulk of the valuation is based on clinical staff time (see section II.L on valuation). However, we want to emphasize that the treating practitioner must direct the service, continue to oversee the beneficiary's care, and perform ongoing management, collaboration and reassessment. If the service (or part thereof) is provided incident to the treating practitioner's services, whether on site or remotely, the clinical staff providing services must have a collaborative, integrated relationship with the treating practitioner. They must also have a continuous relationship with the beneficiary.

Evaluation and management services, such as face-to-face E/M visits, may be separately billed during the service period or on the same day as G0507, provided time is not counted twice towards the same code.

For payment purposes, we are categorizing this service as a designated care management service assigned general supervision for purposes of “incident to” billing, because we do not believe it is clinically necessary for the individuals on the team who provide services other than the treating practitioner (namely, clinical staff) to have the treating practitioner immediately available to them at all times, as would be required under a higher level of supervision. However, general supervision sets the minimum standard for supervision and does not, by itself, meet the requirements we are setting for billing new code G0507. While certain aspects of G0507 might be furnished under general supervision, we do not believe the general supervision requirement adequately describes the nature of the relationship and interactions of the respective team members for services furnished using BHI models of care or the codes we are creating to describe those services. Moreover the general supervision requirement only directly addresses the physical location of the treating practitioner, not the location of clinical staff, necessarily.

Comment: Regarding behavioral health care planning, some commenters noted that there is not necessarily value in accumulating or enumerating a number of different types of care plans addressing different aspects of the beneficiary's problems, such as a behavioral or psychiatric care plan, a CCM care plan, and a cognitive impairment care plan (see G0505 in section II.E.5).

Response: While any care planning should take into account the whole patient, our intent is that the care planning included in the CCM coding (and G0506, the CCM initiating visit add-on code) will be the most comprehensive in nature, addressing all health issues with particular focus on the multiple chronic conditions being managed by the treating practitioner. In contrast, the BHI care planning will focus on behavioral health or psychiatric issues, in particular, just as the cognitive impairment care planning will focus on cognitive impairment issues, in particular (see section II.E.5. of this final rule).

However, we understand that adoption of EHRs may be lower among behavioral health practitioners and note that resources are available to help inform how care plans can support team-based care and BHI. While we understand that practitioners, in general, are exploring a wide variety of innovative approaches and tools that facilitate care plan integration across clinical disciplines, at this time, there may not be sufficient adoption of interoperable health IT interoperability among all practitioners and providers treating a given beneficiary to necessarily have a single, master care plan that adequately addresses the progress of the beneficiary in relation to all of these issues. In general, practitioners are encouraged to pursue approaches that integrate health information from multiple sources into a single care plan, but we understand that practitioners may need to create separate documents or the relevant care planning may be documented in another format within the medical record.

We believe the format of the care plan(s) is less important than having a process whereby feedback and expertise from all relevant practitioners and providers, whether internal or external to the billing practice, are integrated into the beneficiary's treatment plan and goals; that this plan be regularly assessed and revisited by the practitioner who is assuming an overall care management role for the beneficiary in a given month; that the patient is engaged in the care planning process; and that the care planning be documented in the medical record (as with any required element of any PFS service). We have framed the care planning service element for G0507 accordingly, “Behavioral health care planning in relation to behavioral/psychiatric health problems, including revision for patients who are not progressing or whose status changes.”

Comment: We received a few comments recommending codes in addition to the psychiatric CoCM codes that would pay for similar services to inpatients, or for behavioral health services by psychologists to psychologically and medically complex patients in skilled nursing facilities (SNF) and nursing homes. Some of these commenters stated that in SNF and long-term care settings, psychologists work closely with primary care physicians, psychiatrists, nurses, and other consultants to improve outcomes by reducing inappropriate use or dosing of psychotropic medications, improving activities of daily living, and preventing avoidable admissions/falls. These commenters stated that many health systems employ psychologists as BHI team leaders or coordinators, and sought clarification on how psychologist-led teams would operationalize the new BHI codes. These commenters believe that psychology training provides unique skills in facilitating interdisciplinary teams. While they acknowledged that psychologists are not qualified to perform the full range of BHI services and interventions, they believed psychologists should be able to separately report and bill for care coordination and BHI initiation activities.

We received similar comments supporting the addition of psychiatric collaborative care services to the PFS, and other evidence-based models in a variety of primary care-based treatment settings. However, these commenters supported the inclusion of social workers at all levels of licensures as reimbursable providers of these services.

Response: We appreciate the commenters' descriptions of some particular working models of care, and we welcome additional information in this regard. We continue to believe it would be appropriate to have new coding for a range of BHI care models applicable to inpatient as well as outpatient and facility settings. Our goal in separately identifying and paying for BHI services is to prioritize accurate payment for these services, in recognition of the associated time and complexity of the services. We agree that beneficiaries who are admitted to a facility, are in long-term care, or are transitioning among settings during the month are likely to be more complex than other types of patients, and to warrant more- not less- BHI services. Therefore, we have valued G0507 in both facility and non-facility settings (see section II.L on valuation). We are not limiting the time that can be counted towards the monthly time requirement to bill G0507 to time that is spent in the care of an outpatient or a beneficiary residing in the community. As we provide for the psychiatric CoCM services, G0507 may be reported by specialties that are not “traditional” primary care specialties, if such specialists furnish the included services. However, we stress that G0507 can only be reported by a treating physician or other qualified health care professional when he or she has directed the BHI service for the duration of time that he or she is reporting it, and has a qualifying relationship with individuals providing the service under his or her direction and control. Also, time and effort that is spent managing care transitions for CCM or TCM patients and that is counted towards reporting TCM or CCM services, cannot also be counted towards reporting any transitional care management activities reported under a BHI service code(s). We welcome additional input from stakeholders regarding appropriate (or inappropriate) settings of service for G0507.

Since the BHI initiating visit that is required to bill G0507 is not within the scope of practice of a psychologist or social worker (see below), psychologists and social workers will not be able to report G0507 directly (although a psychiatrist may be able to do so). Psychologists and social workers may provide care management services included in G0507 incident to the services of another (billing) practitioner. They may also provide services that are separately billable during the service period. We appreciate the commenters' support for team-based care, and we recognize the substantial role of various types of mental health professionals within a primary care team. We are interested in receiving additional input from stakeholders as to whether and why behavioral health care management services by a social worker, psychologist or similarly qualified professional should be reportable in its own right, rather than incident to the services of a practitioner authorized to bill Medicare for a BHI initiating visit. Consistent with our recent approaches to making proposals under PFS notice and comment rulemaking, we could consider adopting new coding under a different construct that was not defined as BHI, if stakeholders provided sufficient input on how to design, define and value the services. We would also consider such changes if adopted by the CPT Editorial Panel, per our usual process. BHI integrates behavioral health expertise into evaluation and management care. Therefore G0507 is designed to include services that require the oversight and involvement of a practitioner who can perform evaluation and management services, including facilitation of any needed pharmacotherapy, referral for specialty care, and overall management of the beneficiary's treatment in relation to primary care treatment. We note that G0507 would not be independently billed by psychologists or social workers, though from our understanding of various models of BHI, these professionals seem likely to be participants in team-based care for beneficiaries receiving these services.

c. BHI Initiating Visit

Similar to CCM services (see section II.E.4), we proposed to require an initiating visit for all of the BHI codes (G0502, G0503, G0504 and G0507) that would be billable separate from the BHI services themselves. We proposed that the same services that can serve as the initiating visit for CCM services (see section II.E.4.a. of this final rule) could serve as the initiating visit for the proposed BHI codes. The initiating visit would establish the beneficiary's relationship with the billing practitioner (most aspects of the BHI services would be furnished incident to the billing practitioner's professional services), ensure the billing practitioner assesses the beneficiary prior to initiating care management processes, and provide an opportunity to obtain beneficiary consent (discussed below). We solicited public comment on the types of services that are appropriate for an initiating visit for the BHI codes, and within what timeframe the initiating visit should be conducted prior to furnishing BHI services.

Comment: The commenters were largely supportive of our proposal to allow the same services to qualify for the initiating visit to CCM as for the initiating visit to BHI services. We received a few comments stating that in addition to the qualifying E/M services (or an AWV or IPPE), initiating services should include in-depth psychological evaluations delivered by a psychologist including CPT codes 90791, 96116 or 96118 which, in turn, include care plan development. These commenters agreed that psychologists cannot personally furnish all BHI services (for example, medication reconciliation), but believe psychologists effectively coordinate care and perform other aspects of BHI services as part of a team under current practice models. They believe this approach would be particularly effective for reducing inappropriate use or dosing of psychotropic medications in elderly and complex patients, improving activities of daily living, and preventing avoidable admissions and falls.

Response: We appreciate the commenters' feedback. We agree that psychologists would be qualified to perform care coordination that is included in the psychiatric CoCM codes (G0502, G0503 and G0504) and the general BHI code (G0507) under the direction of a physician or other qualified health care professional. In addition, beneficiaries receiving BHI services under any of those codes may be referred to psychologists for psychotherapy or other services that are separately billable and within the scope of practice of psychologists, as discussed elsewhere in this section of our final rule. However many commenters acknowledged, and we agree, that a BHI initiating visit is necessary. The initiating visit is not, in its entirety, within the scope of psychologist practice. Therefore, we are finalizing our proposal that the same services that qualify as the initiating visit for CCM will also qualify as initiating services for BHI, and they do not include in-depth psychological evaluation by a psychologist. Also, we will require an initiating visit for BHI only for new patients or beneficiaries not seen within a year of commencement of BHI services (the same requirement we are finalizing for CCM, see section II.E.4.a). As more experience is gained with the psychiatric CoCM services and other models of BHI care, we may reassess these provisions.

As discussed above, we are interested in receiving input from stakeholders regarding circumstances other than BHI in which behavioral health care management services by a psychologist, social worker or similarly qualified professional should be reportable in its own right, rather than incident to the services of a practitioner authorized to bill Medicare for a BHI initiating visit.

Comment: Some commenters recommended that CMS establish an add-on code to the initiating visit for BHI services, parallel to G0506 (the proposed add-on code for the CCM initiating visit).

Response: We do not believe we have enough information about practice patterns at this time to create an add-on code to the BHI initiating visit, and we did not propose such a code. We may re-examine this issue in the future.

d. Beneficiary Consent for BHI Services

Commenters to the CY 2016 PFS proposed rule indicated that they did not believe a specific patient consent for BHI services is necessary and indicated that requiring special informed consent for these services may reduce access due to stigma associated with behavioral health conditions. Instead, the commenters recommended requiring a more general consent prior to initiating these services whereby the beneficiary gives the initiating physician or practitioner permission to consult with relevant specialists, which would include conferring with a psychiatric consultant. Accordingly, we proposed to require a general beneficiary consent to consult with relevant specialists prior to initiating these services, recognizing that applicable rules continue to apply regarding privacy. The proposed general consent would encompass conferring with a psychiatric consultant when furnishing the psychiatric CoCM codes (G0502, G0503, and G0504) or the proposed broader BHI code (G0507). Similar to the proposed beneficiary consent process for CCM services, we proposed that the billing practitioner must document in the beneficiary's medical record that the beneficiary's consent was obtained to consult with relevant specialists including a psychiatric consultant, and that, as part of the consent, the beneficiary is informed that there is beneficiary cost-sharing, including potential deductible and coinsurance amounts, for both in-person and non-face-to-face services that are provided. We solicited stakeholder comments on this proposal.

We recognized that special informed consent could also be helpful in cases when a particular service is limited to being billed by a single practitioner for a particular beneficiary. We did not believe that there are circumstances where it would reasonable for multiple practitioners to be reporting these codes during the same month. However, we did not propose a formal limit at this time. We solicited comment on whether such a limitation would be beneficial or whether there are circumstances under which a beneficiary might reasonably receive BHI services from more than one practitioner during a given month.

Comment: The commenters were largely supportive of our proposal regarding BHI consent, some noting that physician-to-physician communication as well as communication within treatment teams happens routinely, without an extra layer of formal written consent, for other medical conditions. A few commenters intimated that CMS might pursue a single broad consent that could be used across care management services; for example, applying for both CCM and BHI. We did not receive any public comments delineating the circumstances under which it would be appropriate to bill for services furnished using more than one BHI service model per month, or appropriate for more than one practitioner (whether in the same practice or different practices) to bill for services furnished in a BHI care model per month.

Response: We agree with the commenters that physician-to-physician communication as well as communication within treatment teams happens routinely, without an extra layer of formal written consent, for other medical conditions. However there are particular privacy concerns addressed by other rules and regulations for some behavioral health or substance use care. Also we are concerned that beneficiaries should not incur unexpected expenses for care that is largely, or in significant part, non-face-to-face in nature. Finally, there are issues to consider, that we considered for CCM, regarding prevention of duplicative practitioner billing, and whether BHI services can actually be furnished under the direction and control of any given practitioner if for a given service period, more than one practitioner is furnishing BHI services and billing them.

The public comments were supportive of our proposal for a broad consent that could be verbally obtained but must be documented in the medical record, and we are finalizing as proposed. At this time, we do not believe a single consent process for both BHI and CCM is advisable. It is not clear how frequently BHI and CCM would or should be furnished concurrently. BHI and CCM are distinct, separate services, having significant differences in time thresholds, the nature of the services, types of individuals providing the services, and payment and cost sharing amounts. Therefore, at this time, we are maintaining separate consent processes for CCM and BHI, as provided in the respective sections of this final rule. Also, as discussed in section II.E.4 on CCM, CCM and BHI may be billed during the same service period.

It remains unclear whether it would be reasonable and necessary for more than one practitioner (whether in the same practice or different practices) to bill BHI services for a given beneficiary for a given service period, given the lack of public response and input on this issue. It may depend on the conditions(s) being treated and whether specialty care, other than psychiatric or behavioral health specialty care, and primary care are both involved. We are not proposing a formal limit at this time, but we stress that BHI services can only be reported by a treating physician or other qualified health care professional when he or she has obtained the required beneficiary consent, directed the BHI services he or she reports for the duration of time reported, and has a qualifying relationship with individuals providing the reported services under his or her direction and control. We would not expect a single practitioner to furnish care to a given beneficiary under more than one BHI model of care during a given month. Therefore a single practitioner must choose whether to report psychiatric CoCM code(s) (G0502, G0503, and G0504 as applicable) or the general BHI code (G0507) for a given month for a given beneficiary. We remind stakeholders that time cannot be counted more than once towards any code(s), all services must be medically reasonable and necessary, and that beneficiary cost sharing and advance consent apply. We will be monitoring the claims data and studying the utilization patterns. We will continue to assess appropriate reporting patterns, and we expect that potential coding changes by the CPT Editorial Panel may inform this issue.

Comment: We received a number of comments recommending that cost sharing be removed for all care management services, whether through legislative change, demonstration, waiver safe harbor, or designation as preventive services.

Response: We appreciate commenters' concerns and recognize many of the challenges associated with patient cost-sharing for these kinds of services. At this time, we do not have authority to waive cost sharing for the BHI or other care management services. We appreciate the commenters' acknowledgement of our current limitations and we will continue to consider this issue.

e. Summary of Final BHI Policies

Beginning in CY 2017, we are providing separate payment for a range of BHI services. Specifically, we are providing payment for psychiatric CoCM services under the following codes:

• G0502: Initial psychiatric collaborative care management, first 70 minutes in the first calendar month of behavioral health care manager activities, in consultation with a psychiatric consultant, and directed by the treating physician or other qualified health care professional, with the following required elements:

++ Outreach to and engagement in treatment of a patient directed by the treating physician or other qualified health care professional;

++ Initial assessment of the patient, including administration of validated rating scales, with the development of an individualized treatment plan;

++ Review by the psychiatric consultant with modifications of the plan if recommended;

++ Entering patient in a registry and tracking patient follow-up and progress using the registry, with appropriate documentation, and participation in weekly caseload consultation with the psychiatric consultant; and

++ Provision of brief interventions using evidence-based techniques such as behavioral activation, motivational interviewing, and other focused treatment strategies.

• G0503: Subsequent psychiatric collaborative care management, first 60 minutes in a subsequent month of behavioral health care manager activities, in consultation with a psychiatric consultant, and directed by the treating physician or other qualified health care professional, with the following required elements:

++ Tracking patient follow-up and progress using the registry, with appropriate documentation;

++ Participation in weekly caseload consultation with the psychiatric consultant;

++ Ongoing collaboration with and coordination of the patient's mental health care with the treating physician or other qualified health care professional and any other treating mental health providers;

++ Additional review of progress and recommendations for changes in treatment, as indicated, including medications, based on recommendations provided by the psychiatric consultant;

++ Provision of brief interventions using evidence-based techniques such as behavioral activation, motivational interviewing, and other focused treatment strategies;

++ Monitoring of patient outcomes using validated rating scales; and relapse prevention planning with patients as they achieve remission of symptoms and/or other treatment goals and are prepared for discharge from active treatment.

• G0504: Initial or subsequent psychiatric collaborative care management, each additional 30 minutes in a calendar month of behavioral health care manager activities, in consultation with a psychiatric consultant, and directed by the treating physician or other qualified health care professional (List separately in addition to code for primary procedure) (Use G0504 in conjunction with G0502, G0503).

These psychiatric CoCM services are reported by the treating physician or other qualified health care professional for services furnished during a calendar month service period. These services may be furnished when a beneficiary has a psychiatric or behavioral health condition(s) that in the treating physician or other qualified health care professional's clinical judgment, requires a behavioral health care assessment; establishing, implementing, revising, or monitoring a care plan; and provision of brief interventions. The diagnosis or diagnoses may be pre-existing or made by the treating physician or other qualified health care professional, and may be refined over time. The psychiatric CoCM services may be furnished to beneficiaries with any psychiatric or behavioral health condition(s) that is being treated by the physician or other qualified health care professional, including substance use disorders. Beneficiaries receiving psychiatric CoCM services may, but are not required to have comorbid chronic or other medical condition(s) that are being managed by the treating practitioner.

Psychiatric CoCM services include the services of the treating physician or other qualified health care professional, the behavioral health care manager (see description below) who provides services incident to services of the treating physician or other qualified health care professional, and the psychiatric consultant (see description below) whose consultative services are furnished incident to services of the treating physician or other qualified health care professional. Time spent by administrative or clerical staff cannot be counted towards the time required to bill the psychiatric CoCM service codes.

Beneficiaries receiving psychiatric CoCM services may have newly diagnosed conditions, need help in engaging in treatment, have not responded to standard care delivered in a non-psychiatric setting, or require further assessment and engagement prior to consideration of referral to a psychiatric care setting. Beneficiaries are treated for an episode of care, defined as beginning when the behavioral health care manager engages in care of the beneficiary under the appropriate supervision of the billing practitioner and ending with:

• The attainment of targeted treatment goals, which typically results in the discontinuation of care management services and continuation of usual follow-up with the treating physician or other qualified healthcare professional; or
• Failure to attain targeted treatment goals culminating in referral to a psychiatric care provider for ongoing treatment; or
• Lack of continued engagement with no psychiatric collaborative care management services provided over a consecutive 6-month calendar period (break in episode).

A new episode of care will start after a break in episode of 6 calendar months or more.

The treating physician or other qualified health care professional directs the behavioral health care manager and continues to oversee the beneficiary's care, including prescribing medications, providing treatments for medical conditions, and making referrals to specialty care when needed. The treating physician or other qualified health care professional must remain involved in ongoing oversight, management, collaboration and reassessment as appropriate to bill the psychiatric CoCM codes.

The behavioral health care manager has formal education or specialized training in behavioral health (which could include a range of disciplines, for example, social work, nursing, and psychology). The behavioral health care manager provides care management services, as well as an assessment of needs, including the administration of validated rating scales; behavioral health care planning in relation to behavioral/psychiatric health problems, including revision for patients who are not progressing or whose status changes; provision of brief interventions; ongoing collaboration with the treating physician or other qualified health care professional; maintenance of a registry; all in consultation with the psychiatric consultant. The behavioral health care manager is available to provide these services face-to-face and non-face-to-face, and consults with the psychiatric consultant minimally on a weekly basis.

The behavioral health care manager must have a collaborative, integrated relationship with the rest of the care team members, and be able to perform all of the required elements of the service delineated for the behavioral health care manager. The behavioral health care manager must have the ability to engage the beneficiary outside of regular clinic hours as necessary to perform the behavioral health care manager's duties under the psychiatric CoCM model, and must have a continuous relationship with the beneficiary. The behavioral health care manager may or may not be a professional who meets all the requirements to independently furnish and report services to Medicare. The behavioral health care manager is subject to the incident to rules and regulations and applicable state law, licensure and scope of practice (see 42 CFR 410.26).

The psychiatric consultant is a medical professional trained in psychiatry and qualified to prescribe the full range of medications. The psychiatric consultant advises and makes recommendations, as needed, for psychiatric and other medical care, including psychiatric and other medical diagnoses, treatment strategies including appropriate therapies, medication management, medical management of complications associated with treatment of psychiatric disorders, and referral for specialty services, that are communicated to the treating physician or other qualified health care professional, typically through the behavioral health care manager. The psychiatric consultant does not typically see the beneficiary or prescribe medications, except in rare circumstances, but can and should facilitate referral for direct provision of psychiatric care when clinically indicated. The psychiatric consultant is subject to the incident to rules and regulations and applicable state law, licensure and scope of practice (see 42 CFR 410.26).

Beginning in CY 2017, we are providing separate payment for BHI services furnished under models of care other than the psychiatric CoCM model, under HCPCS code G0507: Care management services for behavioral health conditions, at least 20 minutes of clinical staff time, directed by a physician or other qualified health care professional, per calendar month, with the following required elements:

• Initial assessment or follow-up monitoring, including the use of applicable validated rating scales;
• Behavioral health care planning in relation to behavioral/psychiatric health problems, including revision for patients who are not progressing or whose status changes;
• Facilitating and coordinating treatment such as psychotherapy, pharmacotherapy, counseling and/or psychiatric consultation; and
• Continuity of care with a designated member of the care team.

G0507 is reported by the treating physician or other qualified health care professional for services furnished during a calendar month service period. This service may be furnished when the beneficiary has a psychiatric or behavioral health condition(s) that in the treating physician or other qualified health care professional's clinical judgment, requires a behavioral health care assessment, behavioral health care planning, and provision of interventions. The presenting condition(s) may be pre-existing or newly diagnosed by the treating physician or other qualified health care professional, and may be refined over time. Beneficiaries receiving services reported under G0507 may have any psychiatric or behavioral health condition(s) that is being treated by the physician or other qualified health care professional, including substance use disorders. Beneficiaries receiving services reported under G0507 may, but are not required to have comorbid chronic or other medical condition(s) that are being managed by the treating practitioner.

Services reported under G0507 may be provided directly by the treating physician or other qualified health care professional, or provided by clinical staff under his or her direction, during a calendar month service period. For G0507, there is not necessarily a specific individual designated as a “behavioral health care manager” with formal or specialized education in providing the services (although there could be). Similarly, there is not necessarily a psychiatric or other behavioral health specialist consultant (although there could be) and we note that G0507 is not valued to explicitly account for expert consultation. For G0507, the term “clinical staff” means the CPT definition of this term, subject to the incident to rules and regulations and applicable state law, licensure and scope of practice at 42 CFR 410.26. For G0507, then, we note that the term “clinical staff” will encompass or include any psychiatric or other behavioral health specialist consultant that may provide consultative services. Clinical staff providing services are not required to be employed by the treating practitioner or located on site, and these individuals may or may not be a professional permitted to independently furnish and report services to Medicare. Time spent by administrative or clerical staff cannot be counted towards the time required to report G0507. We emphasize that the physician or other qualified health care professional must direct the service, continue to oversee the beneficiary's care, and perform ongoing management, collaboration and reassessment. If the service (or part thereof) is provided incident to services of the treating practitioner, whether on site or remotely, the clinical staff providing services must have a collaborative, integrated relationship with the treating practitioner. They must also have a continuous relationship with the beneficiary.

For all of the BHI service codes (G0502, G0503, G0504 and G0507), we are requiring an initiating visit that is billable separate from the BHI services themselves. The same services that qualify as initiating visits for CCM services can serve as the initiating visit for BHI services (certain face-to-face E/M services including the face-to-face visit required for TCM services, and the AWV or IPPE). The BHI initiating visit establishes the beneficiary's relationship with the treating practitioner (BHI services may be furnished incident to the treating practitioner's professional services); ensures that the treating practitioner assesses the beneficiary prior to initiating care management processes; and provides an opportunity to obtain beneficiary consent (consent may also be obtained outside of the BHI initiating visit, as long as it is obtained prior to commencement of BHI services).

For all of the BHI service codes, we are also requiring prior beneficiary consent, recognizing that applicable rules continue to apply regarding privacy. The consent will include permission to consult with relevant specialists including a psychiatric consultant, and inform the beneficiary that cost sharing will apply to in-person and non-face-to-face services provided. Consent may be verbal (written consent is not required) but must be documented in the medical record.

For payment purposes, we are assigning general supervision to all of the BHI service codes (G0502, G0503, G0504 and G0507). However we note that general supervision does not, by itself, comprise a qualifying relationship between the treating practitioner and other individuals providing BHI services under the incident to relationship. Also we note that we valued BHI services in both facility and non-facility settings. BHI services may be furnished to beneficiaries in any setting of care. Time that is spent furnishing BHI services to a beneficiary who is an inpatient or in any other facility setting during service provision or for any part of the service period may be counted towards reporting a BHI code(s). We refer the reader to our discussion above on this matter, as well as reporting by specialty, intersection with other services, and potential reporting by more than one practitioner for a given beneficiary within a service period. A single practitioner must choose whether to report psychiatric CoCM code(s) (G0502, G0503, and G0504 as applicable) or the general BHI code (G0507) for a given month (service period) for a given beneficiary.

4. Reducing Administrative Burden and Improving Payment Accuracy for Chronic Care Management (CCM) Services

Beginning in CY 2015, we implemented separate payment for CCM services under CPT code 99490 (Chronic care management services, at least 20 minutes of clinical staff time directed by a physician or other qualified health professional, per calendar month, with the following required elements:

• Multiple (two or more) chronic conditions expected to last at least 12 months, or until the death of the patient;
• Chronic conditions place the patient at significant risk of death, acute exacerbation/decompensation, or functional decline;
• Comprehensive care plan established, implemented, revised, or monitored).

In the CY 2015 final rule with comment period, we finalized a proposal to make separate payment for CCM services as one initiative in a series of initiatives designed to improve payment for, and encourage long-term investment in, care management services (79 FR 67715). In particular, we sought to address an issue raised to us by the physician community, which stated that the care management included in many of the existing E/M services, such as office visits, does not adequately describe the typical non-face-to-face care management work required by certain categories of beneficiaries (78 FR 43337). We began to re-examine how Medicare should pay under the PFS for non-face-to-face care management services that were bundled into the PFS payment for face-to-face E/M visits, being included in the pre- and post-encounter work (78 FR 43337). In proposing separate payment for CCM, we acknowledged that, even though we had previously considered non-face-to-face care management services as bundled into the payment for face-to-face E/M visits, the E/M office/outpatient visit CPT codes may not reflect all the services and resources required to furnish comprehensive, coordinated care management for certain categories of beneficiaries. We stated that we believed that the resources required to furnish complex chronic care management services to beneficiaries with multiple (that is, two or more) chronic conditions were not adequately reflected in the existing E/M codes. Medical practice and patient complexity required physicians, other practitioners and their clinical staff to spend increasing amounts of time and effort managing the care of comorbid beneficiaries outside of face-to-face E/M visits, for example, complex and multidisciplinary care modalities that involve regular physician development and/or revision of care plans; subsequent report of patient status; review of laboratory and other studies; communication with other health care professionals not employed in the same practice who are involved in the patient's care; integration of new information into the care plan; and/or adjustments of medical therapy.

Therefore, in the CY 2014 PFS final rule with comment period, we established a separate payment under the PFS for CPT code 99490 (78 FR 43341 through 43342). We sought to include a relatively broad eligible patient population within the code descriptor, established a moderate payment amount, and established bundled payment for concurrently new CPT codes that were reserved for beneficiaries requiring “complex” CCM services (base CPT code 99487 and its add-on code 99489) (79 FR 67716 through 67719). We stated that we would evaluate the services reported under CPT code 99490 to assess whether the service is targeted to the right population and whether the payment amount is appropriate (79 FR 67719). We remind stakeholders that CMS did not limit the eligible population to any particular list of chronic conditions other than the language in the CPT code descriptor. Accordingly, one or more of the chronic conditions being managed through CCM services could be chronic mental health or behavioral health conditions or chronic cognitive disorders, as long as the chronic conditions meet the eligibility language in the CPT code descriptor for CCM services and the billing practitioner meets all of Medicare's requirements to bill the code including comprehensive, patient-centered care planning for all health conditions.

In finalizing separate payment for CPT code 99490, we considered whether we should develop standards to ensure that physicians and other practitioners billing the service would have the capability to fully furnish the service (79 FR 67721). We sought to make certain that the newly payable PFS code(s) would provide beneficiary access to appropriate care management services that are characteristic of advanced primary care, such as continuity of care; patient support for chronic diseases to achieve health goals; 24/7 patient access to care and health information; receipt of preventive care; patient, family and caregiver engagement; and timely coordination of care through electronic health information exchange. Accordingly, we established a set of scope of service elements and payment rules in addition to or in lieu of those established in CPT guidance (in the CPT code descriptor and CPT prefatory language), that the physician or nonphysician practitioner must satisfy to fully furnish CCM services and report CPT code 99490 (78 FR 74414 through 74427, 79 FR 67715 through 67730, and 80 FR 14854). We established requirements to furnish a preceding qualifying visit, obtain advance written beneficiary consent, use certified electronic health record (EHR) technology to furnish certain elements of the service, share the care plan and clinical summaries electronically, document specified activities, and other items summarized in Table 11 of our CY 2017 proposed rule. For the CCM service elements for which we required use of a certified EHR, the billing practitioner must use, at a minimum, technology meeting the edition(s) of certification criteria that is acceptable for purposes of the EHR Incentive Programs as of December 31st of the calendar year preceding each PFS payment year. (For the CY 2017 PFS payment year, this would mean technology meeting the 2014 edition of certification criteria).

These elements and requirements for separately payable CCM services are extensive and generally exceed those required for payment of codes describing procedures, diagnostic tests, or other E/M services under the PFS. In addition, both CPT guidance and Medicare rules specify that only a single practitioner who assumes the care management role for a given beneficiary can bill CPT code 99490 per service period (calendar month). Because the new CCM service closely overlapped with several Medicare demonstration models of advanced primary care (the Multi-Payer Advanced Primary Care Practice (MAPCP) demonstration and the Comprehensive Primary Care Initiative (CPCI)), we provided that practitioners participating in one of these two initiatives could not be paid for CCM services furnished to a beneficiary attributed by the initiative to their practice (79 FR 67729).

Given the non-face-to-face nature of CCM services, we also sought to ensure that beneficiaries would receive advance notice that Part B cost sharing applies since we currently have no legislative authority to “waive” cost sharing for this service. Also since only one practitioner can bill for CCM each service period, we believed the beneficiary notice requirement would help prevent duplicate payment to multiple practitioners.

Since the establishment of CPT code 99490 for separate payment of CCM services, in a number of forums and in public comments to the CY 2016 PFS final rule (80 FR 70921), many practitioners have stated that the service elements and billing requirements are burdensome, redundant and prevent them from being able to provide the services to beneficiaries who could benefit from them. Stakeholders have stated that CPT code 99490 is underutilized because it is underpaid relative to the resources involved in furnishing the services, especially given the extensive Medicare rules for payment, and they have suggested a number of potential changes to our current payment rules. Stakeholders continue to believe that many of the CCM payment rules are duplicative, and to recommend that we reduce the rules and expand CCM coding and payment to distinguish among different levels of patient complexity. We also note that section 103 of the MACRA requires CMS to assess and report to Congress (no later than December 31, 2017) on access to CCM services by underserved rural and racial and ethnic minority populations and to conduct an outreach/education campaign that is underway.

The professional claims data for CPT code 99490 show that utilization is steadily increasing but may remain low considering the number of eligible Medicare beneficiaries. To date, approximately 513,000 unique Medicare beneficiaries received the service an average of four times each, totaling $93 million in total payments. Since CPT code 99490 describes a minimum of 20 minutes of clinical staff time spent furnishing CCM services during a month and does not have an upper time limit, and since we currently do not separately pay the other codes in the CCM family of CPT codes (which would provide us with utilization data on the number of patients requiring longer service times during a billing period), we do not know how often beneficiaries required more than 20 minutes of CCM services per month. We also do not know their complexity relative to one another, other than meeting the acuity criteria in the CPT code descriptor. Initial information from practitioner interviews conducted as part of our CCM evaluation efforts indicates that practitioners overwhelmingly meet and exceed the 20-minute threshold time for billing CCM. Typically, these practitioners reported spending between 45 minutes and an hour per month on CCM services for each patient, with times ranging between 20 minutes and several hours per month. CCM beneficiaries tend to exhibit a higher disease burden, are more likely to be dually eligible for Medicare and Medicaid, and are older than the general Medicare fee-for-service population. However, absent multiple levels of CCM coding, we do not have comprehensive data on the relative complexity of the CCM services furnished to beneficiaries.

In light of this stakeholder feedback and our mandate under MACRA section 103 to encourage and report on access to CCM services, we proposed several changes in the payment rules for CCM services. Our primary goal, and our statutory mandate, is to pay as accurately as possible for services furnished to Medicare beneficiaries based on the relative resources required to furnish PFS services, including CCM services. In so doing, we also expect to facilitate beneficiaries' access to reasonable and necessary CCM services that improve health outcomes. First, for CY 2017 we proposed to more appropriately recognize and pay for the other codes in the CPT family of CCM services (CPT codes 99487 and 99489 describing complex CCM), consistent with our general practice to price services according to their relative ranking within a given family of services. We direct the reader to section II.L of this final rule for a discussion of valuation for base CPT code 99487 and its add-on CPT code 99489. The CPT code descriptors are:

• CPT code 99487—Complex chronic care management services, with the following required elements:

++ Multiple (two or more) chronic conditions expected to last at least 12 months, or until the death of the patient;

++ Chronic conditions place the patient at significant risk of death, acute exacerbation/decompensation, or functional decline;

++ Establishment or substantial revision of a comprehensive care plan;

++ Moderate or high complexity medical decision making;

++ 60 minutes of clinical staff time directed by a physician or other qualified health care professional, per calendar month.

• CPT code 99489—Each additional 30 minutes of clinical staff time directed by a physician or other qualified health care professional, per calendar month (List separately in addition to code for primary procedure).

As CPT provides, less than 60 minutes of clinical staff time in the service period could not be reported separately, and similarly, less than 30 minutes in addition to the first 60 minutes of complex CCM in a service period could not be reported. We would require 60 minutes of services for reporting CPT code 99487 and 30 additional minutes for each unit of CPT code 99489.

We proposed to adopt the CPT provision that CPT codes 99487, 99489 and 99490 may only be reported once per service period (calendar month) and only by the single practitioner who assumes the care management role with a particular beneficiary for the service period. That is, a given beneficiary would be classified as eligible to receive either complex or non-complex CCM during a given service period, not both, and only one professional claim could be submitted to the PFS for CCM for that service period by one practitioner.

Comment: Several commenters were supportive of separate payment for complex CCM services.

Response: We thank the commenters for their support and are finalizing separate payment for CPT codes 99487 and 99489 as proposed. As finalized, these separate payments for complex CCM services will support care management for the most complex and time-consuming cases of beneficiaries with multiple chronic conditions.

Except for differences in the CPT code descriptors, we proposed to require the same CCM service elements for CPT codes 99487, 99489 and 99490. In other words, all the requirements in Table 11 of our proposed rule would apply, whether the code being billed for the service period is CPT code 99487 (plus CPT code 99489, if applicable) or CPT code 99490. These three codes would differ in the amount of clinical staff service time provided; the complexity of medical decision-making as defined in the E/M guidelines (determined by the problems addressed by the reporting practitioner during the month); and the nature of care planning that was performed (establishment or substantial revision of the care plan for complex CCM versus establishment, implementation, revision or monitoring of the care plan for non-complex CCM). Billing practitioners could consider identifying beneficiaries who require complex CCM services using criteria suggested in CPT guidance (such as number of illnesses, number of medications or repeat admissions or emergency department visits) or the profile of typical patients in the CPT prefatory language, but these would not comprise Medicare conditions of eligibility for complex CCM.

We proposed several changes to our current scope of service elements for CCM, and proposed that the same scope of service elements, as amended, would apply to all codes used to report CCM services beginning in 2017 (i.e., CPT codes 99487, 99489 and 99490). In particular, we proposed changes in the requirements for the initiating visit, 24/7 access to care and continuity of care, format and sharing of the care plan and clinical summaries, beneficiary receipt of the care plan, beneficiary consent and documentation.

Comment: Commenters were broadly supportive of these proposals. We received several comments recommending changes to the scope of service for non-complex CCM that might improve the distinction between non-complex and complex CCM and inform which “level” of service a given beneficiary is eligible for. For example, these commenters suggested changes to the time included in the code descriptor to reflect two or more time increments for CPT code 99490 using add-on codes, or retaining the current low time threshold while allowing practitioners to choose among certain service elements. Some commenters do not believe CPT code 99490 is intended for beneficiaries who require all the current service elements in a given month, and that only a more limited set of elements is medically necessary for the non-complex population.

Response: We appreciate the commenters' recommendations about how we might better distinguish complex CCM services from non-complex CCM services. The CPT Editorial Panel currently maintains the coding for CCM services. Further changes in codes and/or descriptors may be appropriately addressed by CPT and in subsequent PFS rulemaking.

a. CCM Initiating Visit Add-On Code (G0506)

As provided in the CY 2014 PFS final rule with comment period (78 FR 74425) and subregulatory guidance (available at https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/Downloads/Payment_for_CCM_Services_FAQ.pdf), CCM must be initiated by the billing practitioner during a “comprehensive” E/M visit, AWV or IPPE. This face-to-face, initiating visit is not part of the CCM service and can be separately billed to the PFS, but is required before CCM services can be provided directly or under other arrangements. The billing practitioner must discuss CCM with the patient at this visit. While informed patient consent does not have to be obtained during this visit, the visit is an opportunity to obtain the required consent. The face-to-face visit included in transitional care management (TCM) services (CPT codes 99495 and 99496) qualifies as a “comprehensive” visit for CCM initiation. Levels 2 through 5 E/M visits (CPT codes 99212 through 99215) also qualify; CMS does not require the practice to initiate CCM during a level 4 or 5 E/M visit. However, CPT codes that do not involve a face-to-face visit by the billing practitioner or are not separately payable by Medicare (such as CPT code 99211, anticoagulant management, online services, telephone and other E/M services) do not qualify as initiating visits. If the practitioner furnishes a “comprehensive” E/M, AWV, or IPPE and does not discuss CCM with the patient at that visit, that visit cannot count as the initiating visit for CCM.

We continued to believe that we should require an initiating visit in advance of furnishing CCM services, separate from the services themselves, because a face-to-face visit establishes the beneficiary's relationship with the billing practitioner and most aspects of the CCM services are furnished incident to the billing practitioner's professional services. The initiating visit also ensures collection of comprehensive health information to inform the care plan. We continued to believe that the types of face-to-face services that qualify as an initiating visit for CCM are appropriate. We did not propose to change the kinds of visits that can qualify as initiating CCM visits. However, we proposed to require the initiating visit only for new patients or patients not seen within one year instead of for all beneficiaries receiving CCM services. We believed this would allow practitioners with existing relationships with patients who have been seen relatively recently to initiate CCM services without furnishing a potentially unnecessary E/M visit. We solicited public comment on whether a period of time shorter than one year would be more appropriate.

Comment: The commenters were generally supportive of requiring the CCM initiating visit only for beneficiaries who are new patients or have not been seen in a year. A few commenters suggested a 6-month timeframe, or adopting one year and reconsidering as we gain more experience with CCM. Some commenters misinterpreted our proposal as requiring face-to-face visits every year to periodically reassess the beneficiary or the appropriateness of CCM services. Some recommended a similar coding structure for specialists managing a single condition, in place of prolonged services, or for BHI services.

Response: Our intent was to revise the timeframe for the single CCM initiating visit that is required at the outset of services. We did not propose subsequent “re-initiation” of CCM services or face-to-face reassessment within a given timeframe. We discuss further below that we have some concerns about how to ensure that the billing practitioner remains involved in the beneficiary's care and continually reassesses the beneficiary's care, but at this time we do not believe we should require subsequent face-to-face visits within certain timeframes to address those concerns.

We believe that the proposed one-year timeframe for the single, CCM initiating visit is appropriate for CY 2017, so we are finalizing as proposed. We will require the CCM initiating visit only for new patients or patients not seen within the year prior to commencement of CCM (instead of for all beneficiaries receiving CCM services). We will continue to consider in future years whether a different timeframe is warranted. The goal of our final policy is to allow practitioners with existing relationships with beneficiaries who have been seen relatively recently to initiate CCM services (for the first time) without furnishing a potentially unnecessary E/M visit. Regarding subsequent visits (after CCM services begin), practitioners are already permitted to furnish and separately bill subsequent E/M visits (or AWVs) for beneficiaries receiving CCM services. If a face-to-face reassessment is reasonable and necessary and furnished by the billing practitioner, then he or she may bill an appropriate code describing the face-to-face assessment of a beneficiary to whom they have previously furnished CCM services.

We also proposed for CY 2017 to create a new add-on G-code that would improve payment for services that qualify as initiating visits for CCM services. The code would be billable for beneficiaries who require extensive face-to-face assessment and care planning by the billing practitioner (as opposed to clinical staff), through an add-on code to the initiating visit, G0506 (Comprehensive assessment of and care planning by the physician or other qualified health care professional for patients requiring chronic care management services (billed separately from monthly care management services) (Add-on code, list separately in addition to primary service)).

We proposed that when the billing practitioner initiating CCM personally performs extensive assessment and care planning outside of the usual effort described by the billed E/M code (or AWV or IPPE code), the practitioner could bill G0506 in addition to the E/M code for the initiating visit (or in addition to the AWV or IPPE), and in addition to the CCM CPT code 99490 (or proposed 99487 and 99489) if all requirements to bill for CCM services are also met. We proposed valuation for G0506 in a separate section of our proposed rule.

The code G0506 would account specifically for additional work of the billing practitioner in personally performing a face-to-face assessment of a beneficiary requiring CCM services, and personally performing CCM care planning (the care planning could be face-to-face and/or non-face-to-face) that is not already reflected in the initiating visit itself (nor in the monthly CCM service code). We believed G0506 might be particularly appropriate to bill when the initiating visit is a less complex visit (such as a level 2 or 3 E/M visit), although G0506 could be billed along with higher level visits if the billing practitioner's effort and time exceeded the usual effort described by the initiating visit code. It could also be appropriate to bill G0506 when the initiating visit addresses problems unrelated to CCM, and the billing practitioner does not consider the CCM-related work he or she performs in determining what level of initiating visit to bill. We believed that this proposal would more appropriately recognize the relative resource costs for the work of the billing practitioner in initiating CCM services, specifically for extensive work assessing the beneficiary and establishing the CCM care plan that is reasonable and necessary, and that is not accounted for in the billed initiating visit or in the unit of the CCM service itself that is billed for a given service period. In addition, we believed this proposal would help ensure that the billing practitioner personally performs and meaningfully contributes to the establishment of the CCM care plan when the patient's complexity warrants it.

Comment: Several commenters were supportive of the add-on code (G0506) to the CCM initiating visit to describe physician assessment and care planning for patients requiring CCM services. Some commenters raised questions about whether G0506 should be a one-time service or could also be billed as an add-on code to subsequent reassessments by the billing practitioner (whether E/M visits or subsequent AWVs).

Response: At this time, we do not believe we should permit billing of G0506 more than once by the billing practitioner for a given beneficiary. G0506 was proposed as an add-on code to the single initiating visit, to help ensure the billing practitioner's assessment and involvement at the outset of CCM services. At this time there are no requirements for the billing practitioner to “re-initiate” CCM services; therefore we do not believe we should create an add-on code for a CCM “re-initiation” service. We would have to define “re-initiation” and develop rules regarding when subsequent E/M visits or AWVs are related to the performance of CCM. We do not believe beneficiaries would understand why they are incurring additional cost sharing for an add-on code to a “re-initiation” visit that has not been required or defined by CMS.

As we stated in the CY 2017 proposed rule, we were very interested in coding that was presented to the CPT Editorial Panel, but not adopted, to create code(s) that would separately identify and account for monthly CCM work by the billing practitioner. Such coding may be a better means of separately identifying and valuing the subsequent work of the billing practitioner after CCM is initiated. We want to establish policies that help ensure that the billing practitioner is not merely handing the beneficiary off to a remote care manager under general supervision while no longer remaining involved in their care. We believe that the practitioner billing CCM services should be actively re-assessing the beneficiary's chronic conditions, reviewing whether treatment goals are being met, updating the care plan, performing any medical decision-making that is not within the scope of practice of clinical staff, performing any necessary face-to-face care, and performing other related work. However, it would be more straightforward to separately identify this CCM-related work under code(s) that in their own right describe it, instead of add-on codes to very broadly drawn E/M codes where it becomes difficult to assess the relationship between the two services. Also for beneficiaries receiving complex CCM, some of this work is explicitly included in the complex CCM service codes (i.e., medical decision-making of moderate to high complexity). Therefore, at this time, G0506 will only serve as an add-on code to describe work performed by the billing practitioner once, in conjunction with the start or initiation of CCM services.

We note that despite the role of the billing practitioner in the initiation and provision of CCM services provided by clinical staff, non-complex CCM (CPT code 99490) is described based on the time spent by clinical staff. Complex CCM (CPT codes 99487 and 99489) similarly counts only clinical staff time, although it also includes complex medical decision-making by the billing practitioner. This raises issues regarding appropriate valuation in the facility setting that we will continue to consider in future rulemaking. The facility PE RVU for CCM includes indirect PE (which is an allocation based on physician work), but no direct PE (which would be comprised of other labor, supplies and equipment). This is because historically, the PFS facility rate assumes that the billing practitioner is not bearing a significant resource cost in labor by other individuals, equipment or supplies. Medicare assumes that those costs are instead borne by the facility and adequately accounted for in a separate payment made to the facility. The PFS non-facility rate generally does include such costs, assuming that the billing practitioner bears the resource costs in clinical and other staff labor, supplies and equipment.

For CCM, we have been considering whether this approach to valuation remains appropriate, because the service, in whole or in significant part, is provided by clinical staff under the direction of the billing practitioner. These individuals may provide the service or part thereof remotely, and are not necessarily employees or staff of the facility. Under this construct, there may be more direct practice expense borne by the billing practitioner that should be separately identified and valued over and above any institutional payment to the facility for its staff and infrastructure. We plan to explore these issues in future rulemaking and consider other approaches to valuation that would recognize the accurate relative resource costs to the billing practitioner for CCM and similar services furnished to beneficiaries who remain or reside in a facility setting during some or all of the service period.

Consistent with general coding guidance, we proposed that the work that is reported under G0506 (including time) could not also be reported under or counted towards the reporting of any other billed code, including any of the monthly CCM services codes. The care plan that the practitioner must create to bill G0506 would be subject to the same requirements as the care plan included in the monthly CCM services, namely, it must be an electronic patient-centered care plan based on a physical, mental, cognitive, psychosocial, functional and environmental (re)assessment and an inventory of resources and supports; a comprehensive care plan for all health issues. This would distinguish it from the more limited care planning included in the BHI codes G0502, G0503, G0504 or G0507 which focus on behavioral health issues, or the care planning included in G0505 which focuses on cognitive status. We sought public input on potential overlap among these codes and further clinical input as to how the assessments and care planning that is included in them would differ.

We received a number of comments regarding the relationship between proposed G0506, G0505 (Cognition and functional assessment by the physician or other qualified health care professional in office or other outpatient), prolonged non-face-to-face services, and BHI. We address these comments in the sections of this final rule regarding G0505, prolonged non-face-to-face services and BHI services (sections II.E.5, II.E.2 and II.E.3). In brief, we are not allowing G0506 and G0505 to be billed the same day (by a single practitioner). G0506 will not be an add-on code for the BHI initiating visit or BHI services. G0506 will be a one-time service code for CCM initiation, and the billing practitioner must choose whether to report either G0506 or prolonged services in association with CCM initiation (if requirements to bill both are met).

The CCM and BHI service codes differ substantially in potential diagnosis and comorbidity, the expected duration of the condition(s) being treated, the kind of care planning performed (comprehensive care planning versus care planning focused on behavioral/mental health issues), service elements and who performs them, and the interventions the beneficiary needs and receives apart from the CCM and BHI services themselves. The BHI codes include a more focused process than CCM for the clinical integration of primary care and behavioral health/psychiatric care, and for continual reassessment and treatment progression to a target or goal outcome that is specific to mental and behavioral health or substance abuse issues. However there is no explicit BHI service element for managing care transitions or systematic assessment of receipt of preventive services; there is no requirement to perform comprehensive care planning for all health issues (not just behavioral health issues); and there are different emphases on medication management or medication reconciliation, if applicable. In deciding which code(s) to report for services furnished to a beneficiary who is eligible for both CCM and BHI services, practitioners should consider which service elements were furnished during the service period, who provided them, how much time was spent, and should select the code(s) that most accurately and specifically identifies the services furnished without duplicative time counting. Practitioners should generally select the more specific code(s) when an alternative code(s) potentially includes the services provided. We are not precluding use of the CCM codes to report, or count, behavioral health care management if it is provided as part of a broader CCM service by a practitioner who is comprehensively overseeing all of the beneficiary's health issues, even if there are no imminent non-behavioral health needs. However, such behavioral care management activities could not also be counted towards reporting a BHI code(s). If a BHI service code more specifically describes the service furnished (service time and other relevant aspects of the service being equal), or if there is no focus on the health of the beneficiary outside of a narrower set of behavioral health issues, then it is more appropriate to report the BHI code(s) than the CCM code(s). Similarly, it may be more appropriate for certain specialists to bill BHI services than CCM services, since specialists are more likely to be managing the beneficiary's behavioral health needs in relation to a narrow subset of medical condition(s). CCM and BHI services can only be billed the same month for the same beneficiary if all the requirements to bill each service are separately met. We will monitor the claims data, and we welcome further stakeholder input to inform appropriate reporting rules.

b. 24/7 Access to Care, Continuity of Care, Care Plan and Managing Transitions

We proposed several revisions to the scope of service elements of 24/7 Access to Care, Continuity of Care, Care Plan and Managing Transitions. We continued to believe these elements are important aspects of CCM services, but that we should reduce the requirements for the use of specified electronic health information technology (IT) in their provision. In sum, we proposed to retain a core requirement to use a certified electronic health record (EHR), but allow fax to count for electronic transmission of clinical summaries and the care plan; no longer require access to the electronic care plan outside of normal business hours to those providing CCM services; and remove standards for clinical summaries in managing care transitions.

We sought to improve alignment with CPT provisions by removing the requirement for the care plan to be available remotely to individuals providing CCM services after hours. Studies have shown that after-hours care is best implemented as part of a larger practice approach to access and continuity (see for example, the peer-review article available at http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3475839/). There is substantial local variation in how 24/7 access and continuity of care are achieved, depending on the contractual relationships among practitioners and providers in a particular geographic area and other factors. Care models include various contractual relationships between physician practices and after-hours clinics, urgent care centers and emergency departments; extended primary care office hours; physician call-sharing; telephone triage systems; and health information technology such as shared EHRs and systematic notification procedures (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3475839/). Some or all of these may be used to provide access to urgent care on a 24/7 basis while maintaining information continuity between providers.

We recognized that some models of care require more significant investment in practice infrastructure than others, for example resources in staffing or health information technology. In addition, we believed there is room to reduce the administrative complexity of our current payment rules for CCM services to accommodate a range of potential care models. In re-examining what should be included in the CCM scope of service elements for 24/7 Access to Care and Continuity of Care, we believed the CPT language adequately and more appropriately describes the services that should, at a minimum, be included in these service elements. Therefore, we proposed to adopt the CPT language for these two elements. For 24/7 Access to Care, the scope of service element would be to provide 24/7 access to physicians or other qualified health care professionals or clinical staff including providing patients/caregivers with a means to make contact with health care professionals in the practice to address urgent needs regardless of the time of day or day of week. We believed the CPT language more accurately reflects the potential role of clinical staff or call-sharing services in addressing after-hours care needs than our current language does. In addition, the 24/7 access would be for “urgent” needs rather than “urgent chronic care needs,” because we believed after-hours services typically would and should address any urgent needs and not only those explicitly related to the beneficiary's chronic conditions.

We recognized that health information systems that include remote access to the care plan or the full EHR after hours, or a feedback loop that communicates back to the primary care physician and others involved in the beneficiary's care regarding after-hours care or advice provided, are extremely helpful (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3475839/#CR25). They help ensure that the beneficiary receives necessary follow up, particularly if he or she is referred to the emergency department, and follow up after an emergency department visit is required under the CCM element of Management of Care Transitions. Accordingly, we continued to support and encourage the use of interoperable EHRs or remote access to the care plan in providing the CCM service elements of 24/7 Access to Care, Continuity of Care, and Management of Care Transitions. However, adoption of such technology would be optimal not only for CCM services, but also for a number of other PFS services and procedures (including various other care management services), and we have not required adoption of any certified or non-certified health information technology as a condition of payment for any other PFS service. We noted that there are incentives under other Medicare programs to adopt such information technology, and were concerned that imposing too many EHR-related requirements at the service level as a condition of PFS payment could create disparities between these services and others under the fee schedule. Lastly, we recognized that not all after-hours care warrants follow-up or a feedback loop with the practitioner managing the beneficiary's care overall, and that under particular circumstances feedback loops can be achieved through oral, telephone or other less sophisticated communication methods. Therefore, we proposed to remove the requirement that the individuals providing CCM after hours must have access to the electronic care plan.

This proposal reflected our understanding that flexibility in how practices can provide the requisite 24/7 access to care, as well as continuity of care and management of care transitions, for their CCM patients could facilitate appropriate access to these services for Medicare beneficiaries. This proposal was not intended to undermine the significance of standardized communication methods as part of effective care. Instead, we recognized that other CMS initiatives (such as MIPS and APMs under the Quality Payment Program) may be better mechanisms to incentivize increased interoperability of health information systems than conditions of payment assigned to particular services under the PFS. We also anticipated that improved accuracy of payment for care management services and reduced administrative burden associated with billing for them would contribute to practitioners' capacity to invest in the best tools for managing the care of Medicare beneficiaries.

For Continuity of Care, we currently require the ability to obtain successive routine appointments “with the practitioner or a designated member of the care team,” while CPT only references successive routine appointments “with a designated member of the care team.” We do not believe there is any practical difference between these two phrases and therefore proposed to omit the words “practitioner or” from our requirement. The billing practitioner is a member of the CCM care team, so the CPT language already allows for successive routine appointments either with the billing practitioner or another appropriate member of the CCM care team.

Based on review of extensive public comment and stakeholder feedback, we had also come to believe that we should not require individuals providing the beneficiary with the required 24/7 access to care for urgent needs to have access to the care plan as a condition of CCM payment. As discussed above, we believed that in general, provision of effective after-hours care of the beneficiary would require access to the care plan, if not the full EHR. However, we have heard from rural and other practices that remote access to the care plan is not always necessary or possible because urgent care needs after-hours are often referred to a practitioner or care team member who established the care plan or is familiar with the beneficiary. In some instances, the care plan does not need to be available to address urgent patient needs after business hours. In addition, we have not required the use of any certified or non-certified health information technology in the provision of any other PFS services (including various other care management services). We were concerned that imposing EHR-related requirements at the service level as a condition of PFS payment could distort the relative valuation of services priced under the fee schedule. Therefore, we proposed to change the CCM service element to require timely electronic sharing of care plan information within and outside the billing practice, but not necessarily on a 24/7 basis, and to allow transmission of the care plan by fax.

We acknowledged that it is best for practitioners and providers to have access to care plan information any time they are providing services to beneficiaries who require CCM services. This proposal was not intended to undermine the significance of electronic communication methods other than fax transmission in providing effective, continuous care. On the contrary, we believed that fax transmission, while commonly used, is much less efficient and secure than other methods of communicating patient health information, and we encouraged practitioners to adopt and use electronic technologies other than fax for transmission and exchange of the CCM care plan. We continued to believe the best means of exchange of all relevant patient health information is through standardized electronic means. However, we recognized that other CMS initiatives (such as MIPS and APMs under the Quality Payment Program) may be better mechanisms to incentivize increased interoperability of health information systems than conditions of payment assigned to particular services under the PFS. We believed our proposal would still allow timely availability of health information within and outside the practice for purposes of providing CCM, and would simplify the rules governing provision of the service and improve access to the service. The proposed revisions would better align the service with appropriate CPT prefatory language, which may reduce unnecessary administrative complexity for practitioners in navigating the differences between CPT guidance and Medicare rules.

The CCM scope of service element Management of Care Transitions includes a requirement for the creation and electronic transmission and exchange of continuity of care documents referred to as “clinical summaries” (see Table 11 of the CY 2017 PFS proposed rule). We patterned our requirements regarding clinical summaries after the EHR Incentive Program requirement that an eligible professional who transitions their patient to another setting of care or provider of care, or refers their patient to another provider of care, should provide a summary care record for each transition of care or referral. This clinical summary includes demographics, the medication list, medication allergy list, problem list, and a number of other data elements if the practitioner knows them. As a condition of CCM payment, we required standardized content for clinical summaries (that they must be created/formatted according to certified EHR technology). For the exchange/transport function, we did not require the use of a specific tool or service to exchange/transmit clinical summaries, as long as they are transmitted electronically (this can include fax only when the receiving practitioner or provider can only receive by fax).

Based on review of extensive public comment and stakeholder feedback, we had come to believe that we should not require the use of any specific electronic technology in managing a beneficiary's care transitions as a condition of payment for CCM services. Instead, we proposed more simply to require the billing practitioner to create and exchange/transmit continuity of care document(s) timely with other practitioners and providers. To avoid confusion with the requirements of the EHR Incentive Programs, and since we would no longer require standardized content for the CCM continuity of care document(s), we would refer to them as continuity of care documents instead of clinical summaries. We would no longer specify how the billing practitioner must transport or exchange these document(s), as long as it is done timely and consistent with the Care Transitions Management scope of service element. We welcomed public input on how we should refer to these document(s), noting that CPT does not provide model language specific to CCM services. The proposed term “continuity of care document(s)” draws on CPT prefatory language for TCM services, which CPT provides may include “obtaining and reviewing the discharge information (for example, discharge summary, as available, or continuity of care document).”

Again, this proposal was not intended to undermine the significance of a standardized, electronic format and means of exchange (other than fax) of all relevant patient health information, for achieving timely, seamless care across settings especially after discharge from a facility. On the contrary, we believed that fax transmission, while commonly used, is much less efficient and secure than other methods of communicating patient health information, and we encourage practitioners to adopt and use electronic technologies other than fax for transmission and exchange of continuity of care documents in providing CCM services. We continued to believe the best means of exchange of all relevant patient health information is through standardized electronic means. However, as we discussed above regarding the CCM care plan, we have not applied similar requirements to other PFS services specifically (including various other care management services) and had concerns about how doing so may create disparities between these services and others under the PFS. We also recognized that other CMS initiatives (such as MIPS and APMs under the Quality Payment Program) may be better mechanisms to incentivize increased interoperability of health information systems than conditions of payment assigned to particular services under the PFS.

Comment: Most of the commenters supported our proposed revisions to the health IT use requirements for billing the CCM code. They shared CMS' goal of interoperability but believed the changes were necessary to improve CCM uptake. Some commenters favored hardship exceptions or rural or small practice exceptions instead of changes to the current requirements that would apply to all practitioners alike. Some commenters expressed particular concern about relaxing the current rules in instances where CCM outsourcing reduces clinical integration. These commenters noted that CCM is commonly outsourced to third party companies that provide remote care management services (including after hours) via telephone and online contact only, using staff who have no established relationship with the beneficiary or other members of the care team and have no interaction with the office staff and physicians other than electronic communication. These commenters were concerned that our proposed changes to the health IT requirements for CCM payment would result in little to no oversight or guidance of the third party, and recommended that CMS make the proposed changes cautiously. One of these commenters recommended in addition that CMS should seek to increase access to CCM services and reduce administrative burden by pursuing alignment between the provision of CCM and other programs and incentives, such as the Quality Payment Program. Other commenters recommended further reduction in payment rules, such as removing all requirements to use a certified EHR, or movement away from timed codes that require documentation in short time increments and disrupt workflow.

Response: We continue to believe that other Medicare initiatives and programs (such as MIPS and APMs under the Quality Payment Program) are better suited to advance use of interoperable health IT systems than establishing code-level conditions of payment, unique to CCM or other primary care or cognitive services. We also believe that a hardship, rural or small practice exception would greatly increase rather than decrease administrative complexity for practitioners and CMS, and CCM uptake has been relatively high among solo practices. We believe that reducing code-level conditions of payment is necessary to improve beneficiary access to appropriate CCM services. Therefore, we are finalizing revisions to the CCM scope of service elements as proposed.

However, we appreciate the commenters' feedback that relaxing the health IT use requirements may be of particular concern in situations where CCM is outsourced to a third party, reducing clinical integration. As we discuss in the section of this final rule on BHI services (section II.E.3.b), health IT holds significant promise for remote connectivity and interoperability that may assist and be useful (if not necessary) for reducing care fragmentation. However, we agree that remote provision of services by entities having only a loose association with the treating practitioner can detract from continuous, patient-centered care, whether or not those entities employ certified or other electronic technology. We will continue to consider the potential impacts of remote provision of CCM and similar types of services by third parties. We wish to emphasize for CCM, as we did for BHI services, that while the CCM codes do not explicitly count time spent by the billing practitioner, they are valued to include work performed by the billing practitioner, especially complex CCM. We emphasize that the practitioner billing for CCM must remain involved in ongoing oversight, management, collaboration and reassessment as appropriate to bill CCM services. If there is little oversight by the billing practitioner or a lack of clinical integration between a third party providing CCM and the billing practitioner, we do not believe that the CCM service elements are actually being furnished and therefore, in such cases, the practitioner should not bill for CCM.

Finally, we note that activities undertaken as part of participation in MIPS or an APM under the Quality Payment Program may support the ability of a practitioner to meet our final requirements for the continuity of care document(s) and the electronic care plan.

Comment: Several commenters recommended that we define the proposed term “timely” for the creation and transmission of care plan and care transitions health information. Several commenters believed that “timely” implies a time period of 30 to 90 days, or believed some third party vendors would interpret the term in this manner.

Response: Our proposal of the term “timely” originated from the use of this term in the CPT prefatory language for Care Management services, which includes, for example, “provide timely access and management for follow-up after an emergency department visit” and “timely access to clinical information.” We do not believe we should specify a timeframe, because it would vary for individual patients and CCM service elements, we are not aware of any clinical standards referencing specific times, and we are seeking to allow appropriate flexibility in how CCM is furnished. We note that dictionary meanings of the term “timely” include quickly; soon; promptly; occurring at a suitable time; done or occurring at a favorable or useful time; opportune. “Timely” does not necessarily imply speed, and means doing something at the most appropriate moment. Therefore we believe “timely” is an appropriate term to use to govern how quickly the health information in question is transmitted or available. We note that even the current requirements for use of specific electronic technology do not necessarily impact how quickly the health information in question is used to inform care, and addition of the word “timely” implies more regarding actual use of the information. We are monitoring CCM uptake and diffusion through claims analysis and are pursuing claims-based outcomes analyses, to help inform whether the service is being provided as intended and improving health outcomes. We believe these evaluation activities will help us assess moving forward whether health information is being shared or made available timely enough under our revised CCM payment policies.

As we stated in the CY 2017 proposed rule, the policy changes for CCM health IT use are not intended to undermine the importance of interoperability or electronic data exchange. These changes are driven by concerns that we have not applied similar requirements to other PFS services specifically, including various other care management services, and that such requirements create disparities between CCM services and other PFS services. We believe that other CMS initiatives may be better mechanisms to incentivize increased use and interoperability of health information systems than conditions of payment assigned to particular services under the PFS. We anticipate that these CCM policy changes will improve practitioners' capacity to invest in the best tools for managing the care of Medicare beneficiaries.

c. Beneficiary Receipt of Care Plan

We proposed to simplify the current requirement to provide the beneficiary with a written or electronic copy of the care plan, by instead adopting the CPT language specifying more simply that a copy of the care plan must be given to the patient or caregiver. While we believe beneficiaries should and must be provided a copy of the care plan, and that practitioners may choose to provide the care plan in hard copy or electronic form in accordance with patient preferences, we do not believe it is necessary to specify the format of the care plan that must be provided as a condition of CCM payment. Additionally, we recognize that there may be times that sharing the care plan with the caregiver (in a manner consistent with applicable privacy and security rules and regulations) may be appropriate.

Comment: The commenters who provided comments on this particular proposal were supportive of it. In particular, several commenters expressed appreciation for appropriate inclusion of caregivers.

Response: We thank the commenters for their support and are finalizing as proposed.

d. Beneficiary Consent

We continue to believe that obtaining advance beneficiary consent to receive CCM services is important to ensure the beneficiary is informed, educated about CCM services, and is aware of applicable cost sharing. We also believe that querying the beneficiary about whether another practitioner is already providing CCM services helps to reduce the potential for duplicate provision or billing of the services. However, we believe the consent process could be simplified, and that it should be left to the practitioner and the beneficiary to decide the best way to establish consent. Therefore, we proposed to continue to require billing practitioners to inform the beneficiary of the currently required information (that is, inform the beneficiary of the availability of CCM services; inform the beneficiary that only one practitioner can furnish and be paid for these services during a calendar month; and inform the beneficiary of the right to stop the CCM services at any time (effective at the end of the calendar month)). However, we proposed to specify that the practitioner could document in the beneficiary's medical record that this information was explained and note whether the beneficiary accepted or declined CCM services instead of obtaining a written agreement.

We also proposed to remove the language requiring beneficiary authorization for the electronic communication of his or her medical information with other treating providers as a condition of payment for CCM services, because under federal regulations that implement the Health Insurance Portability and Accountability Act (HIPAA) Privacy Rule (45 CFR 164.506), a covered entity is permitted to use or disclose protected health information for purposes of treatment without patient authorization. Moreover, if such disclosure is electronic, the HIPAA Security Rule requires secure transmission (45 CFR 164.312(e)). In previous regulations we have reminded practitioners that for all electronic sharing of beneficiary information in the provision of CCM services, HIPAA Privacy and Security Rule standards apply in the usual manner (79 FR 67728).

Comment: The commenters were largely supportive of our proposed policy changes. The commenters were supportive of verbal instead of written beneficiary consent if a clear requirement remains to transparently inform the beneficiary about the nature and benefit of the services, applicable cost sharing, and document that this information was conveyed; current written agreements qualify; and practitioners can elect to obtain written consent. Some commenters believed that obtaining written consent might be preferable as a means of resolving who is eligible for payment, if more than one practitioner bills. A few commenters suggested CMS require written educational materials about CCM, or conduct beneficiary outreach and education.

Response: We appreciate the commenters' support and recommendations. We are finalizing changes to the beneficiary consent requirements as proposed and clarifying that a clear requirement remains to transparently inform the beneficiary about the nature and benefit of the services, applicable cost sharing, and to document that this information was conveyed. The final beneficiary consent requirements do not affect any written agreements that are already in place for CCM services, and we note that practitioners can still elect to obtain written consent rather than verbal consent.

e. Documentation

We have heard from practitioners that the requirements to document certain information in a certified EHR format are redundant because the CCM billing rules already require documentation of core clinical information in a certified EHR format. Specifically, we already require structured recording of demographics, problems, medications and medication allergies, and the creation of a clinical summary record, using a qualifying certified EHR; and that a full list of problems, medications and medication allergies in the EHR must inform the care plan, care coordination and ongoing clinical care. Therefore, we proposed to no longer specify the use of a qualifying certified EHR to document communication to and from home- and community-based providers regarding the patient's psychosocial needs and functional deficits and to document beneficiary consent. We would continue to require documentation in the medical record of beneficiary consent (discussed above) and of communication to and from home- and community-based providers regarding the patient's psychosocial needs and functional deficits.

Comment: Many commenters were supportive of these proposals.

Response: We thank the commenters for their support and are finalizing changes to the documentation requirements as proposed. We continue to encourage practitioners to utilize health IT solutions for obtaining and documenting health information from sources external to their practice, noting that the 2015 edition of ONC certification criteria (see 80 FR 62601) includes criteria which specifically relate to obtaining information from non-clinical sources and the capture of structured data relating to social, psychological, and behavioral attributes.

f. Summary of Final CCM Policies

We are finalizing changes to the CCM scope of service elements discussed above that will apply for both complex and non-complex CCM services beginning in CY 2017. The final CY 2017 service elements for CCM are summarized in Table 11. We believe these changes will retain elements of the CCM service that are characteristic of the changes in medical practice toward advanced primary care, while eliminating redundancy, simplifying provision of the services, and improving access to the services. For payment of complex CCM services beginning in CY 2017, we are adopting the CPT code descriptors for CPT codes 99487 and 99489 as well as the service elements in Table 11. We are providing separate payment for complex CCM (CPT 99487, 99489) using the RUC-recommended payment inputs for those services. We may reconsider the role of health information technology in CCM service provision in future years. We anticipate that improved accuracy of payment for CCM services, and reduced administrative burden associated with billing CCM services, will contribute to practitioners' capacity to invest in the best tools for managing the care of Medicare beneficiaries.

5. Assessment and Care Planning for Patients with Cognitive Impairment (GPPP6)

For CY 2017 we proposed a G-code that would provide separate payment to recognize the work of a physician (or other appropriate billing practitioner) in assessing and creating a care plan for beneficiaries with cognitive impairment, such as from Alzheimer's disease or dementia, at any stage of impairment, G0505 (Cognition and functional assessment using standardized instruments with development of recorded care plan for the patient with cognitive impairment, history obtained from patient and/or caregiver, in office or other outpatient setting or home or domiciliary or rest home). We understand that a similar code was recently approved by the CPT Editorial Panel and is scheduled to be included in the CY 2018 CPT code set. We intended for G0505 to be a temporary code, perhaps for only one year, to be replaced by the CPT code in CT 2018. We will consider whether to adopt and establish relative value units for the new CPT code under our standard process, presumably for CY 2018.

We reviewed the list of service elements that were considered by the CPT Editorial Panel, and proposed the following as required service elements of G0505:

• Cognition-focused evaluation including a pertinent history and examination.
• Medical decision making of moderate or high complexity (defined by the E/M guidelines).
• Functional assessment (for example, Basic and Instrumental Activities of Daily Living), including decision-making capacity.
• Use of standardized instruments to stage dementia.
• Medication reconciliation and review for high-risk medications, if applicable.
• Evaluation for neuropsychiatric and behavioral symptoms, including depression, including use of standardized instrument(s).
• Evaluation of safety (for example, home), including motor vehicle operation, if applicable.
• Identification of caregiver(s), caregiver knowledge, caregiver needs, social supports, and the willingness of caregiver to take on caregiving tasks.
• Advance care planning and addressing palliative care needs, if applicable and consistent with beneficiary preference.
• Creation of a care plan, including initial plans to address any neuropsychiatric symptoms and referral to community resources as needed (for example, adult day programs, support groups); care plan shared with the patient and/or caregiver with initial education and support.

The proposed valuation of G0505 (discussed in section II.E.1) assumed that this code would include services that are personally performed by the physician (or other appropriate billing practitioner, such as a nurse practitioner or physician assistant) and would significantly overlap with services described by certain E/M visit codes, advance care planning services, and certain psychological or psychiatric service codes that are currently separately payable under the PFS. Accordingly, we proposed that G0505 must be furnished by the physician (or other appropriate billing practitioner) and could not be billed on the same date of service as CPT codes 90785 (Psytx complex interactive), 90791 (Psych diagnostic evaluation), 90792 (Psych diag eval w/med srvcs), 96103 (Psycho testing admin by comp), 96120 (Neuropsych tst admin w/comp), 96127 (Brief emotional/behav assmt), 99201-99215 (Office/outpatient visits new), 99324-99337 (Domicil/r-home visits new pat), 99341-99350 (Home visits new patient), 99366-99368 (Team conf w/pat by hc prof), 99497 (Advncd care plan 30 min), 99498 (Advncd care plan addl 30 min)), since these codes all reflect face-to-face services furnished by the physician or other billing practitioner for related separately billable services that overlap substantially with G0505. In addition, we proposed to prohibit billing of G0505 with other care planning services, such as care plan oversight services (CPT code 99374), home health care and hospice supervision (G0181, G0182), or our proposed add-on code for comprehensive assessment and care planning by the billing practitioner for patients requiring CCM services (GPPP7). We solicited comment on whether there are circumstances where multiple care planning codes could be furnished without significant overlap. We proposed to specify that G0505 may serve as a companion or primary E/M code to the prolonged service codes (those that are currently separately paid, and those we proposed to separately pay beginning in 2017), but were interested in public input on whether there is any overlap among these services. We solicited comment on how to best delineate the post-service work for G0505 from the work necessary to provide the prolonged services code.

We did not believe the services described by G0505 would significantly overlap with proposed or current medically necessary CCM services (CPT codes 99487, 99489, 99490); TCM services (CPT codes 99495, 99496); or the proposed behavioral health integration service codes (HCPCS codes GPPP1, GPPP2, GPPP3, GPPPX). Therefore, we proposed that G0505 could be billed on the same date-of-service or within the same service period as these codes (CPT codes 99487, 99489, 99490, 99495, 99496, and HCPCS codes GPPP1, GPPP2, GPPP3, and GPPPX). There may be overlap in the patient population eligible to receive these services and the population eligible to receive the services described by G0505, but we believed there would be sufficient differences in the nature and extent of the assessments, interventions and care planning, as well as the qualifications of individuals providing the services, to allow concurrent billing for services that are medically reasonable and necessary. We solicited public comment on potential overlap between G0505 and other codes currently paid under the PFS, as well as the other primary care/cognitive services addressed in this section of the final rule.

Comment: Many commenters were supportive of the proposal, including the provisions regarding scope of service elements, conditions of payment, and overlap with other services under the PFS.

Response: We thank commenters for their support of the proposed scope of service, conditions of payment, and overlap with other services under the PFS for G0505. We believe that by improving payment accuracy by paying separately for this service, practitioners will be able to accurately assess patients for cognitive impairment, particularly in early stages.

Comment: We received numerous comments stating that assessment and staging for dementia is very sensitive and should only be conducted by neuropsychologists, who would be unable to bill G0505. Commenters were concerned that untrained professionals conducting assessments for dementia would lead to errors in diagnosis and inappropriate treatment. Commenters encouraged CMS to not finalize this code and maintain the current coding for psychological and neuropsychological assessment or suggested that CMS remove the bullet points associated with medication management or medical decision making so that G0505 could be billed by psychologists.

Response: While we acknowledge and support the work of psychologists and neuropsychologists in the care of Medicare beneficiaries, we continue to believe that this code describes a distinct PFS service that may be reasonable and necessary in the diagnosis and treatment of a beneficiary's illness. We remind interested stakeholders that we routinely examine the valuation and coding for existing services under the potentially misvalued code initiative, and that there is a the process for public nomination of particular codes. If stakeholders have information to suggest that the current coding for neuropsychological and psychological testing is inaccurate, we welcome nominations under the established process.

Comment: A few commenters encouraged CMS to avoid adopting scope of service elements that are exhaustive as these may create barriers to utilization, while other commenters made the following recommendations regarding the scope of service provisions:

• Expand scope of service elements related to medication management.
• Include occupational therapy in the scope of service element pertaining to community resources.
• Rewrite “Creation of a care plan, including initial plans to address any neuropsychiatric symptoms and referral to community resources as needed (for example, adult day programs, support groups); care plan shared with the patient and/or caregiver with initial education and support” to include “identification of caregiver(s), caregiver knowledge, caregiver needs, social supports, and the willingness and availability of caregiver to voluntarily take on caregiving tasks.”
• Make sure that non-paid or informal caregivers are included in care planning and provide resources and support for care givers so as to improve care givers ability to provide care for the beneficiary.
• Require the inclusion of caregiver names in care plan and patients medical record, require that caregivers be assessed for stress and depressive symptoms, as well as care giver skill and education needs.
• State that consultations with the caregiver are permissible under HIPAA and that such conversations may be necessary in the development of a care plan.
• Specify that any advance care planning is consistent with beneficiary preference and addresses any palliative care needs of the patient, and include establishment of durable power of attorney.
• Clarify that diagnosis of dementia is not part of the scope of service by deleting “cognition focused evaluation including pertinent history” from the scope of service.
• Clarify that “functional assessment” is separate from decision making assessments, and that this is a non-legal assessment of competency.
• Stipulate that other decision makers should be identified.
• Consider deleting “use of standardized instruments to stage dementia” because the care plan is the most important aspect of the service and many standardized instruments are not very effective at staging.
• Clarify that the care plan address both medical and non-medical issues, and includes follow-up scheduling for monitoring and evaluation.
• Provide a copy of the written care plan to the patient.
• Refer to the care plan as a “person-centered care plan.”
• Include evaluation of medical problems including review of lab or imaging tests, review of co-morbidities, especially those which are dependent on self-care, evaluation the risk of falls and recommendations for fall prevention, evaluation of possible elder abuse, and documentation of financial issues, as part of the scope of service.

Response: We appreciate the information provided by commenters on the best practices associated with furnishing this service and would encourage stakeholders to adopt any or all of these scope of service provisions, such as the inclusion of caregivers in care planning. The scope of service for assessment and care planning service for patients with cognitive impairment does not prohibit stakeholders from adopting any additional scope of service provisions which may be beneficial for the treatment of the patient. However, we do not believe that the ability to fully furnish this service and establish an appropriate value for it is contingent on meeting such conditions. Therefore, we do not believe they should be added to the scope of service. We concur with commenters on the necessity of avoiding the imposition of overly burdensome restrictions within the scope of service.

Comment: Some commenters requested that CMS clarify that not all elements in the scope of service need to be provided by the billing practitioner and many can be provided by others incident to the billing practitioner's services. One commenter stated that there are circumstances where the best practitioner to provide a specific service element does not work in the same practice as the billing practitioner, and therefore the billing practitioner should be able to contract out for provision of some aspects, provided that the billing practitioner remain in oversight. Other commenters stated that CMS should make G0505 billable by other practitioners, such as occupational therapists, or community based entities.

Response: G0505 is a service that includes central elements, which must be performed by the billing practitioner subject to established E/M guidelines. Only those practitioners eligible to report E/M services should report this service. Outside of the specified elements, the regular incident-to rules apply consistent with other E/M services. We believe that physicians and eligible non-physician practitioners, such as a nurse practitioners and physician assistants should exclusively bill for this code.

Comment: Many commenters suggested that CMS expand HCPCS code G0505 or pay separately for similar services furnished to patients with other advanced or life threatening illnesses.

Response: We appreciate the comments on other conditions that could benefit from assessment and care planning and will consider these for future rulemaking. We are finalizing the G0505 code to pay separately for the assessment and care plan creation for beneficiaries with cognitive impairment, such as from Alzheimer's disease or dementia, at any stage of impairment.

Comment: Commenters provided many examples of how CMS could develop appropriate quality and outreach measures to ensure appropriate utilization of G0505. Commenters encouraged CMS to closely monitor use of G0505 for a few years following implementation, so as to ascertain whether patient eligibility is an issue in uptake for the code.

Response: We appreciate the information on quality and outreach measures. CMS is engaged in the use of measures to improve quality and access to care. CMS intends to monitor utilization and will consider how conditions of payment align with best practices and quality measures.

Comment: One commenter urged CMS to make the proposed coding and payment changes available to physicians in total cost of care models, such as ACOs and bundled payment programs.

Response: Our proposal relates only to payment for services under the Medicare PFS. We note that the codes and payment amounts that we finalize for services will be available for billing and payment under the PFS for CY 2017. In general, we do not address in this final rule, and instead defer to the policies regarding billing and payment for these services that are applicable within individual Center for Medicare Medicaid Innovation models and other programs. However, as our policies regarding payment for new primary care codes are applicable beginning in CY 2017, we note that models may need to update their policies to prevent potential duplication of payment between the PFS and the models. For example, where CCM services have been excluded from separate payment under existing models, newly established care management services (including complex CCM, psychiatric CoCM, and BHI) may likewise be excluded.

Comment: One commenter stated that many small practices do not have the infrastructure to support a multi-disciplinary team of practitioners and urged CMS to allow flexibility for solo and small group practices to share resources. The commenter also suggested that CMS offer a one-time incentive for practices to integrate service elements into workflow. Response: In general, the coding under the PFS is intended to describe services as they are furnished and are valued using typical resource costs. We appreciate the concern of commenters regarding access, and we are eager to hear from stakeholders regarding concerns related to access for these and other PFS services.

6. Improving Payment Accuracy for Care of People With Disabilities (GDDD1)

We estimate that about 7 percent of all Medicare beneficiaries have a potentially disabling mobility-related diagnosis (the Medicare-only prevalence is 5.5 percent and the prevalence for Medicare-Medicaid dual eligible beneficiaries is 11 percent), using 2010 Medicare (and for dual eligible beneficiaries, Medicaid) claims data.

When a beneficiary with a mobility-related disability goes to a physician or other practitioner's office for an E/M visit, the resources associated with providing the visit can exceed the resources required for the typical E/M visit. An E/M visit for a patient with a mobility-related disability can require more physician and clinical staff time to provide appropriate care because the patient may require skilled assistance throughout the visit to carefully move and adjust his/her body. Furthermore, an E/M visit for a patient with a mobility-related disability commonly requires specialized equipment such as a wheel chair accessible scale, floor and overhead lifts, a movable exam table, padded leg supports, a stretcher and transfer board. The current E/M visit payment rates, based on an assumption of “typical” resources involved in furnishing an E/M visit to a “typical” patient, do not accurately reflect these additional resources associated with furnishing appropriate care to many beneficiaries with mobility-related disabilities.

When furnishing E/M services to beneficiaries with mobility-related disabilities, practitioners face difficult choices in deciding whether to take the extra time necessary and invest in the required specialized equipment for these visits even though the payment rate for the service does not account for either expense; potentially providing less than optimal care for a beneficiary whose needs exceed the standard appointment block of time in the standard equipped exam room reflected in the current E/M visit payment rate; or declining to accept appointments altogether for beneficiaries who require additional time and specialized equipment.

Each of these scenarios is potentially problematic. The first two scenarios suggest that the quality of care for this beneficiary population might be compromised by assumptions under the PFS regarding relative resource costs in furnishing services to this population. The third scenario reflects an obvious access problem for these beneficiaries. To improve payment accuracy and help ameliorate potential disparity in access and quality for beneficiaries with mobility-related disabilities, we proposed to create a new add-on G-code, effective for CY 2017, to describe the additional services furnished in conjunction with E/M services to beneficiaries with disabilities that impair their mobility:

G0501: Resource-intensive services for patients for whom the use of specialized mobility-assistive technology (such as adjustable height chairs or tables, patient lifts, and adjustable padded leg supports) is medically necessary and used during the provision of an office/outpatient evaluation and management service visit (Add-on code, list separately in addition to primary procedure).

Effective January 1, 2017, we proposed that this add-on code could be billed with new and established patient office/outpatient E/M codes (CPT codes 99201 through 99205, and 99212 through 99215), as well as transitional care management codes (CPT codes 99495 and 99496), when the additional resources described by the code are medically necessary and used in the provision of care. In addition to seeking comment on this proposal, we are also sought comment on other HCPCS codes that may be appropriate base codes for this proposed add-on code, including those describing preventive visits and services. We reminded potential commenters that the rationale for this proposal is based in large part on the broad use and lack of granularity in coding for E/M services relative to other PFS services in conjunction with the additional resources used.

We received many thoughtful comments on this proposal and thank commenters for their input. Comments received are summarized below.

Comment: Most commenters agreed with the proposed rule's statement of disability disparities and discussed a variety challenges that individuals with disabilities face in accessing the health care system. Several of these commenters cited evidence of existing challenges for individuals with mobility-related disabilities, including a lack of physically accessible equipment within physician offices, barriers to communication, and a lack of existing tools to recognize, track, and consistently meet specialized needs. Commenters applauded CMS for offering a concrete proposal with significant funding to meaningfully address this problem and noted that 26 years after passage of the Americans with Disabilities Act, it is alarming that physical and communication barriers in physicians' and other health care professionals' offices still exist across the country. However, some commenters suggested that the root cause and scope of these issues are not well characterized, and suggested that CMS work with stakeholders to conduct additional studies and gain information as to the underlying reasons for barriers to access to care and lower quality scores on certain measures.

Generally, commenters noted that they appreciate CMS' efforts to address health disparities based on disability, and some then supported this proposal as a first step in providing medically necessary services to patients with disabilities, while others recommended that CMS not finalize the proposal and raised legal, access, and equity concerns.

Response: We agree with commenters that individuals with disabilities face additional barriers to access health care, an issue that contributes to widespread disparities in outcomes. We also agree with commenters that the underlying reasons for these disparities are multifaceted and can include payment challenges, physical accessibility and communication barriers, a lack of awareness among health care providers in assessing and fully addressing the needs and preferences of people with disabilities, and others issues. As a result of all these factors, individuals with disabilities can face challenges in scheduling appointments, and in finding and maintaining a primary care provider, an essential foundation for accessing the health system.

Although there was near universal agreement among commenters regarding problems in health care disparities and barriers to access among individuals with disabilities, there was disagreement about whether establishing payment for code G0501 as proposed was a good solution to help solve these problems. While we believe that improving the payment accuracy of physicians' services is necessary and appropriate, and can help to address the underlying access issues for individuals with disabilities, we also acknowledge that implementation of new or revised payments can result in unanticipated, and potentially undesirable, consequences. Before implementing payment for code G0501, we plan to further analyze and address the concerns raised by commenters. As such, we are not finalizing payment for code G0501 at this time. We appreciate commenters' insights, and our commitment to promoting better primary care for people with disabilities remains strong. Over the next 6 months we will engage with interested beneficiaries, advocates, and practitioners to continue to explore improvements in payment accuracy for care of people with disabilities. We intend to discuss this issue again in future rulemaking.

While we are not finalizing separate payment for code G0501 for CY 2017, we are including the code in the CY 2017 code set as G0501. The HCPCS code G0501 will not be payable under the Medicare PFS for CY 2017, though practitioners will be able to report the code, should they be inclined to do so.

a. Soliciting Comment on Other Coding Changes To Improve Payment Accuracy for Care of People With Disabilities

When furnishing care to a beneficiary with a mobility-related disability, the current E/M visit payment rates may not fully reflect the associated resource costs that are being incurred by practitioners. We recognize that there are other populations for which payment adjustment may be appropriate. Our proposal regarding beneficiaries with mobility-related disabilities reflected the discrete nature of the additional resource costs for this population, the clear lack of differentiation in resource costs regarding particular kinds of frequently-furnished services, and the broad recognition of access problems. We recognize that some physician practices may frequently furnish services to particular populations for which the relative resource costs are similarly systemically undervalued and we sought comment regarding other circumstances where these dynamics can be discretely observed.

Comment: Multiple commenters suggested additional coding changes to improve payment accuracy for services for people with disabilities. Several commenters requested that CMS broaden the scope of G0501 and the codes with which it may be billed, for example by allowing G0501 to be billed with preventive services, such as the Initial Preventive Physical Examination (IPPE) or “Welcome to Medicare Visit”, the Annual Wellness Visit, or other preventive services including those that have been assigned a grade of A or B by the United States Preventive Services Task Force. One commenter suggested that CMS also establish payment for a lower-level, lower payment add-on code for use with patients with a mobility-related disability that may not require the use of specialized equipment. Commenters also suggested that CMS establish certain forms of physician payment incentives, which might more effectively address the accessibility needs of individuals with disabilities and ultimately reduce healthcare disparities. Specifically, one commenter suggested CMS incentivize physicians to establish record-keeping to inquire into patients' accessibility and accommodation needs, record the needs of their patients, and take action to meet those needs over time.

Response: We thank commenters for their thoughtful responses. We reiterate our commitment to addressing disparities for individuals with disabilities and advancing health equity, and will continue to explore and revisit potential solutions for overcoming these significant challenges, including the appropriate changes in payment.

7. Regulation Text

Our current regulations in 42 CFR 410.26(b) provide for an exception to assign general supervision to CCM services (and similarly, for the non-face-to-face portion of TCM services), because these are generally non-face-to-face care management/care coordination services that would commonly be provided by clinical staff when the billing practitioner (who is also the supervising practitioner) is not physically present; and the CPT codes are comprised solely (or in significant part) of non-face-to-face services provided by clinical staff. A number of codes that we proposed to establish for separate payment in CY 2017 under our initiative to improve payment accuracy for primary care and care management are similar to CCM services, in that a critical element of the services is non-face-to-face care management/care coordination services provided by clinical staff or other qualified individuals when the billing practitioner may not be physically present. Accordingly, we proposed to amend 42 CFR 410.26(a)(3) and 410.26(b) to better define general supervision and to assign general supervision not only to CCM services and the non-face-to-face portion of TCM services, but also to proposed codes G0502, G0503, G0504, G0507, CPT code 99487, and CPT code 99489. Instead of adding each of these proposed codes assigned general supervision to the regulation text on an individual basis, we proposed to revise our regulation under 42 CFR 410.26(b)(1) to assign general supervision to the non-face-to-face portion of designated care management services, and we would designate the applicable services through notice and comment rulemaking.

We did not receive any public comments on our proposed regulation text. However we received a number of comments regarding a related proposal to require behavioral health care managers to be located on site. Also for psychiatric CoCM services (G0502, G0503 and G0504), we are finalizing a requirement that the behavioral health care manager is available to perform his or her duties face-to-face and non-face-to-face with the beneficiary. We address these issues at length in the BHI section of this final rule (section II.E.3). Since we are assigning general supervision to psychiatric CoCM behavioral health care manager services that may be provided face-to-face with the beneficiary, we are omitting the phrase “non-face-to-face portion of” in “the non-face-to-face portion of designated care management services.” Accordingly, the final amended regulation text in 42 CFR 410.26(b) assigns general supervision to “designated care management services” that we will designate through notice and comment rulemaking. The services that we are newly designating (finalizing) for general supervision in this final rule are G0502, G0503, G0504, G0507, CPT code 99487 and CPT code 99489. We had initially proposed adding a cross-reference to the existing definition of “general supervision” in current regulations at § 410.32(b)(3)(i), but to better describe general supervision in the context of these services, we are specifying at § 410.26(a)(3) that general supervision means the service is furnished under the physician's (or other practitioner's) overall direction and control, but the physician's (or other practitioner's) presence is not required during the performance of the service. At § 410.26(b)(5), we specify that, in general, services and supplies must be furnished under the direct supervision of the physician (or other practitioner). Designated care management services can be furnished under general supervision of the physician (or other practitioner) when these services or supplies are provided incident to the services of a physician (or other practitioner). The physician (or other practitioner) supervising the auxiliary personnel need not be the same physician (or other practitioner) who is treating the patient more broadly. However, only the supervising physician (or other practitioner) may bill Medicare for incident to services.

8. CCM Requirements for Rural Health Clinics (RHCs) and Federally Qualified Health Centers (FQHCs).

RHCs and FQHCs have been authorized to bill for CCM services since January 1, 2016, and are paid based on the Medicare PFS national average non-facility payment rate when CPT code 99490 is billed alone or with other payable services on a RHC or FQHC claim. The RHC and FQHC requirements for billing CCM services have generally followed the requirements for practitioners billing under the PFS, with some adaptations based on the RHC and FQHC payment methodologies.

To assure that CCM requirements for RHCs and FQHCs are not more burdensome than those for practitioners billing under the PFS, we proposed revisions for CCM services furnished by RHCs and FQHCs similar to the revisions proposed under the section above entitled, “Reducing Administrative Burden and Improving Payment Accuracy for Chronic Care Management (CCM) Services” for RHCs and FQHCs. Specifically, we proposed to:

• Require that CCM be initiated during an AWV, IPPE, or comprehensive E/M visit only for new patients or patients not seen within one year. This would replace the requirement that CCM could only be initiated during an AWV, IPPE, or comprehensive E/M visit where CCM services were discussed.
• Require 24/7 access to a RHC or FQHC practitioner or auxiliary personnel with a means to make contact with a RHC or FQHC practitioner to address urgent health care needs regardless of the time of day or day of week. This would replace the requirement that CCM services be available 24/7 with health care practitioners in the RHC or FQHC who have access to the patient's electronic care plan to address his or her urgent chronic care needs, regardless of the time of day or day of the week.
• Require timely electronic sharing of care plan information within and outside the RHC or FQHC, but not necessarily on a 24/7 basis, and expands the circumstances under which transmission of the care plan by fax is allowed. This would replace the requirement that the electronic care plan be available on a 24/7 basis to all practitioners within the RHC or FQHC whose time counts towards the time requirement for the practice to bill the CCM code, and removes the restriction on allowing the care plan to be faxed only when the receiving practitioner or provider can only receive clinical summaries by fax.
• Require that in managing care transitions, the RHC or FQHC creates, exchanges, and transmits continuity of care document(s) in a timely manner with other practitioners and providers. This would replace the requirements that clinical summaries must be created and formatted according to certified EHR technology, and the requirement for electronic exchange of clinical summaries by a means other than fax.
• Require that a copy of the care plan be given to the patient or caregiver. This would remove the description of the format (written or electronic) and allows the care plan to be provided to the caregiver when appropriate (and in a manner consistent with applicable privacy and security rules and regulations).
• Require that the RHC or FQHC practitioner documents in the beneficiary's medical record that all the elements of beneficiary consent (for example, that the beneficiary was informed of the availability of CCM services; only one practitioner can furnish and be paid for these services during a calendar month; the beneficiary may stop the CCM services at any time, effective at the end of the calendar month, etc.) were provided, and whether the beneficiary accepted or declined CCM services. This would replace the requirement that RHCs and FQHCs obtain a written agreement that these elements were discussed, and removes the requirement that the beneficiary provide authorization for the electronic communication of his or her medical information with other treating providers as a condition of payment for CCM services.
• Require that communication to and from home- and community-based providers regarding the patient's psychosocial needs and functional deficits be documented in the patient's medical record. This would replace the requirement to document this patient health information in a certified EHR format.

We noted that we did not propose an additional payment adjustment for patients who require extensive assessment and care planning as part of the initiating visit, as payments for RHC and FQHC services are not adjusted for length or complexity of the visit.

We stated that we believe these proposed changes would keep the CCM requirements for RHCs and FQHCs consistent with the CCM requirements for practitioners billing under the PFS, simplify the provision of CCM services by RHCs and FQHCs, and improve access to these services without compromising quality of care, beneficiary privacy, or advance notice and consent.

We received 31 comments on the proposed revisions to the CCM requirements for RHCs and FQHCs. The following is a summary of the comments we received:

Comment: Commenters stated that they support CMS's efforts to ensure that CCM requirements for RHCs and FQHCs are not more burdensome than those for practitioners billing under the Medicare PFS.

Response: We appreciate the support of the commenters.

Comment: One commenter sought clarification on the requirements for initiating CCM with patients that have been seen in the RHC within the past year. The commenter asked if CCM could be initiated if the patient had any type of visit within the past year, or if the visit within the past year had to be an AWV, IPPE, or comprehensive E/M visit.

Response: To initiate CCM with a patient that has been seen in the RHC or FQHC within the past year, an AWV, IPPE, or comprehensive E/M visit must have taken place within the past year in the RHC or FQHC that is billing for the CCM service. No other type of visit would meet the requirement for initiating CCM services.

Comment: A few commenters were concerned that RHCs and FQHCs were charging beneficiaries for coinsurance for non-face-to-face services, and recommended that the copayment be waived or that CMS pursue waivers of cost-sharing for care coordination codes. One of these commenters stated that patients are often unwilling to pay the patient share of the CCM services since rural providers often have already been providing similar services without additional cost to the patients.

Response: As previously stated, we do not have the authority to waive the copayment requirements for CCM services. While many practitioners, including those in rural areas, have always provided some care management services, we believe that payment for CCM services will enable many RHCs and FQHCs to furnish comprehensive and systematic care coordination services that were previously unavailable or only sporadically offered.

Comment: A commenter asked for clarification on how claims for patients in RHCs and FQHCs with pre-existing care management plans should be handled, and suggested that CMS permit claims for services for these patients.

Response: We are not entirely clear what this commenter is suggesting. RHCs and FQHCs that bill for CCM services must develop a comprehensive care plan that includes all the elements previously described and also listed in Table 11. When all the requirements for furnishing CCM services are met, including the development of the comprehensive care plan, the RHC or FQHC would submit a claim for CCM payment using CPT code 99490. Only the time spent furnishing CCM services after CCM is initiated with the patient is counted toward the minimum 20 minutes required for CCM billing. There is no additional payment for a pre-existing care plan, and if a comprehensive care plan that meets the CCM requirements was developed before the initiation of CCM services, the time spent developing the plan would not be counted toward the 20 minute minimum requirement.

Comment: A few commenters requested clarification on whether RHCs and FQHCs could bill the new CCM codes for either complex CCM services (CPT 99487 and 99489) or the separately billable comprehensive CCM assessment and care planning (G0506).

Response: As we noted in the proposed rule, we did not propose to adopt codes to provide for an additional payment for patients who require extensive assessment or care planning because payments for RHC and FQHC services are not adjusted for the length or complexity of the visit. Therefore, the codes identified by the commenters are not separately billable by an RHC or FQHC.

Comment: A few commenters recommended that CMS allow RHCs and FQHCs to bill for the new CCM codes, and to allow safety net providers to bill for preventive services in addition to the all-inclusive rate for RHCs and the PPS rate for FQHCs. The commenters stated that the payment structure for RHCs and FQHCs are a disincentive to provide preventative services in addition to E/M services at the same visit.

Response: RHCs and FQHCs are paid for CCM services when CPT code 99490 is billed either alone or with other payable services on a RHC or FQHC claim. The RHC and FQHC payment structures and payment for preventive services is outside the scope of this final rule.

Comment: Several commenters recommended that CMS provide separate payment for psychiatric collaborative care management services furnished in RHCs and FQHCs, including CPT codes G0502, G0503, G0504 and G0507. The commenters stated that allowing RHCs and FQHCs to bill for these services will ensure that their patients who have been diagnosed with a mental health or substance use disorder have access to high-quality care tailored to their individual condition and circumstances.

Response: To be eligible for CCM services, a Medicare beneficiary must have two or more chronic conditions that are expected to last at least 12 months (or until the death of the patient), and place the patient at significant risk of death, acute exacerbation/decompensation, or functional decline. While CCM is typically associated with primary care conditions, patient eligibility is determined by the RHC or FQHC practitioner, and mental health conditions are not excluded. We invite comments on whether an additional code specifically for mental health conditions is necessary for RHCs and FQHCs that want to include beneficiaries with mental health conditions in their CCM services.

After considering the comments, we are finalizing as proposed the revisions to the requirements for CCM services furnished by RHCs and FQHCs.

F. Improving Payment Accuracy for Services: Diabetes Self-Management Training (DSMT)

Section 1861(s)(2)(S) of the Act specifies that medical and other health services include DSMT services as defined in section 1861(qq) of the Act. DSMT services are intended to educate beneficiaries in the successful self-management of diabetes. DSMT includes, as applicable, instructions in self-monitoring of blood glucose; education about diet and exercise; an insulin treatment plan developed specifically for the patient who is insulin-dependent; and motivation for patients to use the new skills for self-management (see 42 CFR 410.144(a)(5)). DSMT services are reported under HCPCS codes G0108 (Diabetes outpatient self-management training services, individual, per 30 minutes) and G0109 (Diabetes outpatient self- management training services, group session (2 or more), per 30 minutes). The benefit, as specified at 42 CFR 410.141, consists of 1 hour of individual and 9 hours of group training unless special circumstances warrant more individual training or no group session is available within 2 months of the date the training is ordered.

Section 1861(qq) of the Act specifies that DMST services are furnished by a certified provider, defined as a physician or other individual or entity that also provides, in addition to DSMT, other items or services for which payment may be made under Medicare. The physician, individual or entity that furnishes the training also must meet certain quality standards. The physician, individual or entity can meet standards established by us or standards originally established by the National Diabetes Advisory Board and subsequently revised by organizations who participated in their establishment, or can be recognized by an organization that represents individuals with diabetes as meeting standards for furnishing the services.

We require that all those who furnish DSMT services be accredited as meeting quality standards by a CMS-approved national accreditation organization (NAO). In accordance with § 410.144, a CMS-approved NAO may accredit an individual, physician or entity to meet one of three sets of DSMT quality standards: CMS quality standards; the National Standards for Diabetes Self-Management Education Programs (National Standards); or the standards of an NAO that represents individuals with diabetes that meet or exceed our quality standards. Currently, we recognize the American Diabetes Association and the American Association of Diabetes Educators as approved NAOs, both of whom follow National Standards. Medicare payment for outpatient DSMT services is made in accordance with 42 CFR 414.63.

An article titled “Use of Medicare's Diabetes Self-Management Training Benefit” was published in Health Education Behavior on January 23, 2015. The article noted that only 5 percent of Medicare beneficiaries with newly diagnosed diabetes used DSMT services. The article recommended that future research identify barriers to DSMT access.

In the CY 2017 PFS proposed rule (81 FR 45215), we identified issues that the DSMT community had brought to our attention which may contribute to the low utilization of these services, and indicated that we plan to address and clarify those issues through Medicare program instructions as appropriate. We also solicited public comment as to other access barriers—including whether Medicare payment for these services is accurate—to help us identify and address them. We appreciate the many comments regarding many issues in response to our solicitation.

Comment: Many commenters stated that the payment rates were too low but did not suggest specific changes in the inputs used to develop payment rates under the PFS for particular services (specifically, work RVUs and direct PE inputs). We also received additional comments identifying multiple other possible barriers to access. These commenters' recommendations primarily addressed issues related to regulatory and statutory DSMT requirements, such as: (a) Expanding of the definition of diabetes to include hemoglobin A1C as one of the criteria for diagnosing diabetes; (b) modifying the definition of certified provider to include the certified diabetes educator (CDE) to permit them to bill for DSMT; (c) allowing physicians and NPPs, other than the one treating the beneficiary's diabetes, as required by regulation, to order DSMT services; and, (d) eliminating the copays and deductible for DSMT services.

Response: We appreciate the comments received and will consider changes in valuation of these services and other regulatory issues raised by commenters for future rulemaking. We also appreciate commenters' feedback on several subregulatory guidelines and other operational issues that we will consider addressing outside of rulemaking.

G. Target for Relative Value Adjustments for Misvalued Services

Section 1848(c)(2)(O) of the Act establishes an annual target for reductions in PFS expenditures resulting from adjustments to relative values of misvalued codes. Under section 1848(c)(2)(O)(ii) of the Act, if the estimated net reduction in expenditures for a year as a result of adjustments to the relative values for misvalued codes is equal to or greater than the target for that year, reduced expenditures attributable to such adjustments shall be redistributed in a budget-neutral manner within the PFS in accordance with the existing budget neutrality requirement under section 1848(c)(2)(B)(ii)(II) of the Act. The provision also specifies that the amount by which such reduced expenditures exceeds the target for a given year shall be treated as a net reduction in expenditures for the succeeding year, for purposes of determining whether the target has been met for that subsequent year. Section 1848(c)(2)(O)(iv) of the Act defines a target recapture amount as the difference between the target for the year and the estimated net reduction in expenditures under the PFS resulting from adjustments to RVUs for misvalued codes. Section 1848(c)(2)(O)(iii) of the Act specifies that, if the estimated net reduction in PFS expenditures for the year is less than the target for the year, an amount equal to the target recapture amount shall not be taken into account when applying the budget neutrality requirements specified in section 1848(c)(2)(B)(ii)(II) of the Act. Under section 1848(c)(2)(O)(v) of the Act, the target that applies to calendar years (CYs) 2017 and 2018 is calculated as 0.5 percent of the estimated amount of expenditures under the PFS for the year.

In CY 2016 PFS rulemaking, we proposed and finalized a methodology to implement this statutory provision.

Because the annual target is calculated by measuring changes from one year to the next, for CY 2016, we considered how to account for changes in values that are best measured over 3 years, instead of 2 years. As we described in the CY 2016 final rule with comment period (80 FR 70932), our general valuation process for potentially misvalued, new, and revised codes was to establish values on an interim final basis for a year in the PFS final rule with comment period. Then, during the 60-day period following the publication of the final rule with comment period, we would accept public comment about those valuations. In the final rule with comment period for the subsequent year, we would consider and respond to public comments received on the interim final values, and make any appropriate adjustments to values based on those comments. Under that process for revaluing new, revised, and misvalued codes, we believe the overall change in valuation for many codes would best be measured across values for 3 years: between the original value in the first year; the interim final value in the second year; and the finalized value in the third year. However, the target calculation for a year would only be comparing changes in RVUs between 2 years and not among 3 years, so the contribution of a particular change towards the target for any single year would be measured against only the preceding year without regard to the overall change that takes place over 3 years.

For recent years, interim final values for misvalued codes (year 2) have generally reflected reductions relative to original values (year 1), and for most codes, the interim final values (year 2) are maintained and finalized (year 3). However, when values for particular codes have changed between the interim final (year 2) and final values (year 3) based on public comment, the general tendency has been that codes increase in the final value (year 3) relative to the interim final value (year 2), even in cases where the final value (year 3) represents a decrease from the original value (year 1). Therefore, for these codes, the year 2 changes compared to year 1 would risk over-representing the overall reduction, while the year 3 to year 2 changes would represent an increase in value. We noted that if there were similar targets in every PFS year, and a similar number of misvalued code changes made on an interim final basis, the incongruence in measuring what is really a 3-year change in 2-year increments might not be particularly problematic since each year's calculation would presumably include a similar number of codes measured between years 1 and 2 and years 2 and 3.

However, including changes that take place over 3 years generated challenges in calculating the target for CY 2016. Because there was no target for CY 2015, any reductions that occurred on an interim final basis for CY 2015 were not counted toward achievement of a target. If we had then included any upward adjustments made to these codes based on public comment as “misvalued code” changes for CY 2016, we would effectively be counting the service-level increases for 2016 (year 3) relative to 2015 (year 2) against achievement of the target without any consideration to the service-level changes relative to 2014 (year 1), even in cases where the overall change in valuation was negative.

Therefore, we proposed and finalized the decision to exclude code-level input changes for CY 2015 interim final values from the calculation of the CY 2016 misvalued code target since the misvalued change occurred over multiple years, including years not applicable to the misvalued code target provision.

For the CY 2017 final rule, we will be finalizing values (year 3) for codes that were interim final in CY 2016 (year 2). Unlike codes that were interim final for CY 2015, the codes that are interim final for CY 2016 were included as misvalued codes and will fall within the range of years for which the misvalued code target provision applies. Thus, overall changes in values for these codes would be measured in the target across 3 full years: The original value in the first year (CY 2015); the interim final value in the second year (CY 2016); and the finalized value in the third year (CY 2017). The changes in valuation for these CY 2016 interim final codes were previously measured and counted towards the target during their initial change in valuation between years 1 and 2.

As such, we proposed to include changes in values of the CY 2016 interim final codes toward the CY 2017 misvalued code target. We believe that this is consistent with the approach that we finalized in the CY 2016 PFS final rule with comment period. The changes in values of CY 2015 interim final codes were not counted towards the misvalued code target in CY 2016 since the valuation change occurred over multiple years, including years not applicable to the misvalued code target provision. However, both of the changes in valuation for the CY 2016 interim final codes, from year 1 to year 2 (CY 2015 to CY 2016) and from year 2 to year 3 (CY 2016 to CY 2017), have taken place during years that occur within the misvalued code target provision. We therefore believe that any adjustments made to these codes based on public comment should be considered towards the achievement of the target for CY 2017, just as any changes in valuation for these same CY 2016 interim final codes previously counted towards the achievement of the target for CY 2016.

We solicited comments regarding this proposal. We also reminded commenters that we revised our process for revaluing new, revised and misvalued codes so that we will be proposing and finalizing values for most of the misvalued codes during a single calendar year. After this year, there will be far fewer instances of interim final codes and changes that are best measured over 3 years.

We refer readers to the regulatory impact analysis section of this final rule for the net reduction in expenditures relative to the 0.5 percent target for CY 2017, and the resulting adjustment required to be made to the conversion factor. Additionally, we refer readers to the public use file that provides a comprehensive description of how the target is calculated, as well as the estimated impact by code family on the CMS Web site under the supporting data files for the CY 2017 PFS final rule at http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/index.html.

The following is summary of the comments we received regarding the target for relative value adjustments for misvalued services.

Comment: Several commenters expressed support for the CMS estimate that there would be no target recapture amount by which to reduce payments made under the PFS in CY 2017.

Response: We appreciate the comments. We remind stakeholders that the final determination of the target recapture amount is based on finalized RVUs for the relevant codes. We refer readers to the regulatory impact analysis section of this final rule for the net reduction in expenditures relative to the 0.5 percent target for CY 2017, and the resulting adjustment that is required to be made to the conversion factor.

Comment: One commenter urged CMS to broaden its approach to counting misvalued code payment adjustments in the final rule. The commenter stated that CMS was taking a narrow approach to the misvalued code target.

Response: We finalized our methodology for calculating the estimated net reduction relative to the misvalued code target in the CY 2016 final rule with comment period (80 FR 70921-70927). For CY 2017, we proposed a modification to that methodology that only addressed how changes to interim final codes would be addressed when both first and second year changes could be counted towards a misvalued code target since CY 2017 is the first year for that circumstance. We did not make a proposal on the more general issue of the methodology used to calculate the net reductions for the misvalued code target, which, as noted above, was finalized in the CY 2016 PFS final rule with comment period.

We did not receive any public comments on our proposal to include changes in values of the CY 2016 interim final codes toward the CY 2017 misvalued code target.

After consideration of comments received, we are finalizing our proposal to count any adjustments to interim final codes towards the misvalued code target when both first and second year changes can be counted towards a misvalued codes target.

H. Phase-In of Significant RVU Reductions

Section 1848(c)(7) of the Act specifies that for services that are not new or revised codes, if the total RVUs for a service for a year would otherwise be decreased by an estimated 20 percent or more as compared to the total RVUs for the previous year, the applicable adjustments in work, PE, and MP RVUs shall be phased in over a 2-year period. In the CY 2016 PFS rulemaking, we proposed and finalized a methodology to implement this statutory provision. To determine which services are described by new or revised codes for purposes of the phase-in provision, we apply the phase-in to all services that are described by the same, unrevised code in both the current and update year, and exclude codes that describe different services in the current and update year.

Because the phase-in of significant reductions in RVUs falls within the budget neutrality requirements specified in section 1848(c)(2)(B)(ii)(II) of the Act, we estimate the total RVUs for a service prior to the budget-neutrality redistributions that result from implementing phase-in values. In implementing the phase-in, we consider a 19 percent reduction as the maximum 1-year reduction for any service not described by a new or revised code. This approach limits the year one reduction for the service to the maximum allowed amount (that is, 19 percent), and then phases in the remainder of the reduction.

The statute provides that the applicable adjustments in work, PE, and MP RVUs shall be phased in over a 2-year period when the RVU reduction for a code for a year is estimated to be equal to or greater than 20 percent. Since CY 2016 was the first year in which we applied the phase-in transition, CY 2017 will be the first year in which a single code could be subject to RVU reductions greater than 20 percent for 2 consecutive years.

Under our finalized policy, the only codes that are not subject to the phase-in are those that are new or revised, which we defined as those services that are not described by the same, unrevised code in both the current and update year, or by the same codes that describe different services in the current and update year. Since CY 2016 was the first year for which the phase-in provision applied, we did not address how we would handle codes with values that had been partially phased in during the first year, but that have a remaining phase-in reduction of 20 percent or greater.

The significant majority of codes with reductions in RVUs that are greater than 20 percent in year one would not be likely to meet the 20 percent threshold in a consecutive year. However, in a few cases, significant changes (for example, in the input costs included in the valuation of a service) could produce reductions of 20 percent or greater in consecutive years.

As stated in the CY 2017 PFS proposed rule, we believed that a consistent methodology regarding the phase-in transition should be applied to these cases. We proposed to reconsider in each year, for all codes that are not new or revised codes and including codes that were assigned a phase-in value in the previous year, whether the total RVUs for the service would otherwise be decreased by an estimated 20 percent or more as compared to the total RVUs for the previous year. Under this proposed policy, the 19 percent reduction in total RVUs would continue to be the maximum one-year reduction for all codes (except those considered new and revised), including those codes with phase-in values in the previous year. In other words, for purposes of the 20 percent threshold, every service is evaluated anew each year, and any applicable phase-in is limited to a decrease of 19 percent. For example, if we were to adopt a 50 percent reduction in total RVUs for an individual service, the reduction in any particular year would be limited to a decrease of 19 percent in total RVUs. Because we do not set rates 2 years in advance, the phase-in transition would continue to apply until the year-to-year reduction for a given code does not meet the 20 percent threshold. We solicited comments regarding this proposal.

The list of codes subject to the phase-in and the associated proposed RVUs that result from this methodology is available on the CMS Web site under downloads for the CY 2017 PFS final rule at http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/PFS-Federal-Regulation-Notices.html.

The following is summary of the comments we received regarding the phase-in of significant RVU reductions.

Comment: Many commenters supported the proposal that a 19 percent reduction in total RVUs would continue to be the maximum one-year reduction for all codes that are not new or revised. These commenters urged CMS to finalize the proposal.

Response: We appreciate the support from the commenters.

Comment: Several commenters suggested that CMS should extend the threshold for triggering the phase-in provision, by using a lower single-year maximum reduction (such as 10 percent), at a rate different than what the statute stipulates. The commenters stated that a lower threshold would provide a greater safeguard against payment cuts and disruption of services.

Response: Section 1848(c)(7) of the Act requires the phase-in if the total RVUs for a service for a year would otherwise be decreased by an estimated 20 percent or larger. We do not believe that we have the statutory authority to establish a different threshold value for when the phase-in applies.

Comment: One commenter objected to CMS' decision to exclude from the phase-in codes with a reduction of 20 percent or more that fall within a family with significant coding revisions. The commenter requested that CMS reconsider this policy.

Response: We understand the commenters' concerns. In the CY 2016 final rule with comment period (80 FR 70927-70931), we finalized a policy to identify services that are not subject to the phase-in because they are new or revised codes. As we wrote at the time, we excluded as new and revised codes those codes that describe a different set of services in the update year when compared to the current year by virtue of changes in other, related codes, or codes that are part of a family with significant coding revisions. Significant coding revisions within a family of codes can change the relationships among codes to the extent that it changes the way that all services in the group are reported, even if some individual codes retain the same number or, in some cases, the same descriptor. We continue to believe that this is the most accurate methodology to use in identifying new and revised codes for the purposes of the phase-in transition. We also note that we did not make a proposal to change how we identify services to which the phase-in does not apply.

Comment: A commenter requested that CMS apply the phase-in policy to services in the PFS with year-to-year reductions of 20 percent or more in payment amount due to the statutory cap that requires payment for the technical component (TC) of certain imaging services furnished in the office setting to be made the lesser of the PFS or OPPS rates. The commenter stated that this application would capture the spirit of the phase-in legislation in dampening the impact of significant payment reductions on a year to year basis.

Response: Section 1848(c)(7) of the Act requires the phase-in of reductions of 20 percent or more in the total RVUs for individual services. The OPPS cap, required under section 1848(b)(4)(A) of the Act, specifies that if the PFS payment rate for the TC of certain imaging services exceeds the OPPS payment amount for the services, the OPPS payment amount must be substituted for the PFS TC payment amount. The OPPS cap refers to, and requires substitution of, payment rates for individual imaging services, and not a reduction in the total RVUs for those services. As such, services that are subject to the OPPS cap are not subject to the phase-in on that basis.

Comment: One commenter opposed the phase-in proposal. The commenter stated that the proposal twisted a plain reading of the law to effectively extend the phase-in period well beyond the 2 years prescribed by the statute. The commenter questioned why Medicare beneficiaries should have to pay a higher fee for overvalued services when identified as such, and pointed out that in the budget-neutral environment of the fee schedule, the proposal would delay the benefit of these RVU reductions to the rest of the services listed in the PFS.

Response: We appreciate the concerns raised by the commenter. As we have addressed over several rulemaking cycles, we are concerned about the impact of misvalued services in creating distortions in relativity across the fee schedule. However, we have already finalized through notice and comment rulemaking and continue to believe that limiting reductions to 19 percent as the maximum 1-year decrease for all codes (except those considered new and revised) is the best and most fair way to apply the phase-in. Additionally, because we do not set rates 2 years in advance, we believe there are significant obstacles to implementing an alternative methodology. For example, codes may be reviewed multiple times in a short period of time, and may have further decreases in total RVUs for a subsequent year due to a variety of reasons in addition to any change inputs from the initial year phase-in. These might include supply and equipment price updates in non-reviewed years, significant changes in specialty mix of practitioners reporting the service, or changes in other PFS ratesetting policies which could lead to several consecutive years of RVU reductions. In any such cases, it would be impractical to identify with certainty what portion of reductions in code values are due to input changes established in a prior year versus input or policy changes from the current year. We also note that all of these circumstances are relatively rare since it is unusual for changes in code inputs to result in reductions of greater than 40 percent. Therefore, while we appreciate the importance of improving payment accuracy as soon as can be practicable for the reasons stated by the commenter, we also believe that, on balance, the best and most fair approach to implementing the required phase-in of RVU reductions over multiple years is to re-examine eligible codes for the phase-in on an annual basis, in conjunction with our annual ratesetting.

After consideration of comments received, we are finalizing the policy as proposed.

I. Geographic Practice Cost Indices (GPCIs)

1. Background

Section 1848(e)(1)(A) of the Act requires us to develop separate Geographic Practice Cost Indices (GPCIs) to measure relative cost differences among localities compared to the national average for each of the three fee schedule components (that is, work, PE, and malpractice (MP)). The PFS localities are discussed in section II.E.3. of this final rule. Although the statute requires that the PE and MP GPCIs reflect the full relative cost differences, section 1848(e)(1)(A)(iii) of the Act requires that the work GPCIs reflect only one-quarter of the relative cost differences compared to the national average. In addition, section 1848(e)(1)(G) of the Act sets a permanent 1.5 work GPCI floor for services furnished in Alaska beginning January 1, 2009, and section 1848(e)(1)(I) of the Act sets a permanent 1.0 PE GPCI floor for services furnished in frontier states (as defined in section 1848(e)(1)(I) of the Act) beginning January 1, 2011. Additionally, section 1848(e)(1)(E) of the Act provided for a 1.0 floor for the work GPCIs, which was set to expire on March 31, 2015. Section 201 of the MACRA amended the statute to extend the 1.0 floor for the work GPCIs through CY 2017 (that is, for services furnished no later than December 31, 2017).

Section 1848(e)(1)(C) of the Act requires us to review and, if necessary, adjust the GPCIs at least every 3 years. Section 1848(e)(1)(C) of the Act requires that, if more than 1 year has elapsed since the date of the last previous GPCI adjustment, the adjustment to be applied in the first year of the next adjustment shall be half of the adjustment that otherwise would be made. Therefore, since the previous GPCI update was implemented in CY 2014 and CY 2015, we proposed to phase in 1/2 of the latest GPCI adjustment in CY 2017.

We have completed a review of the GPCIs and proposed new GPCIs in this final rule. We also calculate a geographic adjustment factor (GAF) for each PFS locality. The GAFs are a weighted composite of each area's work, PE and malpractice expense GPCIs using the national GPCI cost share weights. While we do not actually use GAFs in computing the fee schedule payment for a specific service, they are useful in comparing overall areas costs and payments. The actual effect on payment for any actual service would deviate from the GAF to the extent that the proportions of work, PE and MP RVUs for the service differ from those of the GAF.

As noted above, section 201 of the MACRA extended the 1.0 work GPCI floor for services furnished through December 31, 2017. Therefore, the proposed CY 2017 work GPCIs and summarized GAFs reflect the 1.0 work floor. Additionally, as required by sections 1848(e)(1)(G) and 1848(e)(1)(I) of the Act, the 1.5 work GPCI floor for Alaska and the 1.0 PE GPCI floor for frontier states are permanent, and therefore, applicable in CY 2017. See Addenda D and E to this final rule for the CY 2017 GPCIs and summarized GAFs available on the CMS Web site under the supporting documents section of the CY 2017 PFS final rule located at https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/index.html.

2. GPCI Update

The proposed updated GPCI values were calculated by a contractor. There are three GPCIs (work, PE, and MP), and all GPCIs are calculated relative to the national average for each measure. Additionally, each of the three GPCIs relies on its own data source(s) and methodology for calculating its value as described below. Additional information on the CY 2017 GPCI update may be found in our contractor's draft report, “Draft Report on the CY 2017 Update of the Geographic Practice Cost Index for the Medicare Physician Fee Schedule,” which is available on our Web site. It is located under the supporting documents section for the CY 2017 PFS final rule located at https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/index.html.

a. Work GPCIs

The work GPCIs are designed to reflect the relative costs of physician labor by Medicare PFS locality. As required by statute, the work GPCI reflects one quarter of the relative wage differences for each locality compared to the national average.

To calculate the work GPCIs, we use wage data for seven professional specialty occupation categories, adjusted to reflect one-quarter of the relative cost differences for each locality compared to the national average, as a proxy for physicians' wages. Physicians' wages are not included in the occupation categories used in calculating the work GPCI because Medicare payments are a key determinant of physicians' earnings. Including physician wage data in calculating the work GPCIs would potentially introduce some circularity to the adjustment since Medicare payments typically contribute to or influence physician wages. That is, including physicians' wages in the physician work GPCIs would, in effect, make the indices, to some extent, dependent upon Medicare payments.

The work GPCI updates in CYs 2001, 2003, 2005, and 2008 were based on professional earnings data from the 2000 Census. However, for the CY 2011 GPCI update (75 FR 73252), the 2000 data were outdated and wage and earnings data were not available from the more recent Census because the “long form” was discontinued. Therefore, we used the median hourly earnings from the 2006 through 2008 Bureau of Labor Statistics (BLS) Occupational Employment Statistics (OES) wage data as a replacement for the 2000 Census data. The BLS OES data meet several criteria that we consider to be important for selecting a data source for purposes of calculating the GPCIs. For example, the BLS OES wage and employment data are derived from a large sample size of approximately 200,000 establishments of varying sizes nationwide from every metropolitan area and can be easily accessible to the public at no cost. Additionally, the BLS OES is updated regularly, and includes a comprehensive set of occupations and industries (for example, 800 occupations in 450 industries). For the CY 2014 GPCI update, we used updated BLS OES data (2009 through 2011) as a replacement for the 2006 through 2008 data to compute the work GPCIs.

Because of its reliability, public availability, level of detail, and national scope, we believe the BLS OES data continue to be the most appropriate source of wage and employment data for use in calculating the work GPCIs (and as discussed in section II.E.2.b the employee wage component and purchased services component of the PE GPCI). Therefore, for the proposed CY 2017 GPCI update, we used updated BLS OES data (2011 through 2014) as a replacement for the 2009 through 2011 data to compute the work GPCIs.

b. Practice Expense GPCIs

The PE GPCIs are designed to measure the relative cost difference in the mix of goods and services comprising practice expenses (not including malpractice expenses) among the PFS localities as compared to the national average of these costs. Whereas the physician work GPCIs (and as discussed later in this section, the MP GPCIs) are comprised of a single index, the PE GPCIs are comprised of four component indices (employee wages; purchased services; office rent; and equipment, supplies and other miscellaneous expenses). The employee wage index component measures geographic variation in the cost of the kinds of skilled and unskilled labor that would be directly employed by a physician practice. Although the employee wage index adjusts for geographic variation in the cost of labor employed directly by physician practices, it does not account for geographic variation in the cost of services that typically would be purchased from other entities, such as law firms, accounting firms, information technology consultants, building service managers, or any other third-party vendor. The purchased services index component of the PE GPCI (which is a separate index from employee wages) measures geographic variation in the cost of contracted services that physician practices would typically buy. (For more information on the development of the purchased service index, we refer readers to the CY 2012 PFS final rule with comment period (76 FR 73084 through 73085)). The office rent index component of the PE GPCI measures relative geographic variation in the cost of typical physician office rents. For the medical equipment, supplies, and miscellaneous expenses component, we believe there is a national market for these items such that there is not significant geographic variation in costs. Therefore, the equipment, supplies and other miscellaneous expense cost index component of the PE GPCI is given a value of 1.000 for each PFS locality.

For the previous update to the GPCIs (implemented in CY 2014) we used 2009 through 2011 BLS OES data to calculate the employee wage and purchased services indices for the PE GPCI. As discussed in section II.E.2.a., because of its reliability, public availability, level of detail, and national scope, we continue to believe the BLS OES is the most appropriate data source for collecting wage and employment data. Therefore, in calculating the proposed CY 2017 GPCI update, we used updated BLS OES data (2011 through 2014) as a replacement for the 2009 through 2011 data for purposes of calculating the employee wage component and purchased service index component of the PE GPCI.

c. Malpractice Expense (MP) GPCIs

The MP GPCIs measure the relative cost differences among PFS localities for the purchase of professional liability insurance (PLI). The MP GPCIs are calculated based on insurer rate filings of premium data for $1 million to $3 million mature claims-made policies (policies for claims made rather than services furnished during the policy term). For the CY 2014 GPCI update (seventh update) we used 2011 and 2012 malpractice premium data (78 FR 74382). The proposed CY 2017 MP GPCI update reflects 2014 and 2015 premium data. Additionally, the proposed CY 2017 MP GPCI update reflects several proposed technical refinements to the MP GPCI methodology as discussed later in section 5.

d. GPCI Cost Share Weights

For CY 2017 GPCIs, we proposed to continue to use the current cost share weights for determining the PE GPCI values and locality GAFs. We refer readers to the CY 2014 PFS final rule with comment period (78 FR 74382 through 74383), for further discussion regarding the 2006-based MEI cost share weights revised in CY 2014 that were also finalized for use in the CY 2014 (seventh) GPCI update.

The GPCI cost share weights for CY 2017 are displayed in Table 12.

e. PE GPCI Floor for Frontier States

Section 10324(c) of the Affordable Care Act added a new subparagraph (I) under section 1848(e)(1) of the Act to establish a 1.0 PE GPCI floor for physicians' services furnished in frontier states effective January 1, 2011. In accordance with section 1848(e)(1)(I) of the Act, beginning in CY 2011, we applied a 1.0 PE GPCI floor for physicians' services furnished in states determined to be frontier states. In general, a frontier state is one in which at least 50 percent of the counties are “frontier counties,” which are those that have a population per square mile of less than 6. For more information on the criteria used to define a frontier state, we refer readers to the FY 2011 Inpatient Prospective Payment System (IPPS) final rule (75 FR 50160 through 50161). There are no changes in the states identified as Frontier States for the CY 2017 final rule. The qualifying states are: Montana, Wyoming, North Dakota, South Dakota, and Nevada. In accordance with statute, we would apply a 1.0 PE GPCI floor for these states in CY 2017.

f. Proposed GPCI Update

As explained above in the background section, the periodic review and adjustment of GPCIs is mandated by section 1848(e)(1)(C) of the Act. At each update, the proposed GPCIs are published in the PFS proposed rule to provide an opportunity for public comment and further revisions in response to comments prior to implementation. As discussed later in this section, we are finalizing the GPCIs as proposed (except where we correct technical errors). The final CY 2017 updated GPCIs for the first and second year of the 2-year transition, along with the GAFs, are displayed in Addenda D and E to this final rule available on our Web site under the supporting documents section of the CY 2017 PFS final rule Web page at https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/index.html.

3. Payment Locality Discussion

a. Background

The current PFS locality structure was developed and implemented in 1997. There are currently 89 total PFS localities; 34 localities are statewide areas (that is, only one locality for the entire state). There are 52 localities in the other 16 states, with 10 states having 2 localities, 2 states having 3 localities, 1 state having 4 localities, and 3 states having 5 or more localities. The combined District of Columbia, Maryland, and Virginia suburbs; Puerto Rico; and the Virgin Islands are the remaining three localities of the total of 89 localities. The development of the current locality structure is described in detail in the CY 1997 PFS final rule (61 FR 34615) and the subsequent final rule with comment period (61 FR 59494). We note that the localities generally represent a grouping of one or more constituent counties.

Prior to 1992, Medicare payments for physicians' services were made under the reasonable charge system. Payments were based on the charging patterns of physicians. This resulted in large differences in payment for physicians' services among types of services, geographic payment areas, and physician specialties. Recognizing this, the Congress replaced the reasonable charge system with the Medicare PFS in the Omnibus Budget Reconciliation Act (OBRA) of 1989, and the PFS went into effect January 1, 1992. Payments under the PFS are based on the relative resources involved with furnishing services, and are adjusted to account for geographic variations in resource costs as measured by the GPCIs.

Payment localities originally were established under the reasonable charge system by local Medicare carriers based on their knowledge of local physician charging patterns and economic conditions. These localities changed little between the inception of Medicare in 1967 and the beginning of the PFS in 1992. Shortly after the PFS took effect, we undertook a study in 1994 that culminated in a comprehensive locality revision that was implemented in 1997 (61 FR 59494).

The revised locality structure reduced the number of localities from 210 to the current 89, and the number of statewide localities increased from 22 to 34. The revised localities were based on locality resource cost differences as reflected by the GPCIs. For a full discussion of the methodology, see the CY 1997 PFS final rule with comment period (61 FR 59494). The current 89 fee schedule areas are defined alternatively by state boundaries (for example, Wisconsin), metropolitan areas (for example, Metropolitan St. Louis, MO), portions of a metropolitan area (for example, Manhattan), or rest-of-state areas that exclude metropolitan areas (for example, Rest of Missouri). This locality configuration is used to calculate the GPCIs that are in turn used to calculate payments for physicians' services under the PFS.

As stated in the CY 2011 PFS final rule with comment period (75 FR 73261), changes to the PFS locality structure would generally result in changes that are budget neutral within a state. For many years, before making any locality changes, we have sought consensus from among the professionals whose payments would be affected. In recent years, we have also considered more comprehensive changes to locality configuration. In 2008, we issued a draft comprehensive report detailing four different locality configuration options (https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/downloads/ReviewOfAltGPCIs.pdf). We refer readers to the CY 2014 PFS final rule with comment period for further discussion regarding that report, as well as a discussion about the Institute of Medicine's empirical study of the Medicare GAFs established under sections 1848(e) (PFS GPCI) and 1886(d)(3)(E) (IPPS wage index) of the Act.

The following is a summary of the comments we received regarding the proposed CY 2017 GPCI update.

Comment: A few commenters including a major specialty society expressed support for using more current data in calculating all three GPCIs.

Response: We thank the commenters for their support.

Comment: One commenter expressed support for the elimination of all geographic adjustment factors under the PFS, except those designed to achieve a specific public policy goal, such as to encourage physicians to practice in underserved areas. Another commenter opposed any decrease in the GPCI.

Response: As previously discussed, section 1848(e)(1)(A) of the Act requires us to develop separate GPCIs to measure resource cost differences among localities compared to the national average for each of the three GPCI components, and section 1848(e)(1)(C) of the Act requires us to review and, if necessary, adjust the GPCIs at least every 3 years; and based on new data, GPCI values may increase or decrease.

Comment: A few commenters expressed concern regarding payment for rural localities and recommended that CMS monitor how the GPCI calculation changes affect the sustainability of health services in rural communities. One commenter requested that CMS consider the ongoing data issues regarding the GPCIs raised by stakeholders in the Midwest, and establish 1.0 work and PE GPCI values for Wisconsin and Iowa.

Response: As discussed previously in this section, we are required to update the GPCIs at least every 3 years to reflect the relative cost differences of operating a medical practice in each locality compared to the national average costs. Additionally, as previously discussed in this section, sections 1848(e)(1)(G) and 1848(e)(1)(I) of the Act established the permanent 1.5 work GPCI floor for Alaska and the permanent 1.0 PE GPCI floor for frontier States. We do not otherwise have the authority to establish similar GPCI floors or other policies that do not take into consideration the differences in physicians' resource costs among localities.

Comment: One commenter supported the continuation of the 1.0 PE GPCI floor for frontier states.

Response: As previously discussed, beginning January 1, 2011 section 1848(e)(1)(I) of the Act set a permanent 1.0 PE GPCI floor for services furnished in frontier states (as defined in section 1848(e)(1)(I) of the Act).

Comment: Several commenters stated their objection to the use of residential rents as a proxy for physician office space costs, and stated that CMS should collect commercial rent data and use it either as the basis for measuring geographic differences in physician office rents or, if this is not feasible, use it to validate the residential rents as a proxy. A few commenters requested that CMS provide a specific explanation on the barriers to gaining better commercial rent data.

Response: Because Medicare is a national program, and section 1848(e)(1)(A) of the Act requires us to establish GPCIs to measure relative cost differences among localities compared to the national average, we believe it is important to use the best data that is available on a nationwide basis, that is regularly updated, and retains consistency area-to-area, year-to-year. Since there is currently no national data source available for physician office or other comparable commercial rents, we continue to use county-level residential rent data from the American Community Survey (ACS) as a proxy for the relative cost differences in commercial office rents. The ACS is administered by the United States Census Bureau, which is a leading source of national, robust, quality, publicly available data. We agree that a commercial data source for office rent that provided for adequate representation of urban and rural areas nationally would be preferable to a residential rent proxy. We have previously discussed in the CY 2005, CY 2008, and CY 2011 (69 FR 66262, 72 FR 66376, and 75 FR 73257, respectively) final rules that we recognize that apartment rents may not be a perfect proxy for physician office rent. We have also conducted exhaustive searches for reliable commercial rent data sources that are publicly available in the past and have not found any reliable data that meets our accuracy needs, and we continue to conduct such searches. With regards to suggestion that CMS should collect commercial rent data, we note that we discussed this issue in the CY 2012 PFS final rule with comment period (76 FR 73088) and stated with reference to surveying physicians directly to gather data to compute office rent, we note that the development and implementation of a survey could take several years. Additionally, we have historically not sought direct survey data from physicians related to the GPCI to avoid issues of circularity and self-reporting bias. In the CY 2011 PFS final rule with comment period (75 FR 73259), we solicited public comments regarding the benefits of utilizing physician cost reports to potentially achieve greater precision in measuring the relative cost difference among Medicare localities. We also asked for comments regarding the administrative burden of requiring physicians to routinely complete these cost reports and whether this should be mandatory for physicians' practices. We did not receive any feedback related to that comment solicitation during the open public comment period for the CY 2011 final rule with comment period. We continue to have concerns that physician cost reports could be prohibitively expensive, and as well about the administrative burden this approach would place on physician's office staff. We reiterate that the GPCIs are not an absolute measure of practice costs, rather they are a measure of the relative cost differences for each of the three GPCI components. The U.S. Census Bureau is a federal agency that specializes in data collection, accuracy, and reliability, and we continue to believe that where such a publicly available resource exists that can provide useful data to assess geographic cost differences in office rent, even though it is a proxy for the exact data we seek, that we should utilize that available resource. Therefore, given its national representation, reliability, high response rate and frequent updates, we continue to believe the ACS residential rent data is the most appropriate data source available at this time for the purposes of calculating the rent index of the PE GPCI.

Comment: One commenter stated that it objects to the 8 percent weight that the rent expense category has been given by CMS in calculating the PE GPCI, and stated that office rent should be given a much larger weight to more accurately reflect its impact on physician practice expenses, and CMS should commit resources to update this data since it is based on 10-year old data from the 2006 AMA Physician Practice Information Survey (AMA PPIS).

Response: We would like to clarify that the office rent expense category has a cost share weight of 10.223 percent, not 8 percent as indicated by the commenter. The MEI cost share weights were derived from data collected by the AMA on the AMA PPIS. CMS has previously stated that we believe the AMA PPIS is a reliable data source, however the AMA PPIS is not an ongoing data source that is regularly published. We continued to use the AMA PPIS data source in the CY 2014 revisions to the MEI which have not been further updated since, and therefore, as discussed above, the 2006-based MEI cost share weights finalized for use in the CY 2014 (seventh) GPCI update, were proposed for the CY 2017 (eighth) GPCI update. The AMA is no longer conducting the AMA PPIS survey, and CMS' Office of the Actuary continues to look into viable options for updating the MEI cost share weights going forward. In the CY 2014 PFS final rule with comment period (78 FR 74275), we stated that we continue to investigate possible data sources to use for the purpose of rebasing the MEI in the future.

Comment: A few commenters expressed concern with the use of unrelated proxy data for physician wages in geographic adjustment. The commenters expressed concern about GPCI proxy inputs that result in downward payment adjustments, which they believe do not reflect the actual cost of physician practices. The commenters stated that better data exist for measuring the real physician compensation rates, such as recruitment compensation surveys and wages for physicians employed at federally qualified health centers. The commenters also stated that MedPAC studies have confirmed that the data sources currently relied upon for geographic adjustment bear no correlation to physician earnings. One commenter also stated that CMS has acknowledged that the proxies utilized for the purposes of geographic adjustment have never been validated and there never has been a new data source utilized in the twenty years since the fee schedule was implemented. The commenters urged CMS to undertake the necessary studies to identify reference occupations that will accurately reflect the higher input costs of rural physician earnings, and implement the resulting corrections to the geographic adjustment of the fee schedule as soon as possible.

Response: We appreciate the comments regarding the professional occupations used to determine the relative cost differences in physician earnings for purposes of calculating the work GPCI. In consideration of the ongoing concerns regarding the reference occupations and other proxy data used to calculate the GPCIs, we also note that in the past we received comments suggesting the use of survey data to determine GPCI values, and stated that we would continue to consider the possibility of establishing a physician cost report and requiring a sufficiently large sample of physicians in each locality to report data on actual costs incurred. However we also stated that we believed that a physician cost report could take years to develop and implement, and could be prohibitively expensive (75 FR 73259). We solicited public comment regarding the potential benefits to be gained from establishing a physician cost report and whether this approach is appropriate to achieve potentially greater precision in measuring the relative cost differences in physicians' practices among PFS localities. We also solicited public comments on the potential administrative burden of requiring physicians to routinely complete and submit a cost report. We did not receive any feedback specifically related to that comment solicitation (76 FR 73088). As noted previously in this section, physicians' wages are not included in the occupation categories (reference occupations) used in calculating the work GPCI because Medicare payments are a key determinant of physicians' earnings. We have long maintained that including physician wage data in calculating the work GPCIs would potentially introduce some circularity to the adjustment since Medicare payments typically contribute to physician wages. In other words, including physicians' wages in the physician work GPCI would, in effect, make the indices, to some extent, dependent upon Medicare payments, which in turn are impacted by the indices. We reiterate that the work GPCI is not an absolute measure of physician earnings; rather it is a measure of the relative wage differences for each locality as compared to the national average; additionally, the work GPCI reflects only one quarter of those relative wage differences consistent with the statutory requirement as discussed previously in this section.

Comment: We received a few comments on the PFS locality structure that were not within the scope of the CY 2017 proposed rule. For example, several commenters requested that Prince William and Loudoun counties in Virginia be changed from the Rest of Virginia locality into the DC + MD/VA Suburbs locality. Another commenter stated that it believes large cuts to rural and rest-of-State areas should be avoided or minimized, but locality boundaries with large payment differences should not be in the middle of urban areas, because they create payment cliffs where payment can change by up to eight percent if an office location is moved across a street or down a block. The commenter stated that CMS should act quickly to create locality definitions that are not constrained by county boundaries, and advocated implementing locality definition changes based on Metropolitan Statistical Areas.

Response: We appreciate the suggestions for revisions to the PFS locality structure. As discussed above, we did not propose changes to the PFS locality structure; we note that the update to the California Fee Schedule Areas discussed later in this section is the result of a statutory requirement. Additionally, we would like to note that, absent statutory provisions like those that pertain to California, changes to the locality configuration within a state would lead to significant redistributions in payments within that state. It has been our practice, and we have stated in previous rulemaking (72 FR 38139, and 73 FR 38513), that we have not considered making changes to localities without the support of a State medical association(s) to demonstrate consensus for the change among the professionals whose payments would be affected (with some increasing and some decreasing). Also, we would like to clarify that, just as the localities under the Fee Schedule areas used in the PFS are comprised of one or more constituent counties, so are Metropolitan Statistical Areas. Therefore the concept of a payment cliff between neighboring counties as described by the commenter would not necessarily be mitigated by a change from PFS fee schedule areas to Metropolitan Statistical Areas.

After consideration of the public comments received regarding the proposed CY 2017 GPCI data update, we are finalizing as proposed.

b. California Locality Update to the Fee Schedule Areas Used for Payment Under Section 220(h) of the Protecting Access to Medicare Act

(1) General Discussion and Legislative Change

Section 220(h) of the PAMA added a new section 1848(e)(6) to the Act, that modifies the fee schedule areas used for payment purposes in California beginning in CY 2017.

Currently, the fee schedule areas used for payment in California are based on the revised locality structure that was implemented in 1997 as previously discussed. Beginning in CY 2017, section 1848(e)(6)(A)(i) of the Act requires that the fee schedule areas used for payment in California must be Metropolitan Statistical Areas (MSAs) as defined by the Office of Management and Budget (OMB) as of December 31 of the previous year; and section 1848(e)(6)(A)(ii) of the Act requires that all areas not located in an MSA must be treated as a single rest-of-state fee schedule area. The resulting modifications to California's locality structure would increase its number of localities from 9 under the current locality structure to 27 under the MSA-based locality structure.

However, section 1848(e)(6)(D) of the Act defines transition areas as the fee schedule areas for 2013 that were the rest-of-state locality, and locality 3, which was comprised of Marin County, Napa County, and Solano County. Section 1848(e)(6)(B) of the Act specifies that the GPCI values used for payment in a transition area are to be phased in over 6 years, from 2017 through 2021, using a weighted sum of the GPCIs calculated under the new MSA-based locality structure and the GPCIs calculated under the current PFS locality structure. That is, the GPCI values applicable for these areas during this transition period are a blend of what the GPCI values would have been under the current locality structure, and what the GPCI values would be under the MSA-based locality structure. For example, in the first year, CY 2017, the applicable GPCI values for counties that were previously in rest-of-state or locality 3 and are now in MSAs are a blend of 1/6 of the GPCI value calculated for the year under the MSA-based locality structure, and 5/6 of the GPCI value calculated for the year under the current locality structure. The proportions shift by 1/6 in each subsequent year so that, by CY 2021, the applicable GPCI values for counties within transition areas are a blend of 5/6 of the GPCI value for the year under the MSA-based locality structure, and 1/6 of the GPCI value for the year under the current locality structure. Beginning in CY 2022, the applicable GPCI values for counties in transition areas are the values calculated under the new MSA-based locality structure. For the sake of clarity, we reiterate that this incremental phase-in is only applicable to those counties that are in transition areas that are now in MSAs, which are only some of the counties in the 2013 California rest-of state locality and locality 3.

Additionally, section 1848(e)(6)(C) of the Act establishes a hold harmless for transition areas beginning with CY 2017 whereby the applicable GPCI values for a year under the new MSA-based locality structure may not be less than what they would have been for the year under the current locality structure. There are a total of 58 counties in California, 50 of which are in transition areas as defined in section 1848(e)(6)(D) of the Act. Therefore, 50 counties in California are subject to the hold harmless provision. The other 8 counties, which are metropolitan counties that are not defined as transition areas, are not held harmless for the impact of the new MSA-based locality structure, and may therefore potentially experience slight decreases in their GPCI values as a result of the provisions in section 1848(e)(6) of the Act, insofar as the locality in which they are located now newly includes data from adjacent counties that decreases their GPCI values relative to those that would have applied had the new data not been incorporated. Therefore, the GPCIs for these eight counties under the MSA-based locality structure may be less than they would have been under the current GPCI structure. The eight counties that are not within transition areas are: Orange; Los Angeles; Alameda; Contra Costa; San Francisco; San Mateo; Santa Clara; and Ventura counties.

We emphasize that while transition areas are held harmless from the impact of the GPCI changes using the new MSA-based locality structure, because we proposed other updates for CY 2017 as part of the eighth GPCI update, including the use of updated data, transition areas would still be subject to impacts resulting from those other updates. Table 13 illustrates using GAFs, for CY 2017, the isolated impact of the MSA-based locality changes and hold-harmless for transition areas required by section 1848(e)(6) of the Act, the impact of the use of updated data for GPCIs, and the combined impact of both of these changes.

Additionally, for the purposes of calculating budget neutrality and consistent with the PFS budget neutrality requirements as specified under section 1848(c)(2)(B)(ii)(II) of the Act, we proposed to start by calculating the national GPCIs as if the current localities are still applicable nationwide; then for the purposes of payment in California, we override the GPCI values with the values that are applicable for California consistent with the requirements of section 1848(e)(6) of the Act. This approach is consistent with the implementation of the GPCI floor provisions that have previously been implemented—that is, as an after-the-fact adjustment that is implemented for purposes of payment after both the GPCIs and PFS budget neutrality have already been calculated.

(2) Operational Considerations

As discussed above, under section 1848(e)(6) of the Act, counties that were previously in the rest-of-state locality or locality 3 and are now in MSAs would have their GPCI values under the new MSA-based locality structure phased in gradually, in increments of one-sixth over 6 years. Section 1848(e)(1)(C) of the Act requires that, if more than 1 year has elapsed since the date of the last previous GPCI adjustment, the adjustment to be applied in the first year of the next adjustment shall be 1/2 of the adjustment that otherwise would be made. While section 1848(e)(6)(B) of the Act establishes a blended phase-in for the MSA-based GPCI values, it does not explicitly state whether or how that provision is to be reconciled with the requirement at section 1848(e)(1)(C) of the Act. We believe that since section 1848(e)(6)(A) of the Act requires that we must make the change to MSA-based fee schedule areas for California GPCIs notwithstanding the preceding provisions of section 1848(e) of the Act, and subject to the succeeding provisions of section 1848(e)(6) of the Act, that applying the two-year phase-in specified by the preceding provisions simultaneously with the six-year phase-in would undermine the incremental 6-year phase-in specified in section 1848(e)(6)(B) of the Act. Therefore, we proposed that the requirement at section 1848(e)(1)(C) of the Act to phase in 1/2 of the adjustment in year 1 of the GPCI update would not apply to counties that were previously in the rest-of-state or locality 3 and are now in MSAs, and therefore, are subject to the blended phase-in as described above. Since section 1848(e)(6)(B) of the Act provides for a gradual phase in of the GPCI values under the new MSA-based locality structure, specifically in one-sixth increments over 6 years, if we were to also apply the requirement to phase in 1/2 of the adjustment in year 1 of the GPCI update then the first year increment would effectively be one-twelfth. We note that this issue is only of concern if more than 1 year has elapsed since the previous GPCI update, and would only be applicable through CY 2021 since, beginning in CY 2022, the GPCI values for such areas in an MSA would be fully based on the values calculated under the new MSA-based locality structure for California.

As previously stated, the resulting modifications to California's locality structure increase its number of localities from 9 under the current locality structure to 27 under the MSA-based locality structure. However, both the current localities and the MSA-based localities are comprised of various component counties, and in some localities only some of the component counties are subject to the blended phase-in and hold harmless provisions required by section 1848(e)(6)(B) and (C) of the Act. Therefore, the application of these provisions may produce differing GPCI values among counties within the same fee schedule area under the MSA-based locality structure. For example, the MSA-based San Jose-Sunnyvale-Santa Clara locality, is comprised of 2 constituent counties—San Benito County, and Santa Clara County. San Benito County is in a transition area (2013 rest-of-state), while Santa Clara County is not. Hence, although the counties are in the same MSA, the requirements of section 1848(e)(6)(B) and (C) of the Act may produce differing GPCI values for each county. To address this issue, we proposed to assign a unique locality number to the counties that would be impacted in the aforementioned manner. As a result, although the modifications to California's locality structure increase the number of localities from 9 under the current locality structure to 27 under the MSA-based locality structure, for purposes of payment, the actual number of localities under the MSA-based locality structure would be 32 to account for instances where unique locality numbers are needed as described above. Additionally, while the fee schedule area names are consistent with the MSAs designated by OMB, we proposed to maintain 2-digit locality numbers to correspond to the existing fee schedule areas. Pursuant to the implementation of the new MSA-based locality structure for California, the total number of PFS localities would increase from 89 to 112. Table 14 displays the current fee schedule areas in California, and Table 15 displays the MSA-based fee schedule areas in California required by section 1848(e)(6) of the Act. Additional information on the California locality update may be found in our contractor's draft report, “Draft Report on the CY 2017 Update of the Geographic Practice Cost Index for the Medicare Physician Fee Schedule,” which is available on the CMS Web site. It is located under the supporting documents section of the CY 2017 PFS final rule located at https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/index.html.

We received few comments regarding the California locality update to the fee schedule areas used for payment under section 220(h) of PAMA.

Comment: One commenter stated that it supports the proposed California payment locality implementation plan. The commenter stated that based on its analysis the calculations are accurate except for a few errors. Specifically, the commenter stated that the CY 2016 GAFs for 3 fee schedule areas [Los Angeles-Long Beach-Anaheim (Orange County), San Francisco-Oakland-Hayward (Alameda/Contra Costa County), and San Francisco-Oakland-Hayward (San Francisco County)] in Table 13 of the proposed rule (81 FR 46221 through 46222) were incorrect. The commenter also requested that all of the 2016 GAFs in the table be reported to three decimal places to avoid confusion with rounding. Additionally, the commenter indicated that Sierra County in California was missing from the CY 2017 Proposed GPCI County Data File in the CY 2017 Proposed Rule GPCI Public Use Files available on our Web site under the supporting documents section of the CY 2017 PFS proposed rule Web page at https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/index.html.

Response: We thank the commenter for its support of our proposed California payment locality implementation plan. With regard to the errors noted by the commenter, we thank the commenter for bringing this issue to our attention. We agree that the CY 2016 GAFs for Los Angeles-Long Beach-Anaheim (Orange County), San Francisco-Oakland-Hayward (Alameda/Contra Costa County), and San Francisco-Oakland-Hayward (San Francisco County) were incorrect in Table 13 of the proposed rule, and have been corrected in Table 13 in this final rule. We have also updated all of the 2016 GAFs in Table 13 to reflect 3 decimal places as to avoid confusion with rounding as requested. Additionally, we note that while the GAFs for these 3 fee schedule areas were incorrect in Table 13 of the proposed rule, the GAF values were correct in Addendum D to the proposed rule available on our Web site under the supporting documents section of the CY 2017 PFS Proposed Rule Web page at https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/index.html. Moreover, GAF values are an analysis tool, and are not used to determine service level payment. Additionally, we note Sierra County was omitted from the CY 2017 Proposed GPCI County Data File because we removed counties with 0 total RVUs in 2014. However, for the final rule we have revised the file to include all counties, even those with 0 total RVUs in 2014. The updated file can be viewed in the CY 2017 Final GPCI County Data File in the CY 2017 Final Rule GPCI Public use files available on our Web site.

Comment: One commenter requested that CMS implement the California locality update requirement in a manner that does not require the Medicare Administrative Contractor (MAC) for California to make changes to the enrollment process for physician groups in California or changes in the way that physician groups submit claims to the MAC.

Response: While we note that there are several internal administrative burdens that result from the implementation of the California locality update, we do not believe there should be related burden on practitioners, and California practitioners will continue to follow the existing process for submitting claims.

After consideration of the public comments received regarding the proposed California payment locality implementation plan, we are finalizing as proposed.

4. Update to the Methodology for Calculating GPCIs in the U.S. Territories

In calculating GPCIs within U.S. states, we use county-level wage data from the Bureau of Labor Statistics (BLS) Occupational Employment Statistics Survey (OES), county-level residential rent data from the American Community Survey (ACS), and malpractice insurance premium data from state departments of insurance. In calculating GPCIs for the U.S. territories, we currently use three distinct methodologies—one for Puerto Rico, another for the Virgin Islands, and a third for the Pacific Islands (Guam, American Samoa, and Northern Marianas Islands). These three methodologies were adopted at different times based primarily on the data that were available at the time they were adopted. At present, because Puerto Rico is the only territory where county-level BLS OES, county-level ACS, and malpractice premium data are available, it is the only territory for which we use territory-specific data to calculate GPCIs. For the Virgin Islands, because county-level wage and rent data are not available, and insufficient malpractice premium data are available, CMS has set the work, PE, and MP GPCI values for the Virgin Islands payment locality at the national average of 1.0 even though, like Puerto Rico, the Virgin Islands is its own locality and county-level BLS OES data are available for the Virgin Islands. For the U.S. territories in the Pacific Ocean, we currently crosswalk GPCIs from the Hawaii locality for each of the three GPCIs, and incorporate no local data from these territories into the GPCI calculations even though county-level BLS OES data does exist for Guam, but not for American Samoa or the Northern Mariana Islands.

As noted above, currently Puerto Rico is the only territory for which we calculate GPCIs using the territory-specific information relative to data from the U.S. States. For several years stakeholders in Puerto Rico have raised concerns regarding the applicability of the proxy data in Puerto Rico relative to their applicability in the U.S. states. We believe that these concerns may be consistent across island territories, but lack of available, appropriate data has made it difficult to quantify such variation in costs. For example, some stakeholders previously indicated that shipping and transportation expenses increase the cost of acquiring medical equipment and supplies in islands and territories relative to the mainland. While we have previously attempted to locate data sources specific to geographic variation in such shipping costs, we found no comprehensive national data source for this information (we refer readers to 78 FR 74387 through 74388 for the detailed discussion of this issue). Therefore, we have not been able to quantify variation in costs specific to island territories in the calculation of the GPCIs.

For all the island territories other than Puerto Rico, the lack of comprehensive data about unique costs for island territories has had minimal impact on GPCIs because we have used either the Hawaii GPCIs (for the Pacific territories) or used the unadjusted national averages (for the Virgin Islands). In an effort to provide greater consistency in the calculation of GPCIs given the lack of comprehensive data regarding the validity of applying the proxy data used in the States in accurately accounting for variability of costs for these island territories, we proposed to treat the Caribbean Island territories (the Virgin Islands and Puerto Rico) in a consistent manner. We proposed to do so by assigning the national average of 1.0 to each GPCI index for both Puerto Rico and the Virgin Islands. We did not propose any changes to the GPCI methodology for the Pacific Island territories (Guam, American Samoa, and Northern Marianas Islands) where we already consistently assign the Hawaii GPCI values for each of the three GPCIs. Additional information on the Proposed Update to the Methodology for Calculating GPCIs in the U.S. Territories may be found in our contractor's draft report, “Draft Report on the CY 2017 Update of the Geographic Practice Cost Index for the Medicare Physician Fee Schedule,” which is available on our Web site. It is located under the supporting documents section of the CY 2017 PFS final rule located at https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/index.html.

The following is a summary of the comments we received regarding the proposed update to the methodology for calculating GPCIs in the U.S. territories.

Comment: Several commenters expressed support for CMS' proposal to assign the national average of 1.0 to each GPCI in Puerto Rico, stating that the physicians in Puerto Rico who treat patients enrolled in fee-for-service Medicare will be reimbursed in a manner that more closely aligns with the manner in which physicians in the other U.S. territories are reimbursed, and better reflects the cost of practicing medicine in Puerto Rico. Other commenters supporting the proposal also suggested that there has been a need for revision of Medicare payment in Puerto Rico, and that the territories of the U.S. have not been treated similarly even though the territories are much alike. Another commenter stated that the existing fee schedule for Puerto Rico does not correlate with the cost of caring for patients, and that the proposed policy is long overdue. Some commenters also stated that increasing the GPCI's for Puerto Rico is an important and necessary first step in trying to salvage Puerto Rico's deteriorated health system.

Response: We thank the commenters for their support.

Comment: A few commenters requested that CMS consider raising the GPCI values in Puerto Rico to 1.25.

Response: We proposed assigning the national average of 1.0 to each GPCI index for both Puerto Rico and the Virgin Islands, in an effort to provide greater consistency in the calculation of GPCIs among these island territories, given the lack of information on the validity of applying the proxy data used in the States to accurately account for variability of costs in these territories as compared to the national average costs. Ultimately we proposed to treat the Caribbean Island territories (the Virgin Islands and Puerto Rico) in a consistent manner by assigning the national average of 1.0 to each GPCI index. We do not believe that it would be appropriate to raise the value to 1.25 in the absence of data demonstrating that would be an accurate reflection of costs in those territories relative to national average costs.

Comment: We received several comments that are outside of the scope of the Physician Fee Schedule, requesting that CMS explore every option to determine whether a one-time correction can be made to the Medicare Advantage (MA) regulatory cycle so that the per-person monthly payment to Puerto Rico MA Plans in CY 2017 will reflect the increase to the fee-for-service spending in the territory as a result of the proposed GPCI increase. Some commenters stated that it is imperative that CMS see that the increased physician fees reach the actual providers and are not diverted away from patient care by third parties such as Medicare Advantage Organizations. Some commenters requested that CMS clarify that the new GPCIs will be incorporated into the MA benchmarks in CY 2018.

Response: We appreciate the concerns raised by the commenters. Consistent with the statute, we published the final CY 2017 Rate Announcement for Medicare Advantage on April 4, 2016. Medicare Advantage actuarial bids and benefit packages for 2017 have been approved by CMS and sponsors have begun marketing plan to beneficiaries. Thus, a change in to CY 2017 benchmark would be disruptive to beneficiaries. In future years, including CY 2018, we will follow our normal process for calculating rates. This process incorporates historical Fee for Service expenditures, which would include any updates to Fee for Service payment rates, such as an adjustment to the Puerto Rico GPCI. CMS will not be making any adjustments to CY 2017 Medicare Advantage rates as a result of this final rule. Finally, we note that according to the statute, we are prohibited from interfering or directing the contracting between Medicare Advantage Organizations (MAOs) and contracted providers. As such, we are not permitted to dictate to MAOs how any increase in payment rates can be spent, including on provider rates.

Comment: One commenter suggested that if the MA benchmark cannot be adjusted for CY 2017 that CMS should postpone the applicability of the GPCI change in Puerto Rico until CY 2018 when such an effect is also reflected in the MA benchmarks.

Response: We do not agree that the proposal to update to the methodology for calculating GPCIs in the U.S. territories, which will provide greater consistency in the calculation of GPCIs for these areas, should be delayed based on when the MA benchmarks will reflect the increases as a result of this policy.

After consideration of the public comments received regarding our proposal to treat the Caribbean Island territories (the Virgin Islands and Puerto Rico) in a consistent manner, by assigning the national average of 1.0 to each GPCI index for both Puerto Rico and the Virgin Islands, we are finalizing as proposed.

5. Refinement to the MP GPCI Methodology

In the process of calculating MP GPCIs for the purposes of this final rule, we identified several technical refinements to the methodology that yield improvements over the current method. We also proposed refinements that conform to our proposed methodology for calculating the GPCIs for the U.S. Territories described above. Specifically, we proposed modifications to the methodology to account for missing data used in the calculation of the MP GPCI. Under the methodology used in the CY 2014 GPCI update (78 FR 74380 through 74391), we first calculated the average premiums by insurer and specialty, then imputed premium values for specialties for which we did not have specific data, before adjusting the specialty-specific premium data by market share weights. We proposed to revise our methodology to instead calculate the average premiums for each specialty using issuer market share for only available companies. This proposed methodological improvement would reduce potential bias resulting from large amounts of imputation, an issue that is prevalent for insurers that only write policies for ancillary specialties for which premiums tend to be low. The current method would impute the low premiums for ancillary specialties across the remaining specialties, and generally greater imputation leads to less accuracy. Additional information on the MP GPCI methodology, and the proposed refinement to the MP GPCI methodology may be found in our contractor's draft report, “Draft Report on the CY 2017 Update of the Geographic Practice Cost Index for the Medicare Physician Fee Schedule,” which is available on our Web site. It is located under the supporting documents section of the CY 2017 PFS final rule located at https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/index.html.

We did not receive any comments regarding the proposed technical refinements to the MP GPCI methodology, and we are finalizing as proposed.

J. Payment Incentive for the Transition From Traditional X-Ray Imaging to Digital Radiography and Other Imaging Services

Section 502(a)(1) of Division O, Title V of the Consolidated Appropriations Act, 2016 (Pub. L. 114-113) amended section 1848(b) of the Act by adding new paragraph (b)(9). Effective for services furnished on or after January 1, 2017, section 1848(b)(9)(A) of the Act reduces by 20 percent the payment amounts under the PFS for the technical component (TC) (including the TC portion of a global services) of imaging services that are X-rays taken using film. The reduction is made prior to any other adjustment under this section and without application of this new paragraph.

Section 1848(b)(9) of the Act allows for the implementation of the payment reduction through appropriate mechanisms which may include use of modifiers. In accordance with section 1848(c)(2)(B)(v)(X), the adjustments under section 1848(b)(9)(A) of the Act are exempt from budget neutrality.

To implement this provision, in the CY 2017 PFS proposed rule (81 FR 46224), we proposed to establish a new modifier to be used on claims that include imaging services that are X-rays (including the imaging portion of a service) taken using film. Since the display of the proposed rule, modifier FX has been established for that purpose. Effective January 1, 2017, the modifier must be used on claims for X-rays that are taken using film. The use of this modifier will result in a 20 percent reduction for the X-ray service, as specified under section 1848(b)(9)(A) of the Act.

The proposed rule preamble stated that the applicable HCPCS codes describing imaging services that are X-ray services could be found on the PFS Web site. However, we did not intend this to indicate that we would be developing or displaying an exhaustive list of applicable codes. Instead, we intended to refer to the several lists of PFS imaging codes, including those that describe imaging services that are X-rays.

Comment: Many commenters commented on the merits of the statutory provision. The commenters stated that the reduction of Medicare film-based x-ray payments by 20 percent will have unintended consequences on patient care.

Response: We believe our proposal would implement the required statutory provision and we do not believe that we have the authority to alter the application of the provision based on these comments.

Comment: An overwhelming majority of the commenters requested CMS implement an alternative policy to improve quality of imaging services. The recommended policy would require registered radiologic technicians to perform all Medicare film or digital radiography procedures. Other commenters countered this recommended alternative by stating that it would exclude otherwise qualified professionals who have undergone education to acquire limited scope licenses or certification programs demonstrating As Low As Reasonably Achievable (ALARA) safety practices by either a third party, vendor training, or another didactic course deemed acceptable by any of the four accreditation organizations. One commenter also referenced 35 states that have an entry level certification for X-ray technicians and that throughout the US, there are x-ray technicians and limited scope X-ray machine operators that are also licensed and certified.

Response: We appreciate commenters' interests in standards that might improve quality of care for Medicare beneficiaries, but we did not propose a policy regarding standards for radiologic technicians in the proposed rule. Also, as previously stated, we do not believe we have the authority to implement conditions of payment regarding radiologic technicians as an alternative to the adjustments required by the statutory provision.

Comment: A commenter recommended that a financial incentive be provided for physicians to convert to digital machines as had been done in the case of electronic medical records.

Response: The legislation does not authorize any financial incentive in the form of increased payment, but provides an incentive to use digital images to avoid the 20 percent reduction that applies to imaging services that are X-rays taken using film.

Comment: One commenter requested that in the absence of a meaningful opportunity to comment on the list of codes for which the policy applies, the provision should be limited to traditional diagnostic X-ray procedures only. Two commenters presented separate lists of codes for which the payment reduction should apply. One commenter also provided codes that should be explicitly excluded from the payment reduction, for example, radiographic-fluoroscopic, vascular and mammography X-ray imaging services, radioscopic, radioscopic and radiography services provided in a single examination. Other commenters also provided a list of procedures that should be excluded. The commenter also requested that we publish the list of applicable codes as soon as possible.

Response: As previously stated, we did not publish an exhaustive list of applicable codes, and previously intended to point to existing lists of PFS imaging services. We believe that physicians and non-physician practitioners are in the best position to determine whether a particular imaging service is an X-ray taken using film.

Comment: One commenter suggested that if at least half of the number of discrete X-ray exposures required for the radiographic exam are captured using a DR detector, then the examination should be considered as digital and the payment differential should not be applied. Another commenter requested that we clarify that the law only applies (and requires use of a modifier) to sites that use X-ray as a single method for image capture. The commenter also seeks clarification that if a site uses both X-ray film and electronic capture of images and maintains digital archives, by a picture archiving communication system or other electronic method, that the site is not required to report the modifier.

Response: At this time, in accordance with the statute, we are requiring the FX modifier to be used whenever an imaging service is an X-ray taken using film. As stated, the statute requires that if an imaging service is an X-ray taken using film, a reduction in payment is to occur. The statutory requirement applies at the service level, not based on where the service is furnished or the method used to store images. There is no provision for an exception to the payment reduction based on the availability of various technologies or the use of certain image archiving technology at a particular site.

After consideration of the public comments we received, we are finalizing the establishment of new modifier “FX” to be reported on claims for imaging services that are X-rays that are taken using film.

Beginning January 1, 2017, claims for imaging services that are X-rays taken using film must include the modifier “FX.”

The use of this modifier will result in a 20 percent reduction for the X-ray service, as specified under section 1848(b)(9)(A) of the Act.

K. Procedures Subject to the Multiple Procedure Payment Reduction (MPPR) and the OPPS Cap

Effective January 1, 2012, we implemented an MPPR of 25 percent on the professional component (PC) of advanced imaging services. The reduction applies when multiple imaging procedures are furnished by the same physician (or physician in the same group practice) to the same patient, in the same session, on the same day. Full payment is made for the PC of the highest priced procedure. Payment for the PC of subsequent services is reduced by 25 percent.

Section 502(a)(2)(A) of Division O, Title V of the Consolidated Appropriations Act, 2016 (Pub. L. 114-113, enacted on December 18, 2015) added a new section 1848(b)(10) of the Act which revises the payment reduction from 25 percent to 5 percent, effective January 1, 2017. Section 502(a)(2)(B) added a new subclause at section 1848(c)(2)(B)(v)(XI) which exempts the reduced expenditures attributable to the revised 5 percent MMPR on the PC of imaging from the PFS budget neutrality provision. We proposed to implement these provisions for services furnished on or after January 1, 2017. We refer readers to section VI.C of this final rule regarding the necessary adjustment to the proposed PFS conversion factor to account for the mandated exemption from PFS budget neutrality.

We note that the lists of services for the upcoming calendar year that are subject to the MPPR on diagnostic cardiovascular services, diagnostic imaging services, diagnostic ophthalmology services, and therapy services; and the list of procedures that meet the definition of imaging under section 5102(b) of the Deficit Reduction Act (DRA), and therefore, are subject to the OPPS cap, are displayed in the public use files for the PFS proposed and final rules for each year. The public use files for CY 2017 are available on our Web site under downloads for the CY 2017 PFS final rule with comment period at http://www.cms.gov/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/PFSFederal-Regulation-Notices.html.

Comment: Commenters supported the proposal to implement the statutory provision.

Response: We our finalizing our CY 2017 proposal to revise the MPPR on the PC of diagnostic imaging services.

L. Valuation of Specific Codes

1. Background: Process for Valuing New, Revised, and Potentially Misvalued Codes

Establishing valuations for newly created and revised CPT codes is a routine part of maintaining the PFS. Since inception of the PFS, it has also been a priority to revalue services regularly to make sure that the payment rates reflect the changing trends in the practice of medicine and current prices for inputs used in the PE calculations. Initially, this was accomplished primarily through the 5-year review process, which resulted in revised work RVUs for CY 1997, CY 2002, CY 2007, and CY 2012, and revised PE RVUs in CY 2001, CY 2006, and CY 2011. Under the 5-year review process, revisions in RVUs were proposed and finalized via rulemaking. In addition to the 5-year reviews, beginning with CY 2009, CMS and the RUC have identified a number of potentially misvalued codes each year using various identification screens, as discussed in section II.D.4 of this final rule. Historically, when we received RUC recommendations, our process had been to establish interim final RVUs for the potentially misvalued codes, new codes, and any other codes for which there were coding changes in the final rule for a year. Then, during the 60-day period following the publication of the final rule, we accepted public comment about those valuations. For services furnished during the calendar year following the publication of interim final rates, we paid for services based upon the interim final values established in the final rule. In the final rule with comment period for the subsequent year, we considered and responded to public comments received on the interim final values, and typically made any appropriate adjustments and finalized those values.

In the CY 2015 PFS final rule with comment period, we finalized a new process for establishing values for new, revised and potentially misvalued codes. Under the new process, we include proposed values for these services in the proposed rule, rather than establishing them as interim final in the final rule with comment period. Beginning with the CY 2017 proposed rule, the new process is applicable to all codes, except for new codes that describe truly new services. For CY 2017, we proposed new values in the CY 2017 proposed rule for the vast majority of new, revised, and potentially misvalued codes for which we received complete RUC recommendations by February 10, 2016. To complete the transition to this new process, for codes where we established interim final values in the CY 2016 PFS final rule with comment period, we reviewed the comments received during the 60-day public comment period following release of the CY 2016 PFS final rule with comment period, and re-proposed values for those codes in the CY 2017 proposed rule.

We considered public comments received during the 60-day public comment period for the proposed rule before establishing final values in this final rule. As part of our established process we will adopt interim final values only in the case of wholly new services for which there are no predecessor codes or values and for which we do not receive recommendations in time to propose values. For CY 2017, we are not aware of any new codes that describe such wholly new services. Therefore, we are not establishing any code values on an interim final basis. However, we remind stakeholders that we continually review stakeholder information regarding the valuation of codes under the potentially misvalued code initiative and, under our existing process, could consider proposing any particular changes as early as CY 2018 rulemaking.

2. Methodology for Proposing Work RVUs

We conduct a review of each code identified in this section and review the current work RVU (if any), RUC-recommended work RVU, intensity, time to furnish the preservice, intraservice, and postservice activities, as well as other components of the service that contribute to the value. Our review of recommended work RVUs and time inputs generally includes, but is not limited to, a review of information provided by the RUC, HCPAC (Health Care Professionals Advisory Committee), and other public commenters, medical literature, and comparative databases, as well as a comparison with other codes within the PFS, consultation with other physicians and health care professionals within CMS and the federal government, as well as Medicare claims data. We also assess the methodology and data used to develop the recommendations submitted to us by the RUC and other public commenters and the rationale for the recommendations. In the CY 2011 PFS final rule with comment period (75 FR 73328 through 73329), we discussed a variety of methodologies and approaches used to develop work RVUs, including survey data, building blocks, crosswalks to key reference or similar codes, and magnitude estimation (see the CY 2011 PFS final rule with comment period for more information). When referring to a survey, unless otherwise noted, we mean the surveys conducted by specialty societies as part of the formal RUC process. The building block methodology is used to construct, or deconstruct, the work RVU for a CPT code based on component pieces of the code.

Components used in the building block approach may include preservice, intraservice, or postservice time and post-procedure visits. When referring to a bundled CPT code, the building block components could be the CPT codes that make up the bundled code and the inputs associated with those codes. Magnitude estimation refers to a methodology for valuing work that determines the appropriate work RVU for a service by gauging the total amount of work for that service relative to the work for a similar service across the PFS without explicitly valuing the components of that work. In addition to these methodologies, we have frequently utilized an incremental methodology in which we value a code based upon its incremental difference between another code or another family of codes. The statute specifically defines the work component as the resources in time and intensity required in furnishing the service. Also, the published literature on valuing work has recognized the key role of time in overall work. For particular codes, we refine the work RVUs in direct proportion to the changes in the best information regarding the time resources involved in furnishing particular services, either considering the total time or the intraservice time.

Several years ago, to aid in the development of preservice time recommendations for new and revised CPT codes, the RUC created standardized preservice time packages. The packages include preservice evaluation time, preservice positioning time, and preservice scrub, dress and wait time. Currently there are six preservice time packages for services typically furnished in the facility setting, reflecting the different combinations of straightforward or difficult procedure, straightforward or difficult patient, and without or with sedation/anesthesia. Currently, there are three preservice time packages for services typically furnished in the nonfacility setting, reflecting procedures without and with sedation/anesthesia care.

We have developed several standard building block methodologies to value services appropriately when they have common billing patterns. In cases where a service is typically furnished to a beneficiary on the same day as an E/M service, we believe that there is overlap between the two services in some of the activities furnished during the preservice evaluation and postservice time. Our longstanding adjustments have reflected a broad assumption that at least one-third of the work time in both the preservice evaluation and postservice period is duplicative of work furnished during the E/M visit.

Accordingly, in cases where we believe that the RUC has not adequately accounted for the overlapping activities in the recommended work RVU and/or times, we adjust the work RVU and/or times to account for the overlap. The work RVU for a service is the product of the time involved in furnishing the service multiplied by the intensity of the work. Preservice evaluation time and postservice time both have a long-established intensity of work per unit of time (IWPUT) of 0.0224, which means that 1 minute of preservice evaluation or postservice time equates to 0.0224 of a work RVU.

Therefore, in many cases when we remove 2 minutes of preservice time and 2 minutes of postservice time from a procedure to account for the overlap with the same day E/M service, we also remove a work RVU of 0.09 (4 minutes × 0.0224 IWPUT) if we do not believe the overlap in time has already been accounted for in the work RVU. The RUC has recognized this valuation policy and, in many cases, now addresses the overlap in time and work when a service is typically furnished on the same day as an E/M service.

We note that many commenters and stakeholders have expressed concerns with our ongoing adjustment of work RVUs based on changes in the best information we have regarding the time resources involved in furnishing individual services. We are particularly concerned with the RUC's and various specialty societies' objections to our approach given the significance of their recommendations to our process for valuing services and since much of the information we have used to make the adjustments is derived from their survey process. As explained in the CY 2016 PFS final rule with comment period (80 FR 70933), we recognize that adjusting work RVUs for changes in time is not always a straightforward process, so we apply various methodologies to identify several potential work values for individual codes. However, we want to reiterate that we are statutorily obligated to consider both time and intensity in establishing work RVUs for PFS services.

We have observed that for many codes reviewed by the RUC, final recommended work RVUs appear to be incongruous with recommended assumptions regarding the resource costs in time. This is the case for a significant portion of codes for which we have recently established or proposed work RVUs that are based on refinements to the RUC-recommended values. When we have adjusted work RVUs to account for significant changes in time, we begin by looking at the change in the time in the context of the RUC-recommended work RVU. When the recommended work RVUs do not appear to account for significant changes in time, we employ the different approaches to identify potential values that reconcile the recommended work RVUs with the recommended time values. Many of these methodologies, such as survey data, building blocks, crosswalks to key reference or similar codes, and magnitude estimation have long been used in developing work RVUs under the PFS. In addition to these, we sometimes use the relationship between the old time values and the new time values for particular services to identify alternative work RVUs based on changes in time components.

In so doing, rather than ignoring the RUC-recommended value, we are using the recommended values as a starting reference and then applying one of these several methodologies to account for the reductions in time that we believe have not otherwise been reflected in the RUC-recommended value. When we believe that such changes in time have already been accounted for in the RUC recommendation, then we do not make such adjustments. Likewise, we do not arbitrarily apply time ratios to current work RVUs to calculate proposed work RVUs. We use the ratios to identify potential work RVUs and consider these work RVUs as potential options relative to the values developed through other options.

We clarify that we are not implying that the decrease in time as reflected in survey values must equate to a one-to-one or linear decrease in newly valued work RVUs. Instead, we believe that, since the two components of work are time and intensity, absent an obvious or explicitly stated rationale for why the relative intensity of a given procedure has increased, significant decreases in time should be reflected in decreases to work RVUs. If the RUC recommendation has appeared to disregard or dismiss the changes in time, without a persuasive explanation of why such a change should not be accounted for in the overall work of the service, then we generally use one of the aforementioned referenced methodologies to identify potential work RVUs, including the methodologies intended to account for the changes in the resources involved in furnishing the procedure.

Several commenters, including the RUC, in general have objected to our use of these methodologies and deemed our actions in adjusting the recommended work RVUs as inappropriate. We received several specific comments regarding this issue in response to the CY 2016 PFS final rule with comment period and those comments are summarized below.

Comment: Several commenters, including the RUC, stated that our methodology for adjusting work RVUs appears to be contrary to the statute.

Response: We disagree with these comments. Since section 1848(c)(1)(A) of the Act explicitly identifies time as one of the two types of resources that encompass the work component of the PFS payment, we do not believe that our use of the aforementioned methodologies to adjust the work RVU to account for the changes in time, which is one of the resources involved, is inconsistent with the statutory requirements related to the maintenance of work RVUs, and we have regularly used these and other methodologies in developing values for PFS services. In selecting which methodological approach will best determine the appropriate value for a service, we consider the current and recommended work and time values, as well as the intensity of the service, all relative to other services. In our review of RUC recommended values, we have observed that the RUC also uses a variety of methodologies to develop work RVUs for individual codes, and subsequently validates the results of these approaches through magnitude estimation or crosswalk to established values for other codes.

Comment: Several commenters, including the RUC, stated that we could not take one element of the services that has changed such as intra-service time, and apply an overall ratio for reduction to the work RVU based on changes to time, as that renders the value no longer resource-based in comparison to the RUC-recommended values.

Response: We disagree with the commenters and continue to believe that the use of time ratios is one of several reasonable methods for identifying potential work RVUs for particular PFS services, particularly when the alternative values do not account for information that suggests the amount of time involved in furnishing the service has changed significantly. We reiterate that, consistent with the statute, we are required to value the work RVU based on the relative resources involved in furnishing the service, which include time and intensity. When our review of recommended values determines that changes in the resource of time have been unaccounted for in a recommended RVU, then we believe we have the obligation to account for that change in establishing work RVUs since the statute explicitly identifies time as one of the two elements of the work RVUs. We recognize that it would not be appropriate to develop work RVUs solely based on time given that intensity is also an element of work, but in applying the time ratios we are using derived intensity measures based on current work RVUs for individual procedures. Were we to disregard intensity altogether, the work RVUs for all services would be developed based solely on time values and that is definitively not the case. Furthermore, we reiterate that we use time ratios to identify potential work RVUs, and then use other methods (including estimates of work from CMS medical personnel and crosswalks to key reference or similar codes) to validate these RVUs. We also disagree with several commenters' implications that a work RVU developed through such estimation methods is only resource-based through the RUC process.

Comment: Several commenters, including the RUC, stated that our inconsistent use of the time ratio methodology has rendered it ineffective for valuation purposes and that by choosing the starting base work value and/or physician time at random, we are essentially reverse engineering the work value we want under the guise of a standard algorithm.

Response: We do not choose a starting base work value and/or physician time at random as suggested by the commenters. We use the RUC recommended values or the existing values as the base values; essentially, we are taking one of those values and applying adjustments to account for the change in time that based on our analysis of the RUC recommendation, we determine has not been properly accounted for to determine an appropriate work RVU. In circumstances where adjustments to time and the corresponding work RVU are relatively congruent or persuasively explained, our tendency has been to use those values as recommended. Where the RUC recommendations do not account for changes in time, we have made changes to RUC-recommended values to account for the changes in time.

Comment: Commenters, including the RUC, also stated that the use of time ratio methodologies distills the valuation of the service into a basic formula with the only variable being either the new total physician time or the new intra-service physician time, and that these methodologies are based on the incorrect assumption that the per minute physician work intensity established is permanent regardless of when the service was last valued. Other commenters have suggested that previous assumed times are inaccurate.

Response: We agree with commenters that per minute intensity for a given service may change over time. If we believed that the per-minute intensity for a given service were immutable, then a reverse-building block approach to revaluation based on new time data could be appropriate. However, we have not applied such an approach specifically because we agree that the per-minute intensity of work is not necessarily static over time or even necessarily during the course of a procedure. Instead, we utilize time ratios to identify potential values that account for changes in time and compare these values to other PFS services for estimates of overall work. When the values we develop reflect a similar derived intensity, we agree that our values are the result of our assessment that the relative intensity of a given service has remained similar.

Regarding the validity of comparing new times to the old times, we, too, hope that time estimates have improved over many years especially when many years have elapsed since the last time the service in question was valued. However, we also believe that our operating assumption regarding the validity of the pre-existing values as a point of comparison is critical to the integrity of the relative value system as currently constructed. Pre-existing times are a very important element in the allocation of indirect PE RVUs by specialty, and had the previously recommended times been overestimated, the specialties that furnish such services would be benefitting from these times in the allocation of indirect PE RVUs. As long time observers of the RUC process, we also recognize that the material the RUC uses to develop overall work recommendations includes the data from the surveys about time. We have previously stated concerns regarding the validity of much of the RUC survey data. However, we believe additional kinds of concern would be warranted if the RUC itself were operating under the assumption that its pre-existing data were typically inaccurate.

We understand stakeholders' concerns regarding how best to consider changes in time in improving the accuracy of work RVUs and have considered all of the issues raised by commenters. In conjunction with our review of recommended code values for CY 2017, we conducted a preliminary analysis to identify general tendencies in the relationship between changes in time and changes in work RVUs for CY 2014 and CY 2015. We looked at services for which there were no coding changes to simplify the analysis. The intent of this preliminary analysis was to examine commenters' beliefs that CMS is only considering time when making refinements to RUC recommended work values. For CY 2014, we found that in the aggregate, the average difference between the RUC recommended intraservice time and existing intraservice time was −17 percent, but the average difference between the RUC recommended work RVU and existing work RVU was only −4 percent. However, the average difference between the CMS refined work RVU and existing work RVU was −7 percent. For CY 2015, the average difference between the RUC recommended intraservice time and existing intraservice time was −17 percent, but the average difference between the RUC recommended work RVU and existing work RVU was 1 percent, and the average difference between the CMS refined work RVU and existing work RVU was −6 percent. This preliminary analysis demonstrates that we are not making refinements solely in consideration of time, if that were the case, the changes in the work RVU values that we adopted would be comparable to the changes in the time that we adopted, but that is not the case.

We believe that we should account for efficiencies in time when the recommended work RVU does not account for those efficiencies, otherwise relativity across the PFS can be significantly skewed over periods of time. For example, if when a code is first valued, a physician was previously able to do only 5 procedures per day, but due to new technologies, the same physician can now do 10 procedures per day, resource costs in time have empirically been lessened, and we believe that relative reduction in resources involved in furnishing that service should be accounted for in the assignment of work RVUs for that service, since the statute explicitly identifies time as one of the two components of work. Of course, if more resource intensive technology has allowed for the increased efficiency in furnishing the procedure, then the nonfacility PE RVUs for the service should also be adjusted to account for this change. Additionally, we believe it may be that the intensity per minute of the procedure may have changed with the greater efficiency in time. Again, that is why we do not generally reduce work RVUs in strict proportion to changes in time. We understand that intensity is not entirely linear, and that data related to time as obtained in the RUC survey instrument may improve over time, and that the number of survey respondents may improve over time. However, we also understand time as a tangible resource cost in furnishing PFS services, and a cost that by statute, is one of the two kinds of resources to be considered as part of the work RVU.

Therefore, in the proposed rule, we stated that we were interested in receiving comments on whether, within the statutory confines, there are alternative suggestions as to how changes in time should be accounted for when it is evident that the survey data and/or the RUC recommendation regarding the overall work RVU does not reflect significant changes in the resource costs of time for codes describing PFS services. We solicited comment on potential alternatives, including the application of the reverse building block methodology, to making the adjustments that would recognize overall estimates of work in the context of changes in the resource of time for particular services.

The following is a summary of the comments we received in response to our solicitation regarding potential alternatives, including the application of the reverse building block methodology, to making the adjustments that would recognize overall estimates of work in the context of changes in the resource of time for particular services.

Comment: One commenter stated that it continues to support CMS in its efforts to adjust work RVUs commensurate with changes in intra-service and total time, as well as post-operative visits despite RUC recommendations to the contrary. The commenter agreed with our changes and encouraged CMS to continue to identify and address such incongruities. The commenter stated that it is routine to encounter recommended decreases in physician time and/or post-procedure visits combined with RUC recommendations to maintain or increase the work RVUs. The commenter agreed that when physician time decreases, physician work should decrease comparatively, absent a compelling argument that the intensity of the service has increased sufficiently to offset the decrease in physician time. The commenter did not have alternative suggestions for how CMS should make these adjustments, and believes the approaches that CMS has taken are reasonable and defensible.

Another commenter stated that it appreciates that CMS provided information about how it reviews recommendations for work RVUs that come from the RUC. Additionally, one commenter stated appreciation for the consideration and effort that CMS gives in valuing the work RVUs for a service. The commenter stated that the accuracy of RVU estimates has improved as a result of CMS' various validation processes for collecting data and its consideration of feedback from the RUC and public commenters. The commenter stated that CMS should account for efficiencies in the resource costs of time when the recommended work RVU does not account for emerging efficiencies, such as advances in surgical techniques, and that by considering time in these situations, CMS will be able to effectively adjust both emerging technological trends and their impact on resource costs needed to deliver care to beneficiaries.

Response: We appreciate the commenters' support for our ongoing adjustment of work RVUs based on changes in the best information we have regarding the time resources involved in furnishing individual services. We also agree that CMS should account for efficiencies in the resource costs of time, as indicated by one commenter, and will endeavor to do so when we consider the work RVU and how the effect of advancements such as emerging technology and improvements in surgical techniques impact the resource costs of time.

Comment: A few commenters, including the RUC, stated that all adjustments to work RVUs should be solely based on the resources involved in performing each procedure or service. The commenters stated that all adjustments to work RVUs should either be work neutral to the family or result in budget neutral adjustment to the conversion factor, and that broadly redistributing work RVUs would distort the relative value system and create unintended consequences.

Response: We agree that adjustments to work RVUs should be based on the resources involved with each procedure or service, and consistent with the statute, the work RVUs should reflect the relative resources costs of time and intensity. We also agree with the commenter regarding how changes in work RVUs affect PFS relativity. We have a long-standing practice of making an adjustment to the CF to account for increases or decreases in work RVUs across the PFS instead of scaling the work RVUs to maintain overall relativity. The practical effect of a positive adjustment to the CF is that the value of a single work RVU is greater than it previously had been. In other words, the relative value of the other work RVUs has increased, across the PFS, whenever we apply a positive budget neutrality adjustment to the CF to account for an overall decrease in work RVUs.

Comment: A few commenters, including the RUC, stated that they appreciate CMS agreeing with the RUC's assertion that the usage of time ratios to reduce work RVUs is typically not appropriate, as often a change in physician time coincides with a change in the physician work intensity per minute. The commenters stated that CMS acknowledges that physician work intensity per minute is typically not linear and also that making reductions in RVUs in strict proportion to changes in time is inappropriate.

Response: We do not agree with the commenters' characterization of our statements, and believe it misinterprets our view on this matter. We specifically stated in the CY 2017 proposed rule that we are not implying that the decrease in time as reflected in survey values must necessarily equate to a one-to-one or linear decrease in newly valued work RVUs, given that intensity for any given procedure may change over several years or within the intraservice period. Nevertheless, we believe that since the two components of work are time and intensity, that absent an obvious or explicitly stated rationale for why the relative intensity of a given procedure has specifically increased or that the reduction in time is disproportionally from less-intensive portions of the procedure, that significant decreases in time should generally be reflected in decreases to work RVUs.

Comment: A few commenters, including the RUC, stated that they wanted to remind CMS of the Agency's and the RUC's longstanding position that treating all components of physician time as having identical intensity is incorrect, and inconsistently applying this treatment to only certain services under review creates inherent payment disparities in a payment system that is based on relative valuation. The commenters stated that when physician times are updated in the fee schedule, the ratio of intra-service time to total time, the number and level of bundled post-operative visits, the length of pre-service, and the length of immediate post-service time may all potentially change for the same service. These changing components of physician time result in the physician work intensity per minute often changing when physician time also changes, and the commenters recommended that CMS always account for these nuanced variables. A few commenters also stated that the RUC recommendations now explicitly state when physician time has changed and address whether and to what magnitude these changes in time impact the work involved.

Response: We appreciate the commenters' feedback. We understand that not all components of physician time have identical intensity and are mindful of this point when we are determining what the appropriate work RVU values should be. We also agree that the nuanced variables involved in the changing components of physician time must be accounted for, and it is our goal to do so when determining the appropriate valuation. We appreciate when the RUC recommendations provide as much detailed information regarding the recommended valuations as possible, including thorough discussions regarding physician time changes and how the RUC believes such changes should or should not impact the work involved, and we consider that information when conducting our review of each code.

Comment: A few commenters stated that CMS places undue emphasis on time and not enough emphasis on intensity or whether a value is appropriately ranked in the Medicare fee schedule. The commenters stated that CMS ignores compelling evidence that work has changed if the time has not also changed, and that CMS uses codes as supporting references for new lower values that make no clinical sense. The commenters urged CMS to always consider all elements of relative work in every review, including time, relative intensity and relative work.

Response: We disagree with commenters' statement that CMS ignores compelling evidence that work has changed if the time has not also changed. As previously stated, we are not making refinements solely in consideration of time, and if that were the case, changes in work RVU values that we adopted would consistently be comparable to the changes in the time that we adopted, and that is not the case. It is our practice to consider all elements of the relative work when we are reviewing and determining work RVU valuations. Additionally, our review of recommended work RVUs and time inputs generally includes review of various sources such as information provided by the RUC, other public commenters, medical literature, and comparative databases.

Comment: A few commenters, including the RUC, stated that they do not agree with any suggested methodology to use a reverse building block methodology to systematically reduce work RVUs for services. The commenters stated that any purely formulaic approach should never be used as the primary methodology to value services, and that it is highly inappropriate due to the fact that magnitude estimation was used to establish work RVUs for services.

Response: We note that a formulaic approach is not being used as the primary methodology to value services. Instead, we use various methodologies to identify values to consider relative to other PFS services. We reiterate that we use the RUC-recommended values or the existing values as the base values. We then apply adjustments to the RUC-recommended values where, for example, the RUC's recommendations do not account for changes in time.

Comment: Another commenter stated that the establishment of a time formula or use of reverse building block methodology as the primary method for valuation would completely disregard the possibility that physicians actually get better at what they do in favor of the erroneous conclusion that physicians only find new ways to cut corners. The commenter provided an example to demonstrate why time alone does not create value, and it is instead just one component of valuation. The commenter described an example of two watchmakers that make watches at different rates—one makes two watches per day, the other makes four watches per day. Each watch involves the same number of gears, sprockets, jewels, and escapements. One watchmaker is faster than the other: More focused, more experienced, more agile, and able to accomplish fastidious work more efficiently. At the end of one workday, the first watchmaker has two finished watches on the bench, while the other has four. The commenter questioned that if the watches are identical, why should the faster (better) watchmaker be paid half the price for each watch?

Response: We understand some stakeholders' interest in the maintenance of work RVUs regardless of efficiencies gained. The work RVU is not a measure of our appreciation for the work ethic of the physician. Instead, the work RVU reflects the time and intensity of a particular service relative to others on the PFS. For this reason, we do not agree with the implication that we should ignore efficiencies in time, and instead believe that we are obligated to recognize when efficiencies change the relative resource costs involved in particular procedures. Of course, such efficiencies can occur as physicians become more proficient and can therefore complete a service or procedure in less time. We believe that time is a tangible resource, particularly the time of a physician or other practitioner paid on the PFS, and the statute specifically identifies it as such.

Comment: A few commenters urged CMS to always enlist the assistance of medical officers familiar with procedures under review to examine CMS staff recommendations that reject the RUC recommendation. Similarly, a few commenters also urged CMS to work with the RUC to ensure that the robust discussions and key points that are discussed during RUC meetings are transferred to CMS in a way that is meaningful to staff to develop the proposed relative value recommendations.

Response: We note that the values proposed by CMS are developed through consultation with, and input from CMS staff including medical officers, who often attend RUC meetings as observers, and therefore, have had the opportunity to listen to the discussions that take place and key points that are raised during the RUC meetings.

Comment: One commenter stated that the recent rejections of RUC recommendations by CMS to instead reduce the work RVUs for almost every code, even if only by one or two percent, are illogical.

Response: We do not agree with the suggestion that we reject the RUC-recommended values for most codes. Furthermore, given the numerical specificity of the RUC-recommended values and that so many PFS services reviewed under the misvalued code initiative are high-volume, we do not believe that relatively minor adjustments are unimportant or illogical because a minor adjustment to the work RVU of a high-volume code may have a significant dollar impact. However, we would be interested to know if stakeholders generally agree that the RUC-recommended values represent only rough estimates, and because of that belief, minor refinements would be considered illogical, as indicated by the commenter.

Comment: A few commenters stated that they are concerned with the CMS trend to discredit intensity when assigning work RVUs to procedures. These commenters stated that intensity is a key factor when specialties are making work RVU recommendations and needs to remain an equal force along with time in the relative value system. One commenter stated that it is concerned that CMS is repeatedly ignoring intensity discussions and picking arbitrary crosswalks to justify lowering work RVU values. One commenter stated that by placing the same value on clearly different services that vary both in intensity and in types of patients treated, CMS ignores its statutory requirement to consider time and intensity in the valuation of services. One commenter stated that CMS does not mention how it considers, weights or measures intensity, and there is no validity to the assumption that reduced time equals less work. The commenter stated that it found no published evidence supporting this, and states that if the same amount of work is performed for a shorter period of time, it is logical that the intensity of work per unit of time increases. The commenter stated that CMS must be transparent and demonstrate why current intensity measurements are not appropriate, and if there is a more accurate way to measure intensity, this must be clearly elucidated with evidence for superiority of the alternative proposal.

Response: We disagree that we discredit intensity when we establish work RVUs for procedures. We reiterate that we use RUC-recommended values or existing values, which we understand to incorporate an assessment of intensity, as the base values, and then subsequently apply adjustments as necessary. Additionally, as we have previously stated, we recognize that it would not be appropriate to develop work RVUs solely based on time given that intensity is also an element of work. Additionally, if we were to disregard intensity altogether, the work RVUs for all services would be developed solely based on time values, and that is absolutely not the case. We have previously stated that in cases where the RUC's recommendations do not account for changes in time, but do provide a persuasive explanation regarding why the time has drastically changed but the work RVU value has remained the same, our tendency has been to use those values as recommended. When the RUC's recommendations do not account for changes in time, and provide no explanation as to why this is appropriate, we have made changes to the RUC-recommended values to account for changes in time.

We also disagree that we ignore the statutory requirement to consider time and intensity in the valuation of services. Based on the assessments of CMS medical officers and other reviewers, as well as upon consideration of the survey results, and the rationales in the recommendations, we make determinations about the overall work valuations. We acknowledge that the degree to which intensity varies among different procedures is a relatively subjective assessment, and we understand that sometimes stakeholders may have a different perspective in cases where the intensities of procedures differ. We recognize that the IWPUT measure is a derived value with specific uses for quantifying intensity. However, the limited way in which that derived value is used under the RUC valuation process, we believe reflects a general consensus that there are not widely accepted metrics for intensity. As a part of recommendations for misvalued codes, we welcome any information from stakeholders for us to more objectively measure intensity.

Comment: A few commenters stated that they are concerned with the current implied methodology that the 25th survey percentile is the ceiling for RUC recommendations, and stated that if codes are continually sent forth for re-survey and the 25th percentile is the ceiling, a built in reduction is applied to all surveyed codes just by the nature of surveying the codes, regardless of other factors.

Response: We disagree with commenters' statement that the 25th survey percentile is the ceiling for RUC recommendations. We note that, as previously stated in the CY 2011 final rule with comment period (75 FR 73328), we had concerns that surveys conducted on existing codes produced predictable results, and upon clinical review of a number of these situations, we were concerned over the validity of the survey results since the survey values often were very close to the current code values. Increasingly, the RUC is choosing to recommend the 25th percentile survey value, potentially responding to the same concern we have identified, rather than recommending the median survey value that had historically been the most commonly used. We reiterate that this does not designate the 25th percentile as the ceiling, rather suggests that in many instances the 25th percentile is the most appropriate since it is more frequently being identified through the RUC process as the recommended value.

Comment: One commenter stated that the time data obtained through the RUC survey process based on subjective physician perceptions of time, may not be the most accurate data available on intraoperative time. The commenter stated that CMS should be open to reviewing additional sources of objective validated time data, and that such sources might include peer reviewed and published studies of comparative surgery times amongst different procedures in the same institution using standardized metrics. Another commenter stated that if CMS seeks specific information to substantiate time and intensity changes associated with services, they should specify these clearly so stakeholders can provide the necessary data detailing changes over time.

Response: As previously discussed, our review of work RVUs and time inputs utilizes information from various resources. It generally includes, but is not limited to a review of information provided by the RUC, HCPAC, and other public commenters, medical literature, and comparative databases, as well as comparison with other codes with the PFS, consultation with other physicians and health care professionals within CMS and the federal government, as well as Medicare claims data. Additionally, we also assess the methodology and data used to develop the recommendations submitted to us by the RUC and other public commenters and the rationale for the recommendations. However, we continue to seek information regarding the best sources of objective, routinely-updated, auditable, and robust data regarding the resource costs of furnishing PFS services.

We thank the commenters for their feedback. We did not receive any comments regarding specific potential alternatives to making the adjustments that would recognize overall estimates of work in the context of changes in the resource of time for particular services as requested. However, we appreciate the commenters' sharing their concerns and suggestions and will continue to consider them as we continue examining the valuation of services, and as we explore the best way to address these issues.

3. Requests for Refinement Panel

Consistent with the policy finalized in the CY 2016 PFS final rule with comment period, we have retained the Refinement Panel process for use with codes with interim final values where additional input by the panel is likely to add value as a supplement to notice and comment rulemaking. Because there are no codes with interim final values in this final rule, the refinement panel is not necessary for CY 2017. We note that many commenters requested inclusion of codes with proposed values for a refinement panel. While these requests are not consistent with our established process, given the number of requests we received, we are addressing them here. Many commenters appear to believe that that the purpose of the refinement panel process was to serve as a kind of “appeals” or reconsideration process outside of notice and comment rulemaking and that we have effectively eliminated a useful appeals process. We understand that the refinement panel has been perceived as an appeals process by many stakeholders. However, as we have previously clarified, the purpose of the refinement panel process was to assist us in reviewing the public comments on CPT codes with interim final work RVUs and to consider more fully the interests of the specialty societies who provide input on RVU work time and intensity with the budgetary and redistributive effects that could occur if we accepted extensive increases in work RVUs across a broad range of services. From our perspective, the objective of the refinement panel has long been to provide a needed venue for stakeholders to present any new clinical information that was not available at the time of the RUC valuation for interim final values in order that we arrive at the most appropriate final valuation, especially since the initial values for such codes were generally established approximately 2 months prior to being used for Medicare payment. In recent years, we have continually observed that the material presented to the refinement panel largely raised and discussed issues and perspective already included as part of the RUC meetings and considered by us.

We believe that our new process, in which we propose the vast majority of code values in the proposed rule for public comment on those proposed values prior to their taking effect, provides stakeholders and the public with several opportunities to present data or information that might affect code valuation. We believe that this is generally consistent with the purpose of the rulemaking process and reflects our efforts to increase transparency and accountability to the public. We also note that we continue to seek new information that is relevant to valuation of particular services, including those with values recently finalized, for use in future rulemaking. We believe that notice and comment rulemaking provides the most appropriate means for valuing services under the PFS. We note that in several instances in this final rule, thoughtful and informative comments have helped us to finalize values for CY 2017 that we believe are improved from those we had proposed. In many cases, these changes reflect the RUC-recommended value. Therefore, we urge commenters to review this information and continue to consider how we might continue to improve the notice and comment rulemaking process rather than establish a process outside of notice and comment rulemaking.

Table 27 contains a list of codes for which we proposed work RVUs; this includes all RUC recommendations received by February 10, 2016, and codes for which we established interim final values in the CY 2016 PFS final rule with comment period. When the proposed work RVUs vary from those recommended by the RUC or for which we do not have RUC recommendations, we address those codes in the portions of this section that are dedicated to particular codes. The final work RVUs and work time and other payment information for all CY 2017 payable codes are available on the CMS Web site under downloads for the CY 2017 PFS final rule at http://www.cms.gov/physicianfeesched/downloads/.

4. Methodology for Proposing the Direct PE Inputs To Develop PE RVUs

a. Background

On an annual basis, the RUC provides us with recommendations regarding PE inputs for new, revised, and potentially misvalued codes. We review the RUC-recommended direct PE inputs on a code by code basis. Like our review of recommended work RVUs, our review of recommended direct PE inputs generally includes, but is not limited to, a review of information provided by the RUC, HCPAC, and other public commenters, medical literature, and comparative databases, as well as a comparison with other codes within the PFS, consultation with physicians and health care professionals within CMS and the federal government, as well as Medicare claims data. We also assess the methodology and data used to develop the recommendations submitted to us by the RUC and other public commenters and the rationale for the recommendations. When we determine that the RUC's recommendations appropriately estimate the direct PE inputs (clinical labor, disposable supplies, and medical equipment) required for the typical service, are consistent with the principles of relativity, and reflect our payment policies, we use those direct PE inputs to value a service. If not, we refine the recommended PE inputs to better reflect our estimate of the PE resources required for the service. We also confirm whether CPT codes should have facility and/or nonfacility direct PE inputs and refine the inputs accordingly.

Our review and refinement of RUC-recommended direct PE inputs includes many refinements that are common across codes, as well as refinements that are specific to particular services. Table 28 details our finalized refinements of the RUC's direct PE recommendations at the code-specific level. In this final rule, we address several refinements that are common across codes, and refinements to particular codes are addressed in the portions of this section that are dedicated to particular codes. We note that for each refinement, we indicate the proposed impact on direct costs for that service. We note that, on average, in any case where the impact on the direct cost for a particular refinement is $0.32 or less, the refinement has no impact on the proposed PE RVUs. This calculation considers both the impact on the direct portion of the PE RVU, as well as the impact on the indirect allocator for the average service. We also note that nearly half of the proposed refinements listed in Table 28 result in changes under the $0.32 threshold and are unlikely to result in a change to the proposed RVUs.

We also note that the final direct PE inputs for CY 2017 are displayed in the CY 2017 direct PE input database, available on the CMS Web site under the downloads for the CY 2017 final rule at www.cms.gov/PhysicianFeeSched/. The inputs displayed there have also been used in developing the final CY 2017 PE RVUs as displayed in Addendum B.

b. Common Refinements

(1) Changes in Work Time

Some direct PE inputs are directly affected by revisions in work time. Specifically, changes in the intraservice portions of the work time and changes in the number or level of postoperative visits associated with the global periods result in corresponding changes to direct PE inputs. The direct PE input recommendations generally correspond to the work time values associated with services. We believe that inadvertent discrepancies between work time values and direct PE inputs should be refined or adjusted in the establishment of proposed direct PE inputs to resolve the discrepancies.

(2) Equipment Time

Prior to CY 2010, the RUC did not generally provide CMS with recommendations regarding equipment time inputs. In CY 2010, in the interest of ensuring the greatest possible degree of accuracy in allocating equipment minutes, we requested that the RUC provide equipment times along with the other direct PE recommendations, and we provided the RUC with general guidelines regarding appropriate equipment time inputs. We continue to appreciate the RUC's willingness to provide us with these additional inputs as part of its PE recommendations.

In general, the equipment time inputs correspond to the service period portion of the clinical labor times. We have clarified this principle over several years of rulemaking, indicating that we consider equipment time as the time within the intraservice period when a clinician is using the piece of equipment plus any additional time that the piece of equipment is not available for use for another patient due to its use during the designated procedure. For those services for which we allocate cleaning time to portable equipment items, because the portable equipment does not need to be cleaned in the room where the service is furnished, we do not include that cleaning time for the remaining equipment items, as those items and the room are both available for use for other patients during that time. In addition, when a piece of equipment is typically used during follow-up post-operative visits included in the global period for a service, the equipment time would also reflect that use.

We believe that certain highly technical pieces of equipment and equipment rooms are less likely to be used during all of the preservice or postservice tasks performed by clinical labor staff on the day of the procedure (the clinical labor service period) and are typically available for other patients even when one member of the clinical staff may be occupied with a preservice or postservice task related to the procedure. We also note that we believe these same assumptions would apply to inexpensive equipment items that are used in conjunction with and located in a room with non-portable highly technical equipment items since any items in the room in question would be available if the room is not being occupied by a particular patient. For additional information, we refer readers to our discussion of these issues in the CY 2012 PFS final rule with comment period (76 FR 73182) and the CY 2015 PFS final rule with comment period (79 FR 67639).

(3) Standard Tasks and Minutes for Clinical Labor Tasks

In general, the preservice, intraservice, and postservice clinical labor minutes associated with clinical labor inputs in the direct PE input database reflect the sum of particular tasks described in the information that accompanies the RUC-recommended direct PE inputs, commonly called the “PE worksheets.” For most of these described tasks, there are a standardized number of minutes, depending on the type of procedure, its typical setting, its global period, and the other procedures with which it is typically reported. The RUC sometimes recommends a number of minutes either greater than or less than the time typically allotted for certain tasks. In those cases, we review the deviations from the standards and any rationale provided for the deviations. When we do not accept the RUC-recommended exceptions, we refine the proposed direct PE inputs to conform to the standard times for those tasks. In addition, in cases when a service is typically billed with an E/M service, we remove the preservice clinical labor tasks to avoid duplicative inputs and to reflect the resource costs of furnishing the typical service.

In general, clinical labor tasks fall into one of the categories on the PE worksheets. In cases where tasks cannot be attributed to an existing category, the tasks are labeled “other clinical activity.” We believe that continual addition of new and distinct clinical labor tasks each time a code is reviewed under the misvalued code initiative is likely to degrade relativity between newly reviewed services and those with already existing inputs. This is because codes more recently reviewed would be more likely to have a greater number of clinical labor tasks as a result of the general tendency to increase the number of clinical labor tasks. To mitigate the potential negative impact of these additions, we review these tasks to determine whether they are fully distinct from existing clinical labor tasks, typically included for other clinically similar services under the PFS, and thoroughly explained in the recommendation. For those tasks that do not meet these criteria, we do not accept these newly recommended clinical labor tasks.

(4) Recommended Items That Are Not Direct PE Inputs

In some cases, the PE worksheets included with the RUC recommendations include items that are not clinical labor, disposable supplies, or medical equipment or that cannot be allocated to individual services or patients. We have addressed these kinds of recommendations in previous rulemaking (78 FR 74242), and we do not use items included in these recommendations as direct PE inputs in the calculation of PE RVUs.

(5) New Supply and Equipment Items

The RUC generally recommends the use of supply and equipment items that already exist in the direct PE input database for new, revised, and potentially misvalued codes. Some recommendations, however, include supply or equipment items that are not currently in the direct PE input database. In these cases, the RUC has historically recommended that a new item be created and has facilitated our pricing of that item by working with the specialty societies to provide us copies of sales invoices. For CY 2017, we received invoices for several new supply and equipment items. Tables 30 and 31 detail the invoices received for new and existing items in the direct PE database. As discussed in section II.A. of this final rule, we encourage stakeholders to review the prices associated with these new and existing items to determine whether these prices appear to be accurate. Where prices appear inaccurate, we encourage stakeholders to provide invoices or other information to improve the accuracy of pricing for these items in the direct PE database during the 60-day public comment period for this final rule. We expect that invoices received outside of the public comment period would be submitted by February 10th of the following year for consideration in future rulemaking, similar to our new process for consideration of RUC recommendations.

We remind stakeholders that due to the relativity inherent in the development of RVUs, reductions in existing prices for any items in the direct PE database increase the pool of direct PE RVUs available to all other PFS services. Tables 30 and 31 also include the number of invoices received, as well as the number of nonfacility allowed services for procedures that use these equipment items. We provide the nonfacility allowed services so that stakeholders will note the impact the particular price might have on PE relativity, as well as to identify items that are used frequently, since we believe that stakeholders are more likely to have better pricing information for items used more frequently. A single invoice may not be reflective of typical costs and we encourage stakeholders to provide additional invoices so that we might identify and use accurate prices in the development of PE RVUs.

In some cases, we do not use the price listed on the invoice that accompanies the recommendation because we identify publicly available alternative prices or information that suggests a different price is more accurate. In these cases, we include this in the discussion of these codes. In other cases, we cannot adequately price a newly recommended item due to inadequate information. Sometimes, no supporting information regarding the price of the item has been included in the recommendation. In other cases, the supporting information does not demonstrate that the item has been purchased at the listed price (for example, vendor price quotes instead of paid invoices). In cases where the information provided on the item allows us to identify clinically appropriate proxy items, we might use existing items as proxies for the newly recommended items. In other cases, we have included the item in the direct PE input database without any associated price. Although including the item without an associated price means that the item does not contribute to the calculation of the proposed PE RVU for particular services, it facilitates our ability to incorporate a price once we obtain information and are able to do so.

(6) Service Period Clinical Labor Time in the Facility Setting

Generally speaking, our proposed inputs did not include clinical labor minutes assigned to the service period because the cost of clinical labor during the service period for a procedure in the facility setting is not considered a resource cost to the practitioner since Medicare makes separate payment to the facility for these costs. We address proposed code-specific refinements to clinical labor in the individual code sections.

(7) Procedures Subject to the Multiple Procedure Payment Reduction (MPPR) and the OPPS Cap

We note that the public use files for the PFS proposed and final rules for each year display both the services subject to the MPPR lists on diagnostic cardiovascular services, diagnostic imaging services, diagnostic ophthalmology services and therapy services and the list of procedures that meet the definition of imaging under section 1848(b)(4)(B) of the Act, and therefore, are subject to the OPPS cap for the upcoming calendar year. The public use files for CY 2017 are available on the CMS Web site under downloads for the CY 2017 PFS final rule at http://www.cms.gov/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/PFSFederal-Regulation-Notices.html.

4. Specialty-Mix Assumptions for Proposed Malpractice RVUs

The final CY 2017 malpractice crosswalk table is displayed in the public use files for the PFS final rule. The public use files for CY 2017 are available on the CMS Web site under downloads for the CY 2017 PFS final rule at http://www.cms.gov/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/PFSFederal-Regulation-Notices.html. The table lists the CY 2017 HCPCS codes and their respective source codes used to set the final CY 2017 MP RVUs where the source code for this calculation deviates from the source code for the utilization otherwise used for purposes of PFS ratesetting. The MP RVUs for all PFS services and the utilization crosswalk used to identify the source codes for all other codes are reflected in Addendum B on the CMS Web site at https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/.

5. Valuation of Specific Codes

(1) Anesthesia Services Furnished in Conjunction With Lower Gastrointestinal (GI) Procedures (CPT Codes 00740 and 00810)

CPT codes 00740 and 00810 are used to report anesthesia furnished in conjunction with lower gastrointestinal (GI) procedures. In the CY 2016 PFS proposed rule (80 FR 41686), we discussed that in reviewing Medicare claims data, a separate anesthesia service is typically reported more than 50 percent of the time that various colonoscopy procedures are reported. We discussed that given the significant change in the relative frequency with which anesthesia codes are reported with colonoscopy services, we believed the relative values of the anesthesia services should be reexamined. We proposed to identify CPT codes 00740 and 00810 as potentially misvalued and sought public comment regarding valuation for these services.

The RUC recommended maintaining the base unit value of 5 as an interim base value for both CPT code 00740 and 00810 on an interim basis, due to their concerns about the specialty societies' surveys. The RUC suggested that the typical patient vignettes used in the surveys for both CPT codes 00740 and 00810 were not representative of current typical practice and recommended that the codes be resurveyed with updated vignettes. We stated in the CY 2017 proposed rule that we believed it premature to propose any changes to the valuation of CPT codes 00740 and 00810, continued to believe that these services are potentially misvalued, and sought additional input from stakeholders for consideration during future rulemaking.

Comment: Commenters were supportive of CMS' proposal to maintain the current values for CPT codes 00740 and 00810 for CY 2017. One commenter requested that CMS ensure that reimbursement for anesthesia services remains adequate to compensate providers for the cost of furnishing these services. Commenters also stated that due to greater complexity of furnishing anesthesia services compared to moderate sedation, payment for anesthesia services should not be lower than the values established for moderate sedation.

One commenter stated that CMS' perception that these codes are misvalued is related to the distinction between screening, diagnostic, and therapeutic endoscopies. The commenter further stated that there are no differences in the clinical risk and anesthesia preparation regardless of the indication for these procedures and suggested that the current base unit value of 5 units for CPT codes 00740 and 00810 is appropriate and should be maintained. Another commenter stated that the frequency of use of separate anesthesia services concurrent with colonoscopy procedures is not due to any potential misvaluation, but rather due to changes in Medicare coverage and payment policies that encourage Medicare beneficiaries to undergo screening colonoscopies.

Response: We appreciate the information provided by commenters. We continue to encourage feedback from interested parties and specialty societies, all of which we will take under consideration for future rulemaking.

(2) Soft Tissue Localization (CPT Codes 10035 and 10036)

In the CY 2016 PFS final rule with comment period, we established the RUC-recommended work value as interim final for CPT codes 10035 and 10036. We also made standard refinements to remove duplicative clinical labor and utilize standard equipment time formulas for the PACS workstation proxy (ED050).

Comment on the CY 2016 PFS final rule with comment period: A commenter stated that the clinical labor task “Review/read X-ray, lab, and pathology reports” occurs during the preservice period, and it is a separate activity than “Review examination with interpreting MD”, which occurs during the service period.

Response in the CY 2017 PFS proposed rule: We continued to believe that the clinical labor was duplicative with the clinical labor for “Review examination with interpreting MD” because we believed that the two descriptors detailed the same clinical labor activity taking place, rather than two separate and distinct tasks.

In the CY 2017 proposed rule, we proposed to maintain our previous refinement to 0 minutes for this clinical labor task for CPT codes 10035 and 10036. We also proposed to maintain the interim final work RVUs for CPT codes 10035 and 10036.

We did not receive any comments in response to our proposed valuation on CPT codes 10035 and 10036 and we are finalizing the clinical labor task and work RVUs as proposed.

(3) Removal of Nail Plate (CPT Code 11730)

We identified CPT code 11730 through a screen of high expenditures by specialty. The HCPAC recommended a work RVU of 1.10. We believed the recommendation for this service overestimates the work involved in performing this procedure, specifically given the decrease in physician intraservice and total time concurrently recommended by the HCPAC. We believed that a work RVU of 1.05, which corresponds to the 25th percentile of the survey results, more accurately represents the time and intensity of furnishing the service. To further support the validity of the use of the 25th percentile of the survey, we identified two crosswalk codes, CPT code 20606 (Arthrocentesis, aspiration and/or injection, intermediate joint or bursa), with a work RVU of 1.00, and CPT code 50389 (Removal of nephrostomy tube, requiring fluoroscopic guidance), with a work RVU of 1.10, both of which have identical intraservice times, similar total times and similar intensity. We noted that our proposed work RVU of 1.05 for CPT code 11730 falls halfway between the work RVUs for these two crosswalk codes. CPT code 11730 may be reported with add-on CPT code 11732 to report performance of the same procedure for each additional nail plate procedure.

Since CPT code 11732 was not reviewed by the HCPAC for CY 2017, we proposed a new work value to maintain the consistency of this add-on code with the base code, CPT code 11730. We proposed to remove 2 minutes from the physician intraservice time to maintain consistency with the HCPAC-recommended reduction of 2 minutes from the physician intraservice time period for the base code. We are using a crosswalk from the value for CPT code 77001 (Fluoroscopic guidance for central venous access device placement, replacement (catheter only or complete), or removal (includes fluoroscopic guidance for vascular access and catheter manipulation, any necessary contrast injections through access site or catheter with related venography radiologic supervision and interpretation, and radiographic documentation of final catheter position) (List separately in addition to code for primary procedure)), which has similar physician intraservice and total time values; therefore, we proposed a work RVU of 0.38 for CPT code 11732. As further support for this proposal, we noted that this proposed RVU reduction is similar to the value obtained by subtracting the incremental difference in the current and recommended work RVUs for the base code from the current value of CPT code 11732.

We proposed to use the HCPAC-recommended direct PE inputs for CPT code 11730. We proposed to apply some of the HCPAC-recommended refinements for CPT code 11730 to CPT code 11732, including the removal of the penrose drain (0.25 in × 4 in), lidocaine 1%-2% inj (Xylocaine), applicator (cotton-tipped, sterile) and silver sulfadiazene cream (Silvadene), as well as the reduction of the swab-pad, alcohol from 2 to 1. In addition, we proposed not to include the recommended supply items “needle, 30g, and syringe, 10-12ml” since other similar items are present, and we believe inclusion of these additional supply items would be duplicative. For clinical labor, we proposed to assign 8 minutes to “Assist physician in performing procedure” to maintain a reduction that is proportionate to that recommended for CPT code 11730. For the supply item “ethyl chloride spray,” we believed that the listed input price of $4.40 per ounce overestimates the cost of this supply item, and we solicited comment on the accuracy of this supply item price. Finally, we proposed to add two equipment items as was done in the base code, basic instrument pack and mayo stand, and proposed to adjust the times for all pieces of equipment to eight minutes to reflect the clinical service period time.

Comment: A commenter states that the work for CPT code 11730 has not changed since the previous recommendation, thus maintenance of a work RVU of 1.10 is proper.

Response: We continue to believe that the HCPAC-recommended reduction in intraservice and total time supports a reduction in our estimation of the physician work value of furnishing this service.

Comment: The HCPAC stated that it did not support the proposed decrease in the work RVU for CPT code 11732.

Response: We welcome any additional input as to the appropriate valuation of CPT code 11732. At this time, we continue to believe that a work RVU of 0.38 is appropriate, considering its relationship to CPT code 11730. We proposed values for CPT code 11732 based on its being an add-on code for CPT code 11730. We remind commenters and stakeholders that they may nominate this code family as potentially misvalued if they believe that both codes should be evaluated through the standard process, which would involve use of physician survey data and input from the HCPAC for both codes. We are finalizing work RVUs of 1.05 for CPT code 11730 and 0.38 for CPT code 11732, as well as the proposed PE refinements.

(4) Bone Biopsy Excisional (CPT Code 20245)

In CY 2014, CPT code 20245 was identified by the RUC's 10-Day Global Post-Operative Visits Screen.

For CY 2017, the RUC recommended a work RVU of 6.50 for CPT code 20245, including a change in global period from 10 to 0 days. We disagreed with this value given the significant reductions in the intraservice time, total time, and the change in the office visits assuming the change in global period. The intraservice and total times were decreased by approximately 33 and 53 percent respectively; while the elimination of three post-operative visits (one CPT code 99214 and two CPT code 99213 visits) alone would reduce the overall work RVU by at least 38 percent under the reverse building block methodology. We also note that the RUC-recommended work RVU of 6.50 only represents a 27 percent reduction relative to the previous work RVU of 8.95. To develop a work RVU for this service, we used a crosswalk from CPT code 19298 (Placement of radiotherapy after loading brachytherapy catheters (multiple tube and button type) into the breast for interstitial radioelement application following (at the time of or subsequent to) partial mastectomy, includes imaging guidance), since we believe the codes share similar intensity and total time and the same intraservice time of 60 minutes. Therefore, for CY 2017, we proposed a work RVU of 6.00 for CPT code 20245.

Comments: Several commenters, including the RUC, stated their objection to the proposed crosswalk, indicating that it underestimated the total time by 10 minutes and the physician work involved in furnishing the service. Commenters recommended CMS accept the RUC-recommended work RVU of 6.50.

The RUC also noted the current time of CPT code 20245 was based on a survey of 35 individuals more than 15 years ago and due to the previous flawed survey, the resulting IWPUT was almost zero. Given these discrepancies, the surveyed time of 60 minutes better reflects an appropriate level of intensity and complexity (IWPUT= 0.071) for this service relative to other 0-day global procedures.

Another commenter stated concern that the values proposed by CMS have been arrived at using methodologies that are not consistent with the RUC-recommended values, and therefore, are not appropriately relative to other similar services.

Response: Thank you for your comments. We present the information in Table 16 to illustrate the differences between the CMS crosswalked code and the additional RUC comparator codes.

Although the total times for CPT codes 19298 and 20245 are not identical, we continue to believe it is a more accurate comparison than the additional codes submitted by the RUC, which have 22-29 minutes less total time.

We note that according to the most recent survey, respondents lowered the work RVU of the 25th percentile, which we typically accept, from 6.06 RVUs to 4.94 RVUs when the code was revalued with a 0-day global period.

For CY 2017, we are finalizing the work RVU of 6.00 for CPT code 20245.

(5) Insertion of Spinal Stability Distractive Device (CPT Codes 22867, 22868, 22869, and 22870)

For CY 2016, the CPT Editorial Panel converted two Category III codes to Category I codes describing the insertion of an interlaminar/interspinous process stability device (CPT codes 22867 and 22869) and developed two corresponding add-on codes (CPT codes 22868 and 22870). The RUC recommended a work RVU of 15.00 for CPT code 22867, 4.00 for CPT code 22868, 7.39 for CPT code 22869, and 2.34 for CPT code 22870.

We believe that the RUC recommendations for CPT codes 22867 and 22869 overestimate the work involved in furnishing these services. We believe that a crosswalk to CPT code 36832 (Revision, open, arteriovenous fistula; without thrombectomy, autogenous or nonautogenous dialysis graft (separate procedure)), which has a work RVU of 13.50 is a more accurate comparison. CPT code 36832 is similar in total time, work intensity, and number of visits to CPT code 22867. This crosswalk is supported by the ratio between total time and work in the key reference service, CPT code 63047 (Laminectomy, facetectomy and foraminotomy (unilateral or bilateral with decompression of spinal cord, cauda equina and/or nerve root[s], [eg, spinal or lateral recess stenosis]), single vertebral segment; lumbar). Therefore, we proposed a work RVU of 13.50 for CPT code 22867. For CPT code 22869, we believed that CPT code 29881 (Arthroscopy, knee, surgical; with meniscectomy (medial OR lateral, including any meniscal shaving) including debridement/shaving of articular cartilage (chondroplasty), same or separate compartment(s), when performed) is an appropriate crosswalk based on clinical similarity, as well as intensity and total time. CPT code 29881 has a work RVU of 7.03; therefore, we proposed a work RVU of 7.03 for CPT code 22869. We proposed the RUC-recommended work RVU for CPT codes 22868 and 22870 without refinement.

Comment: Several commenters disagreed with our proposed valuation of the work RVU for CPT codes 22867 and 22869. They stated that the RUC crosswalk for each of these codes, respectively, is either identical to or a better match than the proposed CMS crosswalk.

Response: We recognize that the RUC crosswalk of CPT code 29915 for CPT code 22867 has a total time that is more similar to the new code than the crosswalk we proposed (CPT code 36832). We consider multiple factors when identifying appropriate crosswalk codes. We note that RUC's crosswalk, CPT code 29915, had very low service utilization, 355 in 2015, and was last reviewed by CMS and the RUC in April 2010. CPT code 36832, in contrast, had service utilization of 21,529 in 2015, and was most recently reviewed in October 2013. We considered the combination of these factors in choosing a crosswalk and determining a proposed work RVU. Commenters did not present any additional clinical information or data about this code that would lead us to reconsider our proposed valuation; therefore, we are finalizing the work RVU of 13.50 for CPT code 22867.

With regard to CPT code 22869, we disagree that the RUC crosswalk to CPT code 29880 is a closer comparison than CPT code 29881. The intraservice time for the newly created CPT code 22869 (43 minutes) is between that of the RUC recommended crosswalk CPT code 29880 (45 minutes) and the CMS crosswalk CPTcode 29881 (40 minutes). Total time for CPT code 29881, however, is identical to total time for CPT code 22869 (194 minutes), whereas the RUC recommended crosswalk CPT code 29880 has a higher total time (199 minutes). We continue to believe, therefore, that our crosswalk is appropriate and we are finalizing the proposed work RVU of 7.03 for CPT code 22869.

(6) Biomechanical Device Insertion (CPT Codes 22853, 22854, and 22859)

For CY 2016, the CPT Editorial Panel established three new Category I add-on codes and deleted one code to provide a more detailed description of the placement and attachment of biomechanical spinal devices. For CPT code 22853, the RUC recommended a work RVU of 4.88. For CPT codes 22854 and 22859, the RUC-recommended work RVUs are 5.50 and 6.00, respectively.

In reviewing the code descriptors, descriptions of work and vignettes associated with CPT codes 22854 and 22859, we concluded that the two procedures, in addition to having identical work time, contain many clinical similarities and do not have quantifiable differences in overall intensity. Therefore, we proposed the RUC-recommended work RVU of 5.50 for both CPT code 22854 and CPT code 22859. We believe that the RUC-recommended work RVU of 4.88 for CPT code 22853 overestimates the work in the procedure relative to the other codes in the family. We proposed a work RVU of 4.25 for CPT code 22853 based a crosswalk from CPT code 37237 (Transcatheter placement of an intravascular stent(s) (except lower extremity artery(s) for occlusive disease, cervical carotid, extracranial vertebral or intrathoracic carotid, intracranial, or coronary), open or percutaneous, including radiological supervision and interpretation and including all angioplasty within the same vessel, when performed; each additional artery (List separately in addition to code for primary procedure)), which is similar in time and intensity to the work described by CPT code 22853.

Comment: Several commenters disagreed with our proposed valuation of the work RVU of 4.25 for CPT code 22853 rather than the RUC-recommended work RVU of 4.88. They requested clarification regarding our crosswalk for this new code to CPT code 37237 instead of the RUC-recommended crosswalk of CPT code 57267.

Response: We take many factors into consideration when valuing a work RVU for a new code. We note that CPT code 57267 and CPT code 37237 have identical intraservice times and very similar total work times. We note that CPT code 37237 was most recently valued in April 2013, whereas the RUC crosswalk CPT code 57267 was last reviewed in 2004. We continue to believe that CPT code 37237 is an appropriate crosswalk for valuing the new CPT code 22859. Therefore, we are finalizing our proposed work RVU of 4.25 for CPT code 22853.

Comment: We received several comments objecting to our proposed work RVU of 5.50 for CPT code 22859, which is identical to the work RVU proposed by the RUC and accepted by CMS for CPT code 22854. Commenters provided detailed descriptions of the two procedures in an effort to demonstrate the higher intensity required by CPT code 22859 compared with CPT code 22854, thereby justifying the RUC-recommended work RVU of 6.00 for CPT code 22859. Several commenters expressed confusion about the descriptors for all three of the new CPT codes (CPT codes 22853, 22854, and 22859), in general, and stated their concern that the code descriptors do not clearly differentiate the work involved in furnishing the services.

Response: While we are somewhat persuaded by commenters' detailed descriptions of the two procedures and the higher intensity of work involved in furnishing CPT code 22859 compared with CPT code 22854, we are concerned about a substantive disagreement between the RUC and survey respondents about the intensities of work involved in furnishing the services described by these new codes. The RUC and the survey respondents valued the relative intensities of the two codes in the reverse order. The survey results indicated a work RVU of 8.16 (with 25th percentile of 7.0) for CPT code 22854 and a work RVU of 8.0 (with 25th percentile of 6.0) for CPT code 22859. The RUC reviewed the survey results and agreed that respondents overvalued the work involved in performing CPT code 22854. The RUC-recommended work RVU for CPT code 22854, which we are accepting as recommended, was established through a crosswalk to CPT code 37234. We agree that this is an appropriate crosswalk and valuation of this service. For CPT code 22859, the RUC also believed that the survey recommended work RVU of 8.0 was overvalued. The RUC recommended the 25th percentile of survey results, with a work RVU of 6.0. We find it difficult to reconcile the conflicting valuations by the survey and the RUC of the absolute and relative intensity of these new codes.

In addition to the survey results and RUC recommendations, we reviewed the descriptors of these codes and agree with commenters who found them vague and unclear. We share the concern of stakeholders who indicated that the lack of differentiation in the codes may lead to inconsistent use and reporting.

Given the disagreement between the RUC and survey respondents regarding the order and level of intensity of these services, along with confusion about the code descriptors, we find that valuing the services of 22854 and 22859 differently from each another is difficult to justify. Therefore, we are finalizing our proposed work RVU of 5.50 for CPT code 22859.

(7) Repair Flexor Tendon (CPT Codes 26356, 26357, and 26358)

In the CY 2016 PFS final rule with comment period, we established an interim final work RVU of 9.56 for CPT code 26356 after considering both its similarity in time to CPT code 25607 (Open treatment of distal radial extra-articular fracture) and the recommended reduction in time relative to the current times assumed for this procedure. We established an interim final work RVU of 10.53 for CPT code 26357 based on a direct crosswalk from CPT code 27654 (Repair, secondary, Achilles tendon, with or without graft), as we believed that this work RVU better reflected the changes in time for this procedure. For the last code in the family, we established an interim final work RVU of 12.13 for CPT code 26358, based on the RUC-recommended increment of 1.60 work RVUs relative to CPT code 26357.

Comment on the CY 2016 PFS final rule with comment period: We received several comments regarding the interim final work values for this family of codes. One commenter stated that it was inappropriate to use time ratios to evaluate CPT code 26356 as it was last valued in 1995, noting that there was an anomalous relationship between the current work RVU and the imputed time components in the RUC database. This commenter also pointed out that when the previous time was developed, fabrication of a splint was considered to be part of the intraservice work, while in the current survey instrument, the fabrication of the splint is considered to be part of the postservice work since it is a dressing. This commenter urged CMS to adopt the RUC recommendations. A different commenter agreed that the CMS crosswalk to CPT code 25607 was an appropriate crosswalk for CPT code 26356 and supported the CMS work RVU of 9.56.

Response in the CY 2017 PFS proposed rule: We appreciate the support from the commenter. We continue to believe that our crosswalk for this code is an appropriate choice, due to our estimate of overall work between CPT code 26356 and CPT code 25607. We appreciate the commenters' concerns regarding the time ratio methodologies and have responded to these concerns about our methodology in section II.L of this final rule. Although we note the commenter's statement about how the service period in which fabrication of a splint takes place may have evolved over time, we do not agree that this task would be responsible for a decrease in intraservice survey time, as the postservice survey time for CPT code 26356 remained unchanged at 30 minutes. If the decrease in intraservice time had been due to the shift of splinting from the intraservice period to the postservice period, then we would have expected to see an increase in the postservice period minutes. However, they remained exactly the same in the physician survey for CPT 26356. As we wrote earlier in this section, we believe in the validity of using pre-existing time values as a point of comparison, and we believe that we should account for efficiencies in time when the recommended work RVU does not account for those efficiencies. After consideration of comments received, we proposed to maintain CPT code 26356 at its current work RVU of 9.56 for CY 2017.

Comment on the CY 2016 PFS final rule with comment period: Several commenters disagreed with the work RVU for CPT code 26357. One commenter stated that the CMS crosswalk to CPT code 27654 had less total time and resulted in an inappropriately lower derived intensity. This commenter urged CMS to adopt the RUC-recommended work value. Another commenter stated that a better crosswalk for CPT code 26357 would be CPT code 25608 (Open treatment of distal radial intra-articular fracture or epiphyseal separation), the next code in the same upper extremity family that CMS used for the initial crosswalk. This commenter stated that the CMS crosswalk for CPT code 26357 created a rank order anomaly in terms of intensity within this family, and that the commenter's suggested crosswalk would create two pairs of matched codes, survey CPT codes 26356/26357 with crosswalk CPT codes 25607/25608.

Response in the CY 2017 PFS proposed rule: We appreciate the suggested crosswalk from the commenters, and we agree that the choice of the initial CMS crosswalk creates a rank order anomaly within the family in terms of intensity. As a result, after consideration of comments received, we proposed to instead value CPT code 26357 at the 25th percentile survey work RVU of 11.00 for CY 2017. This valuation corrects the anomalous intensity within the Repair Flexor Tendon family of codes, and preserves the RUC-recommended increment between CPT codes 26356 and 26357.

Comment on the CY 2016 PFS final rule with comment period: The commenters agreed that the RUC-recommended increment of 1.60 was appropriate for the work RVU of CPT code 26358 when added to the work RVU of CPT code 26357. However, commenters stated that this increment of 1.60 should be added to the RUC-recommended work value for CPT code 26357, and not the CMS refined value from the CY 2016 PFS final rule with comment period.

Response in the CY 2017 PFS proposed rule: We also continue to believe that the increment of 1.60 is appropriate for the work RVU of CPT code 26358. After consideration of comments received, we therefore proposed to set the work RVU for this code at 12.60 for CY 2017, based on the increment of 1.60 from CPT code 26357's proposed work RVU of 11.00.

In the CY 2017 proposed rule, we proposed to maintain the current direct PE inputs for all three codes.

The following is a summary of the comments we received regarding our proposed valuation of the Repair Flexor Tendon codes:

Comment: One commenter expressed support for the proposed work RVU for the flexor tendon codes.

Response: We appreciate the support from the commenters.

After consideration of comments received, we are finalizing our proposed valuation of the Repair Flexor Tendon codes.

(8) Closed Treatment of Pelvic Ring Fracture (CPT Codes 27197 and 27198)

For CY 2017, the CPT Editorial Panel deleted CPT codes 27193 and 27194 and replaced them with new CPT codes 27197 and 27198. The RUC recommended a work RVU of 5.50 for CPT code 27193, and a work RVU of 9.00 for CPT code 27198. We proposed to change the global period for these services from 90 days to 0 days because these codes typically represent emergent procedures with which injuries beyond pelvic ring fractures are likely to occur; we believe it is typical that multiple practitioners would be involved in providing post-operative care and it is likely that a practitioner furnishing a different procedure is more likely to be providing the majority of post-operative care. If other practitioners are typically furnishing care in the post-surgery period, we believe that the six post-service visits included in CPT code 27197, and the seven post-service visits included in CPT code 27198, would likely not occur. This is similar to our CY 2016 review and valuation of CPT codes 21811 (Open treatment of rib fracture(s) with internal fixation, includes thoracoscopic visualization when performed, unilateral; 1-3 ribs), 21812 (Open treatment of rib fracture(s) with internal fixation, includes thoracoscopic visualization when performed, unilateral; 4-6 ribs), and 21813 (Open treatment of rib fracture(s) with internal fixation, includes thoracoscopic visualization when performed, unilateral; 7 or more ribs). In our valuation of those codes, we determined that a 0-day, rather than a 90-day global period was preferable, in part because those codes describe rib fractures that would typically occur along with other injuries, and the patient would likely already be receiving post-operative care because of the other injuries. We believe that the same rationale applies here. To establish a work RVU for CPT code 27197, we proposed crosswalking this code to CPT code 65800 (Paracentesis of anterior chamber of eye (separate procedure); with removal of aqueous), due to its identical intraservice time and similar total time, after removing the work associated with postoperative visits, and its similar level of intensity. Therefore, we proposed a work RVU of 1.53 for CPT code 27197. For CPT code 27198, we proposed crosswalking this code to CPT code 93452 (Left heart catheterization including intraprocedural injection(s) for left ventriculography, imaging supervision and interpretation, when performed) which has an identical intraservice time and similar total time, after removing the work associated with post-operative visits from CPT code 27198. We proposed a work RVU of 4.75 for CPT code 27198.

Comment: Some commenters stated that the new coding for these services was designed, in part, to address the appropriateness of a 90-day global period by differentiating between higher energy and lower energy fractures. According to these commenters, the CPT Editorial Panel redefined these codes as treating injuries from higher energy and more unstable posterior pelvic ring injuries, and added a parenthetical directing physicians to use E/M billing for closed treatment of isolated lower energy fractures. These commenters say that the new coding clarifies when to use E/M coding for these services and when to bill these two codes. They state that these codes should thus remain valued with 90-day global periods while less complicated fractures will be billed with E/M coding.

Response: We took into consideration many factors when determining the appropriate global period of this service. While we understand that the new coding was partly designed to address the appropriateness of a 90-day global period, we continue to believe that a 0-day, rather than a 90-day, global period is more appropriate for this code, since we believe that the patient would likely already be receiving post-operative care because of other injuries. We also believe that the practitioner who performs the original procedure may not typically be performing the follow-up care, and shifting to a 0-day global period will allow the appropriate practitioner to report the follow up care, when appropriate.

Comment: A commenter stated that assigning a 0-day global period to this code will cause these codes to be different from all other closed fracture codes, which the commenter believes will lead to confusion for physicians and rank order anomalies.

Response: The commenter did not present sufficient information to explain why the variation in global periods for these kinds of services would uniquely cause rank order anomalies. We agree that it is preferable that codes for similar procedures have similar global periods; however, other factors specific to each code are taken into consideration when determining the appropriate global period. In the case of CPT codes 27197 and 27198, we continue to believe that the emergent nature inherent with the injuries considered typical would mean that other physicians would typically perform follow-up care. For detailed guidance on billing global surgical procedures, we direct readers to the Medicare Claims Processing Manual, Pub. 100-04; billing requirements and adjudication of claims requirements for global surgeries are under chapter 12, sections 40.2 and 40.4. We also note that if this procedure is billed concurrently with another procedure that is valued with a 10-day or 90-day global period, that the follow up visits associated with the latter procedure would occur as part of that package, while follow-up visits for these two codes would be reported using E/M coding.

Comment: A commenter states that, for procedures valued as part of a 90-day global period, the physician who is performing the primary portion of the treatment is obligated to follow the patient throughout the entire global period and furnish follow-up care.

Response: We understand the commenter's perspective that the treating physician is obligated to provide follow-up care within the global period; however, we do not believe that this necessitates the valuation of every surgical procedure with a 10-day or 90-day global period. While the treating physician would ideally provide follow-up care for these codes were they to be assigned 90-day global periods, we continue to believe that this would be an atypical situation for these types of treatments and for these types of injuries. We note that the assignment of a global period occurs in the process of evaluation of codes and we take into consideration factors specific to each procedure. There may be many instances when codes with similar procedures have different global periods. We are finalizing as proposed the work RVUs of 1.53 for CPT code 27197 and 4.75 for CPT code 27198, as well as an assignment of 0-day global periods.

(9) Bunionectomy (CPT Codes 28289, 28291, 28292, 28295, 28296, 28297, 28298, and 28299)

The RUC identified CPT code 28293 as a 90-day global service with more than 6 office visits and CPT codes 28290-28299 as part of the family of services. In October 2015, the CPT Editorial Panel created two new CPT codes (28291, 28295), deleted CPT codes 28290, 28293, and 28294 and revised CPT codes 28289, 28292, 28296, 28297, 28298 and 28299 based on the rationale that more accurate descriptions of the services needed to be developed.

For CPT codes 28289, 28292, 28296, 28297, 28298, and 28299, the RUC recommended and we proposed work RVUs of 6.90, 7.44, 8.25, 9.29, 7.75, and 9.29 respectively. For CPT code 28291, the RUC recommended a work RVU of 8.01 based on the 25th percentile of the survey. We believed the recommendation for this service overestimates the overall work involved in performing this procedure given the decrease in intraservice time, total time, and post-operative visits when compared to deleted predecessor CPT code 28293. Due to similarity in intraservice and total times, we believed a direct crosswalk of the work RVUs for CPT code 65780 (Ocular surface reconstruction; amniotic membrane transplantation, multiple layers) to CPT code 28291 more accurately reflects the time and intensity of furnishing the service. Therefore, for CY 2017, we proposed a work RVU of 7.81 for CPT code 28291.

For CPT code 28295, the RUC recommended a work RVU of 8.57 based on the 25th percentile of the survey. We believed the recommendation for this service overestimates the work involved in performing this procedure given the similarity in the intensity of the services and identical intraservice and total times as CPT code 28296. Therefore, we proposed a direct RVU crosswalk from CPT code 28296 to CPT code 28295. For CY 2017, we proposed a work RVU of 8.25 for CPT code 28295.

Comments: A few commenters, including the RUC, objected to the proposed work RVUs for CPT codes 28291 and 28295. Commenters noted that deleted CPT code 28293 was marked by the RUC as “not to use for validation of physician work”. The RUC noted the previous time was based on Harvard time and when reviewed in 1995, the RUC maintained the physician work and Harvard time because there was no compelling evidence to revise the value at that time.

The RUC acknowledged that the deleted CPT code 28293 had 30 minutes more intra-service time and a higher work RVU of 11.48 compared to the recommended work RVU of 8.01 for CPT code 28291. However, the RUC stated the differences in the physician work, time, intensity and the actual new service as described in CPT code 28291 were appropriately accounted for in its recommendation.

The RUC also stated disagreement with the proposed crosswalk of work RVUs from CPT code 28291 to CPT code 65780. The RUC stated it compared the family and relative ranking and believed CPT code 28291 was more complex and intense than CPT code 28298. The relative difference in work and complexity was reviewed and correctly ranked by the survey respondents at the 25th percentile, which corresponds with the RUC-recommended value.

One commenter stated that CPT code 28293 was deleted and a new CPT code was established because the two procedures were no longer synonymous. Also, the slight decrease in the intraoperative intensity with the new value is barely measurable, and therefore, the commenter does not agree with CMS that a work RVU of 7.81 is a more accurate valuation.

One commenter stated that CPT code 28295 is more intense than CPT code 28296 because CPT code 28295 requires separate areas of dissection. With CPT code 28296, the osteotomy and soft tissue procedure are performed at the same anatomic location. The commenter stated this nuance in complexity is the rationale for separate codes and is similar to the rationale for separate cervical versus lumbar spine codes or artery versus vein codes for vascular work.

Response: We appreciate additional information offered by the commenters. After consideration of comments received, we agreed with the additional information provided by commenters and are finalizing the RUC-recommend work RVUs of 6.90, 8.01, 7.44, 8.57, 8.25, 9.29, 7.75 and 9.29 for CPT codes 28289, 28291, 28292, 28295, 28296, 28297, 28298 and 28299; respectively.

(10) Endotracheal Intubation (CPT Code 31500)

In the CY 2016 PFS final rule with comment period (80 FR 70914), we identified CPT code 31500 as potentially misvalued. The specialty societies surveyed this code, and after reviewing the survey responses (which included increases in time) the RUC recommended a work RVU of 3.00 for CPT code 31500. After reviewing the RUC's recommendation, we proposed a work RVU of 2.66, based on a direct crosswalk to CPT code 65855 (Trabeculoplasty by laser surgery), which has similar intensity and service times.

Comment: Commenters requested that CMS finalize the RUC-recommended work RVU of 3.00 instead of CMS' proposed 2.66 work RVUs. The RUC stated that the surveyed median intraservice time is 10 minutes, representing a doubling of the current intraservice time of 5 minutes. Commenters also disagreed with CMS' proposed crosswalk from CPT code 65855. The RUC stated that given the emergent nature of the services reported with CPT code 31500, there are few relevant physician work and time-based comparisons within the resource-based relative value scale (RBRVS).

Response: We appreciate commenters' feedback on our proposal. As pointed out by the commenters, the survey data shows increased intraservice and total times for these services. We agree with commenters that due to the emergent nature of these services, there are few relevant physician work and time-based comparisons for this service. Therefore, due to the emergent nature of these services and service time increases, for CY 2017, we are finalizing a work RVU of 3.00 for CPT code 31500.

(11) Flexible Laryngoscopy (CPT Codes 31572, 31573, 31574, 31575, 31576, 31577, 31578, and 31579)

After we identified CPT codes 31575 and 31579 as potentially misvalued (80 FR 70912-70914), the RUC referred the entire flexible laryngoscopy family of codes back to the CPT Editorial Panel for revision and the addition of several codes representing new technology within this family of services. At the May 2015 CPT meeting, the CPT Editorial Panel added three new codes to describe laryngoscopy with ablation or destruction of lesion and therapeutic injection. Based on the survey results, the time resources involved in furnishing the procedures described by this code family experienced a significant reduction in the intraservice period, yet the recommended work RVUs were not similarly reduced. Therefore, in reviewing the recommended values for this family of codes we looked for a rationale for increased intensity and absent such rationale, proposed to adjust the recommend work RVUs to account for significant changes in time.

For CPT code 31575, we disagreed with the RUC-recommended work RVU of 1.00, and we instead proposed a work RVU of 0.94. We looked at the total time ratio for CPT code 31575, which is decreasing from 28 minutes to 24 minutes, and applied this ratio of 0.86 times the current work RVU of 1.10 to derive our proposed work RVU of 0.94. We supported this value for CPT code 31575 through a crosswalk to CPT code 64405 (Injection, anesthetic agent; greater occipital nerve), which shares 5 minutes of intraservice time and also has a work RVU of 0.94.

We agreed with the RUC that CPT code 31575 serves as the base code for the rest of the Flexible Laryngoscopy family. As a result, we proposed to maintain the same RUC-recommended increments for the rest of the codes in this family, measuring the increments from CPT code 31575's refined work RVU of 0.94 instead of the RUC-recommended work RVU of 1.00. This meant that each of the work RVUs for the codes in the rest of the family decreased by 0.06 when compared to the RUC-recommended value. We therefore proposed a work RVU of 1.89 for CPT code 31576, a work RVU of 2.19 for CPT code 31577, a work RVU of 2.43 for CPT code 31578, a work RVU of 3.01 for CPT code 31572, a work RVU of 2.43 for CPT code 31573, a work RVU of 2.43 for CPT code 31574, and a work RVU of 1.88 for CPT code 31579.

Regarding the direct PE inputs, we proposed to use refined clinical labor time for “Obtain vital signs” for CPT codes 31577 and 31579 from 3 minutes to 2 minutes. We believe that this extra clinical labor time is duplicative, as these codes are typically performed with a same day E/M service. Each procedure is only allotted a maximum of 5 minutes for obtaining vital signs, and since 3 minutes are already included in the E/M code, we proposed to reduce the time to 2 minutes for these services. Similarly, we proposed to remove the 3 minutes of clinical labor time for “Clean room/equipment by physician staff” from CPT codes 31575, 31577, and 31579. These procedures are typically reported with a same day E/M service, making the clinical labor minutes for cleaning the room in these procedure codes duplicative of the time already included in the E/M codes.

For CPT code 31572, we proposed to remove the “laser tip, diffuser fiber” supply (SF030) and replace it with the “laser tip, bare (single use)” supply (SF029) already present in our direct PE database. We believe that the invoice for SF030 submitted with the RUC recommendation is not current enough to establish a new price for this supply; as a result, we substituted the SF029 supply for this input. We welcomed the submission of new invoices to accurately price the diffuser fiber with laser tip.

We also proposed to make significant changes to the prices of several of the supplies and equipment related to Flexible Laryngoscopy, as well as to the prices of scopes more broadly. We proposed to set the price of the disposable biopsy forceps supply (SD318) at $26.84, based on the submission of an invoice with a price of $536.81 for a unit size of 20. In our search for additional information regarding scope inputs, we obtained a quote from a vendor listing the current price for several equipment items related to the use of scopes. Since we believe that the prices in vendor quotes would typically be equal to or higher than prices actually paid by practitioners, we are updating the prices in our direct PE database to reflect this new information. As part of this process, we proposed to increase the price of the “light source, xenon” (EQ167) from $6,723.33 to $7,000 to reflect current pricing information. We also proposed to adjust the price of the “fiberscope, flexible, rhinolaryngoscopy” (ES020) from $6,301.93 to $4,250.00.

In accordance with the wider proposal that we made involving the use of scope equipment, we proposed to separate the scopes used in these procedures from the scope video systems. In the course of researching different kinds of scopes, we obtained vendor pricing for two different types of scopes used in these procedures. We proposed to price the “rhinolaryngoscope, flexible, video, non-channeled” (ES063) at $8,000 and the “rhinolaryngoscope, flexible, video, channeled” (ES064) at $9,000 in accordance with our vendor quotes. We proposed to use the non-channeled scope for CPT codes 31575, 31579, and 31574 and the channeled scope for CPT codes 31576, 31577, 31578, 31572, and 31573 in accordance with the RUC-recommended video systems that stipulated channeled versus non-channeled scope procedures.

We believe that the “Video-flexible laryngoscope system” listed in the recommendations is not a new form of equipment, but rather constitutes a version of the existing “video system, endoscopy” equipment (ES031). We did not add a new equipment item to our direct PE database; instead, we proposed to use the submitted invoices to update the price of the ES031 endoscopy video system. As the equipment code for ES031 indicates, we proposed to define the endoscopy video system as containing a processor, digital capture, monitor, printer, and cart. We proposed to price ES031 at $15,045.00; this reflected a price of $2,000.00 for the monitor, $9,000.00 for the processor, $1,750.00 for the cart, and $2,295.00 for the printer. These prices were obtained from our vendor invoice, with the exception of the printer, which is a crosswalk to the “video printer, color (Sony medical grade)” equipment (ED036).

We did not agree that there is a need for multiple different video systems for this collection of Flexible Laryngoscopy codes based on our understanding of the clinical differences among the codes. In keeping with this understanding, we proposed to use the same existing “video system, endoscopy” equipment (ES031) for the remaining codes in the family that included RUC recommendations for new equipment items named “Video-flexible channeled laryngoscope system” and “Video-flexible laryngoscope stroboscopy system.” For CPT codes 31576, 31577, 31578, 31572, and 31573, we proposed to replace the Video-flexible channeled laryngoscope system with the existing endoscopy video system (ES031) along with a channeled flexible video rhinolaryngoscope (ES064). For CPT code 31579, we proposed to rename the RUC-recommended “Video-flexible laryngoscope stroboscopy system” to the shortened “stroboscopy system” (ES065) and assign it a price of $19,100.00. This reflected the price of the StrobeLED Stroboscopy system included on the submitted invoice. We proposed to treat the stroboscopy system as a scope accessory, which was included along with the “video system, endoscopy” equipment (ES031) and the “rhinolaryngoscope, flexible, video, non-channeled” (ES063) for CPT code 31579. When the price of the scope, the scope video system, and the stroboscopy system were summed together, the total proposed equipment price was $42,145.00.

We proposed to refine the recommended equipment times for several equipment items to conform to changes in clinical labor time. These are: The fiberoptic headlight (EQ170), the suction and pressure cabinet (EQ234), the reclining exam chair with headrest (EF008), and the basic instrument pack (EQ137). We proposed to use the standard equipment time formula for scope accessories for the endoscopy video system (ES031) and the stroboscopy scope accessory system (ES065). We also proposed to refine the equipment time for the channeled and non-channeled flexible video rhinolaryngoscopes to use the standard equipment time formula for scopes. For this latter pair of two new equipment items, this proposal resulted in small increases to their respective equipment times.

The following is a summary of the comments we received regarding our proposed valuation of the Flexible Laryngoscopy codes:

Comment: Several commenters disagreed with the proposed work RVU for CPT code 31575. Commenters stated that the use of a work/time ratio was inconsistent with the methodology of magnitude estimation, and that reducing work RVUs by mathematical formula can arbitrarily manipulate intensities without allowing input from survey recommendations provided by experts who perform the service. Commenters indicated their disapproval for a reverse building block methodology that assumes that if times for individual services change, work values must also change.

Response: We continue to believe that the use of these methodologies, including the use of time ratios, is an appropriate process for code valuation when recommended work RVUs do not appear to account for significant changes in time. As we stated earlier in our discussion on this topic in this final rule, we use time ratios to identify potential work RVUs and consider these work RVUs as potential options relative to the values developed through other methodologies for code valuation. We continue to believe that the decrease in total time for CPT code 31575 from 28 minutes to 24 minutes was not accounted for in the recommended work RVU, and as a result we proposed a work RVU of 0.94, supported by a crosswalk to CPT code 64405. We continue to believe that this valuation for CPT code 31575 more accurately captures the reduction in physician work caused by the decrease in the time required to perform the procedure, noting again that the statute specifically defines the work component as the resources in time and intensity required in furnishing the service. We believe that our crosswalk to CPT code 64405, which has very similar time and intensity values to CPT code 31575 at the same work RVU of 0.94, supports our valuation for this service.

Comment: Several commenters objected to the application of the work RVU increment to the rest of the codes in this family, measuring the increments from CPT code 31575's refined work RVU of 0.94 instead of the RUC-recommended work RVU of 1.00. Commenters stated that these codes were reviewed individually, not incrementally, and the use of an increment to reduce the work RVU of each code in the family by 0.06 was inappropriate. Commenters disagreed with the notion that when a base code's value is modified or reduced all other codes in the family should be reduced accordingly.

Response: We review codes individually for valuation. When we apply an increment from a base code to the rest of a code family, we do so only after reviewing each code individually and determining that the RUC-recommended relativity between the codes in the family is correct. For this particular family of codes, we stated our belief that the relativity between the codes in the family was accurate, and that the increment between the codes should be maintained after adjusting the work RVU for the base code (CPT code 31575) to account for its significant decrease in time. As we detailed in our discussion of code valuation methodologies earlier in this final rule, we use a variety of different methods, such as survey data, building blocks, crosswalks to key reference or similar codes, time ratios, and increments between codes within the same family. In our review of RUC-recommended values, we have observed that the RUC also uses a variety of methodologies to develop work RVUs for individual codes, and subsequently validates the results of these approaches through magnitude estimation or crosswalk to established values for other codes. We continue to believe that the use of an incremental methodology is the most accurate way to value this particular code family because it maintains the appropriate relativity among the Flexible Laryngoscopy codes.

Comment: One commenter disagreed with our refinement to remove the clinical labor time for “Clean room/equipment by physician staff” from the three codes in this family performed with a same day E/M service. The commenter stated that the clinical staff have to clean the equipment for procedure not used during the E/M service. According to the commenter, they clean that equipment separately and are assisting the physician during the entire procedure.

Response: In response to the commenter, we investigated this issue and determined that in the past we have sometimes provided 1 minute of clinical labor time for cleaning additional equipment beyond what would be cleaned during the E/M visit. As a result, we are restoring 1 minute of clinical labor time for “Clean room/equipment by physician staff” for CPT codes 31575, 31577, and 31579.

Comment: One commenter stated that there was a lack of clarity regarding the removal of the laser tip, diffuser fiber supply (SF030) from CPT code 31572. The commenter stated that the commenter supplied an invoice for the fibers, believed the invoice price was accurate, and believed the invoice should be utilized to set the price for this item.

Response: We continue to believe that the invoice for SF030 submitted with the RUC recommendation, which dates from 2009, is not current enough to establish a new price for this supply. We are continuing to maintain the laser tip, bare (single use) supply (SF029) in its place for CPT code 31572. As we discuss in the PE section of this final rule (II.A), we have concerns that the pricing for the laser tip, diffuser fiber supply has become outdated, and we are requesting the submission of additional current pricing information. We are maintaining the current pricing for this supply at $850 pending the submission of additional data.

We note as well that there were many comments addressing our proposal to reclassify scope equipment, as well as the proper pricing of the scope equipment utilized in this family of codes. These comments are summarized with responses in the PE section of this final rule (II.A).

After consideration of comments received, we are finalizing the work RVUs of the codes in the Flexible Laryngoscopy family at the proposed values. We are also finalizing the proposed direct PE inputs, with the exception of the refinement to the “Clean room/equipment by physician staff” clinical labor detailed above.

(12) Laryngoplasty (CPT Codes 31580, 31584, 31587, 31551-31554, 31591, and 31592)

CPT code 31588 (Laryngoplasty, not otherwise specified (e.g., for burns, reconstruction after partial laryngectomy)) was identified as potentially misvalued based on the RUC's 90-Day Global Post-Operative Visits screen. When this code family was reviewed by the RUC, it was determined that some codes in the family required revision to reflect the typical patient before a survey could be conducted and the code family was referred to the CPT Editorial Panel for revision. At its October 2015 meeting, the CPT Editorial Panel approved the creation of six new codes, revision of three codes, and deletion of three codes. For CPT codes 31580, 31587, 31551, 31552, 31553, 31554, and 31592, CMS proposed the RUC-recommended work RVUs.

For CPT code 31584, the RUC recommended a work RVU of 20.00. We believed that the 25th percentile of the survey, which is a work RVU of 17.58, better represents the time and intensity involved with furnishing this service based on a comparison with and assessment of the overall intensity of other codes with similar instraservice and total time. This value is also supported by a crosswalk code of CPT code 42844 (Radical resection of tonsil, tonsillar pillars, and/or retromolar trigone; closure with local flap (e.g., tongue, buccal)), which has identical intraservice time and identical total time. Therefore, we proposed a work RVU of 17.58 for CPT code 31584.

Comment: Several commenters requested that we provide an explanation for our proposed work RVU of 17.58 for the revised CPT code 31584 instead of the RUC-recommended work RVU of 20.00. They stated that the modified code now represents the combination of two previously separate CPT codes (the existing CPT code 31584 combined with CPT code 31600) and that the work RVU should better reflect the sum of the total time for these combined procedures. Commenters further noted that the proposed work RVU of 17.58 is lower, even, than the existing work RVU for CPT code 31584. A commenter requested that CMS consider two additional codes for comparison: CPT code 37660 and CPT code 43280.

Response: We take multiple factors into account when valuing a service that replaces two previously separate codes. We consider the efficiencies of combining two services, as reflected in the adjustment upwards of the intra-service and total time for this code. We also review the code description and identify a value that is consistent with other, similar, 90-day global codes. Our valuation is above the median work RVU for a group of 28 codes with similar intraservice and total time. Commenters have not provided any additional information that would suggest this code should be valued differently from other 90-day global codes with similar time and intensity.

We reviewed the two additional codes that commenters recommended as comparisons. We note that CPT code 43280 (work RVU of 18.1) was most recently valued in 1997 and that for low-volume code CPT code 37660, physician intensity is considerably higher than that for CPT code 31584, suggesting a poor reference for comparing the work involved in furnishing the service. For these reasons, we do not believe this code is an appropriate comparison for CPT code 31584 and we are finalizing our work RVU of 17.58 for CPT code 31584.

For CPT code 31591, the RUC recommended a work RVU of 15.60. We believed that the 25th percentile of the survey, which is a work RVU of 13.56, better represents the time and intensity involved with furnishing this service based on a comparison of the overall intensity of other codes with similar instraservice and total time. The 25th percentile of the survey is additionally bracketed by two crosswalk codes that we estimate have slightly lower and slighter higher overall intensities, CPT code 36819 (Arteriovenous anastomosis, open; by upper arm basilic vein transposition), which has a work RVU of 13.29, and CPT code 49654 (Laparoscopy, surgical, repair, incisional hernia (includes mesh insertion, when performed); reducible), which has a work RVU of 13.76; both of these codes have identical intraservice time and similar total time. Therefore, we proposed a work RVU of 13.56 for CPT code 31591.

Comment: Several commenters disagreed with our proposed work RVU of 13.56 for CPT code 31591, stating that the RUC-recommended work RVU of 15.60 better reflects the work required to perform the procedure.

Response: In developing our proposed valuation, we looked at other 90-day global codes with identical intraservice time and similar total time (between 275 and 335), and we note that the median work RVU of the resulting values (reflecting 33 codes) is 13.76. We chose the 25th percentile of the survey because of its closeness to the median work RVU of comparable services. We recognize that the RUC's crosswalk to CPT code 58544, with a work RVU of 15.60, has a lower total time than the codes we used as comparisons, but we note that this code has very low utilization, with 103 procedures billed in 2015. We continue to believe that two codes bracketing the 25th percentile of the work RVU for CPT code 31591 (CPT codes 36819 and 49654), as noted in the CY2017 PFS proposed rule, provide a better reference for valuing the new code, and that a work RVU of 13.56 adequately represents the time and intensity involved with furnishing the service. Therefore, we are finalizing our proposed work RVU of 13.56 for CPT code 31591.

Additionally, the RUC forwarded invoices provided by a medical specialty society for the video-flexible laryngoscope system used in these services. We discussed our proposed changes to the items included in equipment item ES031 (video system, endoscopy) in the CY 2017 proposed rule (81 FR 46247). Consistent with those proposed changes, we proposed to add a Nasolaryngoscope, non-channeled, to the list of equipment items used for CPT codes 31580, 31584, 31587, 31551-31554, 31591, and 31592, along with the modified equipment item ES031.

Comment: We received several comments, including from the RUC, about our proposal to implement a separate pricing approach for equipment inputs for this family of codes. Commenters requested a delay in implementing our approach until the RUC convened a PE subcommittee and provided CMS with specific recommendations for these codes.

Response: We appreciate the commenters' interests in making certain that there is appropriate opportunity for stakeholders to provide feedback and recommendations on the reclassification and pricing of scopes. Because these codes are currently under review, however, we believe that they should be valued according to a scheme that accurately describes the scope equipment typically used in the services. We continue to believe that our proposed classification system for scopes is the more proper methodology to use for valuation of these codes for the CY 2017. Please refer to II.A of the final rule for additional discussion on the new pricing process.

(13) Closure of Left Atrial Appendage With Endocardial Implant (CPT Code 33340)

The CPT Editorial Panel deleted category III CPT code 0281T (Percutaneous transcatheter closure of the left atrial appendage with implant, including fluoroscopy, transseptal puncture, catheter placement(s), left atrial angiography, left atrial appendage angiography, radiological supervision and interpretation) and created new CPT code 33340 to describe percutaneous transcatheter closure of the left atrial appendage with implant. The RUC recommended a work RVU of 14.00. We proposed a work RVU of 13.00 for CPT code 33340, which is the minimum survey result. Based on our clinical judgment and that the key reference codes discussed in the RUC recommendations have higher intraservice and total service times than the median survey results for CPT code 33340, we stated in the CY 2017 proposed rule that we believe a work RVU of 13.00 would more accurately represent the work value for this service.

Comment: We received several comments, including from the RUC. Commenters noted inaccuracies in CMS' description of the RUC recommendations including descriptions of the relationship between the RUC-recommended work RVU, survey results, and service times for the two key reference codes. Commenters requested that CMS finalize the RUC-recommended work RVU of 14.00.

Response: We appreciate the commenters' feedback and acknowledge that we inadvertently mischaracterized the RUC's recommendations related to this service. We agree that the survey results showed a 25th percentile survey result of 19.88 and that during the RUC meeting, this code was referred to the facilitation committee whereby the RUC identified two comparable codes with 14.00 work RVUs, which the RUC factored into its analysis and recommended valuation for this service. After consideration of the comments, we are finalizing the RUC-recommended work RVU of 14.00 for CPT code 33340.

(14) Valvuloplasty (CPT Codes 33390 and 33391)

The CPT Editorial Panel created new codes to describe valvuloplasty procedures and deleted existing CPT code 33400 (Valvuloplasty, aortic valve; open, with cardiopulmonary bypass). New CPT code 33390 represents a simple valvuloplasty procedure and new CPT code 33391 describes a more complex valvuloplasty procedure. We proposed to use the RUC-recommended values for CPT code 33390. For CPT code 33391, the RUC recommended a work RVU of 44.00, the 25th percentile survey result. The RUC estimated that approximately 70 percent of the services previously reported using CPT code 33400 would be reported using CPT code 33391, with 30 percent reported using new CPT code 33390. Therefore, the typical service previously reported with CPT code 33400 ought to now be reported with CPT code 33391. Compared to deleted CPT code 33400, the survey results for CPT code 33391 showed similar median intraservice times and decreased total times. Therefore, we proposed a work RVU of 41.50 for CPT code 33391, which is the current value of CPT code 33400. Given that the typical service should remain consistent between the two codes, we stated that we believe the work RVUs should remain consistent as well.

Comment: Commenters disagreed with CMS' proposed valuation of CPT code 33391, citing increased intensity and complexity of the procedures. Commenters noted that more complex patients are undergoing valvuloplasty (for instance, adult cardiac patients) when historically these patients would have received aortic valve replacements.

Response: As discussed in the CY 2017 proposed rule, the deleted CPT code 33400 is being replaced with two CPT codes that identify simple and complex procedures. The RUC's utilization crosswalk suggests that approximately 70 percent of the services that would previously have been reported using the combined code (CPT code 33400) would now be reported with CPT code 33391, the complex procedure. Based on the RUC's utilization crosswalk, the complex procedure would be the typical procedure reported under the combined code (CPT code 33400). The survey data for the complex procedure (CPT code 33391) showed similar median intraservice times and decreased total times compared to CPT code 33400. Therefore, for CY 2017, we are finalizing a work RVU of 35.00 for CPT code 33390 and a work RVU of 41.50 for CPT code 33391.

(15) Mechanochemical Vein Ablation (MOCA) (CPT Codes 36473 and 36474)

At the October 2015 CPT meeting, the CPT Editorial Panel established two Category I codes for reporting venous mechanochemical ablation, CPT codes 36473 and 36474. We proposed the RUC-recommended work RVU of 3.50 for CPT code 36473. For CPT code 36474, we proposed a work RVU of 1.75 and stated that we believed the RUC-recommended work RVU of 2.25 does not accurately reflect the typical work involved in furnishing this procedure. The specialty society survey showed that this add-on code has half the work of the base code (CPT code 36473). This value is supported by the ratio between work and time in the key reference service (CPT code 36476: Endovenous ablation therapy of incompetent vein, extremity, inclusive of all imaging guidance and monitoring, percutaneous, radiofrequency; second and subsequent veins treated in a single extremity, each through separate access sites (List separately in addition to code for primary procedure)).

The RUC-recommended direct PE inputs for CPT codes 36473 and 36474 included inputs for an ultrasound room (EL015). Based on the clinical nature of these procedures, we stated in our proposal that we do not believe that an ultrasound room would typically be used to furnish these procedures. We proposed to remove inputs for the ultrasound room and subsequently include a portable ultrasound (EQ250), power table (EF031), and light (EF014). The RUC also recommended that the ultrasound machine be allocated clinical staff time based on the PACS workstation formula. We stated that we did not believe that an ultrasound machine would be used like a PACS workstation, as images are generated and reviewed in real time. Therefore, we proposed to remove all direct PE inputs associated with the PACS workstation.

Comment: We received several comments, including from the RUC. Commenters disagreed with CMS' proposed work RVU of 1.75 for CPT code 36474 and requested that CMS finalize the RUC's recommendation of 2.25 work RVUs. The RUC disagreed with CMS' rationale for the proposed work RVU for CPT code 36474. The RUC stated that the ratio between CMS' proposed physician time and physician work for the survey code is 0.058, whereas that same ratio for the key reference code used by the RUC is 0.0883, and that the divergent ratios between the two services are not comparable.

Response: The commenters recommended that we accept the RUC-recommended ratio of 36 percent between the RUC-recommended work RVUs for CPT codes 36473 and 36474. We disagree. The RUC survey reported 79 minutes of total time for CPT code 36473 and 30 minutes of total time for CPT code 36474, a decrease of greater than 50 percent between the base code and the add-on code. As discussed in the proposed rule, our proposed work RVU of 1.75 for CPT code 36474 is supported by the ratio between work and time in the key reference service. The RUC recommendations made reference to two identical sets of services that use differing mechanisms for ablating the vein (radiofrequency procedures reported with CPT codes 36475 and 36476 (work RVUs of 5.30 and 2.65); laser procedures reported with CPT codes 36478 and 36479 (work RVUs of 5.30 and 2.65)). Both key reference code sets have a work RVU ratio of 50 percent (5.30 versus 2.65) between the base codes and the add-on codes. Therefore, for CY 2017, we are finalizing a work RVU of 3.50 for CPT code 36473 and a work RVU of 1.75 for CPT code 36474.

Comment: Commenters requested that CMS restore the direct PE inputs for the ultrasound room, which includes the PACS workstation. Commenters stated that the PACS workstation is needed for these procedures to store and make images available for future use.

Response: Commenters suggested that the ultrasound room was necessary for this procedure since the ultrasound room includes a PACS workstation that would allow for storage of the images and subsequent future use. As we discussed in the proposed rule, during the typical procedure, the images would be used in real time rather than being stored for subsequent interpretation. Further, the ultrasound room would not be typically used during these procedures. Our proposal included a portable ultrasound that allows for use of the images during the course of the procedure.

Comment: One commenter requested that CMS include an additional direct PE input for a ClariVein catheter for both CPT codes 36473 and 36474, and included invoices related to this item. The commenter suggested that an additional catheter is necessary to prevent contamination during treatment of subsequent vessels if the catheter used in an initial vessel were reused in a subsequent vessel.

Response: The invoice data submitted by the commenter appears to be applicable to the ClariVein catheters in some instances and in others to the ClariVein kits. Our review of the ClariVein kits indicated that the ClariVein catheters are part of the ClariVein kits. Because we lack clear product data regarding the cost of the ClariVein kits versus the ClariVein catheters and whether the catheters are included in the price of the kits, for CY 2017, we are finalizing our proposed direct PE inputs for the ClariVein kits for CPT codes 36473 and 36474 without modification. We welcome additional feedback from stakeholders regarding the product data and costs for the ClariVein catheters and ClariVein kits for consideration in future rulemaking.

(16) Dialysis Circuit (CPT Codes 36901, 36902, 36903, 36904, 36905, 36906, 36907, 36908, 36909)

In January 2015, a CPT/RUC workgroup identified the following CPT codes as being frequently reported together in various combinations: 35475 (Transluminal balloon angioplasty, percutaneous; brachiocephalic trunk or branches, each vessel), 35476 (Transluminal balloon angioplasty, percutaneous; venous), 36147 (Introduction of needle and/or catheter, arteriovenous shunt created for dialysis (graft/fistula); initial access with complete radiological evaluation of dialysis access, including fluoroscopy, image documentation and report), 36148 (Introduction of needle and/or catheter, arteriovenous shunt created for dialysis (graft/fistula); additional access for therapeutic intervention), 37236 (Transcatheter placement of an intravascular stent(s) (except lower extremity artery(s) for occlusive disease, cervical carotid, extracranial vertebral or intrathoracic carotid, intracranial, or coronary), open or percutaneous, including radiological supervision and interpretation and including all angioplasty within the same vessel, when performed; initial artery), 37238 (Transcatheter placement of an intravascular stent(s), open or percutaneous, including radiological supervision and interpretation and including angioplasty within the same vessel, when performed; initial vein), 75791 (Angiography, arteriovenous shunt (eg, dialysis patient fistula/graft), complete evaluation of dialysis access, including fluoroscopy, image documentation and report (includes injections of contrast and all necessary imaging from the arterial anastomosis and adjacent artery through entire venous outflow including the inferior or superior vena cava), radiological supervision and interpretation), 75962 (Transluminal balloon angioplasty, peripheral artery other than renal, or other visceral artery, iliac or lower extremity, radiological supervision and interpretation), and 75968 (Transluminal balloon angioplasty, each additional visceral artery, radiological supervision and interpretation). These codes are frequently reported together for both dialysis circuit services and transluminal angioplasty services. At the October 2015 CPT Editorial Panel meeting, the panel approved the creation of nine new codes and deletion of four existing codes used to describe bundled dialysis circuit intervention services, and the creation of four new codes and deletion of 13 existing codes used to describe bundled percutaneous transluminal angioplasty services (see discussion of the latter code family in the next section). The Dialysis Circuit family of codes overlaps with the Open and Percutaneous Transluminal Angioplasty family of codes (CPT codes 37246-37249), as they are both being constructed from the same set of frequently reported together codes. We reviewed these two families of codes concurrently to maintain relativity between these clinically similar procedures based upon the same collection of deleted codes.

For CPT code 36901, we proposed a work RVU of 2.82 instead of the RUC-recommended work RVU of 3.36. When we compared CPT code 36901 against other codes in the RUC database, we found that the RUC-recommended work RVU of 3.36 would be the highest value in the database among the 32 0-day global codes with 25 minutes of intraservice time. Generally speaking, we are particularly skeptical of RUC-recommended values for newly “bundled” codes that appear not to recognize the full resource overlap between predecessor codes. Since the recommended values would establish a new highest value when compared to other services with similar time, we believed it likely that the recommended value for the new code does not reflect the efficiencies in time. Of course, were there compelling evidence for this valuation accompanying the recommendation, we would consider such information. We also noted that the reference code selected by the survey participants, CPT code 36200 (Introduction of catheter, aorta), has a higher intraservice time and total time, but a lower work RVU of 3.02 We believe that there are more accurate CPT codes that can serve as a reference for CPT code 36901. As a result, we proposed to crosswalk CPT code 36901 to CPT code 44388 (Colonoscopy through stoma; diagnostic). CPT code 44388 has a work RVU of 2.82, and we believe it is a more accurate crosswalk for valuation due to its similar overall intensity and shared intraservice time of 25 minutes with 36901 and similar total time of 65 minutes.

We proposed a work RVU of 4.24 for CPT code 36902 instead of the RUC-recommended work RVU of 4.83. The RUC-recommended work RVU is based upon a direct crosswalk to CPT code 43253 (Esophagogastroduodenoscopy, flexible, transoral), which shares the same 40 minutes of intraservice time with CPT code 36902. However, CPT code 43253 has significantly longer total time than CPT code 36902, 104 minutes against 86 minutes, which we believe reduces its utility for comparison. We instead proposed to crosswalk the work RVU for CPT code 36902 from CPT code 44408 (Colonoscopy through stoma), which has a work RVU of 4.24. In addition to our assessment that the two codes share similar intensities, CPT code 44408 also shares 40 minutes of intraservice time with CPT code 36902 but has only 95 minutes of total time and matches the duration of the procedure under review more closely than the RUC-recommended crosswalk to CPT code 43253. We also note that the RUC-recommended work increment between CPT codes 36901 and 36902 was 1.47, and by proposing a work RVU of 4.24 for CPT code 36902, we would maintain a very similar increment of 1.42. As a result, we proposed a work RVU of 4.24 for CPT code 36902, based on this direct crosswalk to CPT code 44408. For CPT code 36903, we proposed a work RVU of 5.85 instead of the RUC-recommended work RVU of 6.39. The RUC-recommended value is based on a direct crosswalk to CPT code 52282 (Cystourethroscopy, with insertion of permanent urethral stent). Like the previous pair of RUC-recommended crosswalk codes, CPT code 52282 shares the same intraservice time of 50 minutes with CPT code 36903, but has substantially longer total time (120 minutes against 96 minutes) which we believe limits its utility as a crosswalk. We proposed a work RVU of 5.85 based on maintaining the RUC-recommended work RVU increment of 3.03 as compared to CPT code 36901 (proposed at a work RVU of 2.82), the base code for this family of related procedures. We also point to CPT code 44403 (Colonoscopy through stoma; with endoscopic mucosal resection) as a reference point for this value. CPT code 44403 has a work RVU of 5.60, but also lower intraservice time (45 minutes as compared to 50 minutes) and total time (92 minutes as compared to 96 minutes) in relation to CPT code 36903, suggesting that a work RVU a bit higher than 5.60 would be an accurate valuation. Therefore, we proposed a work RVU of 5.85 for CPT code 36903, based on an increment of 3.03 from the work RVU of CPT code 36901.

We proposed a work RVU of 6.73 instead of the RUC-recommended work RVU of 7.50 for CPT code 36904. Our proposed value comes from a direct crosswalk from CPT code 43264 (Endoscopic retrograde cholangiopancreatography), which shares the same intraservice time of 60 minutes with CPT code 36904 and has a higher total time. We also looked to the intraservice time ratio between CPT codes 36901 and 36904; this works out to 60 minutes divided by 25 minutes, for a ratio of 2.4, and a suggested work RVU of 6.77 (derived from 2.4 times CPT code 36901's work RVU of 2.82). This indicates that our proposed work RVU of 6.73 maintains relativity within the Dialysis Circuit family. As a result, we proposed a work RVU of 6.73 for CPT code 36904, based on a direct crosswalk to CPT code 43264.

We proposed a work RVU of 8.46 instead of the RUC-recommended work RVU of 9.00 for CPT code 36905. We looked at the intraservice time ratio between CPT codes 36901 and 36905 as one potential method for valuation, which is a 1:3 ratio (25 minutes against 75 minutes) for this case. This means that one potential value for CPT code 36905 would be triple the work RVU of CPT code 36901, or 2.82 times 3, which results in a work RVU of 8.46. We also investigated preserving the RUC-recommended work RVU increment between CPT code 36901 and 36905, which was an increase of 5.64. When this increment is added to the work RVU of 2.82 for CPT code 36901, it also resulted in a work RVU of 8.46 for CPT code 36905. Therefore, we proposed a work RVU of 8.46 for CPT code 36905, based on both the intraservice time ratio with CPT code 36901 and the RUC-recommended work increment with the same code.

For CPT code 36906, we proposed a work RVU of 9.88 instead of the RUC-recommended work RVU of 10.42. We based the proposed value upon the RUC-recommended work RVU increment between CPT codes 36901 and 36906, which is 7.06. When added to the work RVU of 2.82 for CPT code 36901, the work RVU for CPT code 36906 would be 9.88. We are supporting this value through the use of two crosswalks that both share the same 90 minutes of intraservice time with 36906. These are CPT code 31546 (Laryngoscopy, direct, with submucosal removal of non-neoplastic lesion(s) of vocal cord) at a work RVU of 9.73 and CPT code 61623 (Endovascular temporary balloon arterial occlusion, head or neck) at a work RVU of 9.95.

The final three codes in the Dialysis Circuit family are all add-on codes, which make comparisons difficult to the global 0-day codes that make up the rest of the family. We proposed a work RVU of 2.48 instead of the RUC-recommended work RVU of 3.00 for CPT code 36907. Due to the difficulty of comparing CPT code 36907 with the non-add-on codes in the rest of the Dialysis Circuit family, we looked instead to compare the value to the add-on codes in the Open and Percutaneous Transluminal Angioplasty family of codes (CPT codes 37246-37249). As we stated previously, both of these groups of new codes are being constructed from the same set of frequently reported together codes. We reviewed these two families of codes together to maintain relativity across the two families, and so that we could compare codes that shared the same global period.

We proposed the RUC-recommended work RVUs for all four codes in the Open and Percutaneous Transluminal Angioplasty family of codes. As a result, we compared CPT code 36907 with the RUC-recommended work RVU of 2.97 for CPT code 37249, which is also an add-on code. These procedures should be clinically very similar, since both of them are performing percutaneous transluminal angioplasty on a central vein, and both of them are add-on procedures. We looked at the intraservice time ratio between these two codes, which was a comparison between 25 minutes for CPT code 36907 against 30 minutes for CPT code 37249. This produces a ratio of 0.83, and a proposed work RVU of 2.48 for CPT code 36907 when multiplied with the RUC-recommended work RVU of 2.97 for CPT code 37249. We noted as well that the intensity was markedly higher for CPT code 36907 as compared to CPT code 37249 when using the RUC-recommended work values, which did not make sense since CPT code 36907 would typically be a clinically less intense procedure. Using the intraservice time ratio results in the two codes having exactly the same intensity. As a result, we therefore proposed a work RVU of 2.48 for CPT code 36907, based on this intraservice time ratio with the RUC-recommended work RVU of CPT code 37249.

For CPT code 36908, we disagree with the RUC-recommended work RVU of 4.25, and we instead proposed a work RVU of 3.73. We did not consider the RUC work value of 4.25 to be accurate for CPT code 36908, as this was higher than our proposed work value for CPT code 36902 (4.24), and we did not believe that an add-on code should typically have a higher work value than a similar non-add-on code with the same intraservice time. We identified two appropriate crosswalks for valuing CPT code 36908: CPT code 93462 (Left heart catheterization by transseptal puncture through intact septum or by transapical puncture) and CPT code 37222 (Revascularization, endovascular, open or percutaneous, iliac artery). Both of these codes share the same intraservice time as CPT code 36908, and both of them also have the same work RVU of 3.73, which results in these codes also sharing the same intensity since they are all add-on codes. We therefore proposed a work value of 3.73 for CPT code 36908, based on a direct crosswalk to CPT codes 93462 and 37222.

Finally, we proposed a work RVU of 3.48 for CPT code 36909 instead of the RUC-recommended work RVU of 4.12. The RUC-recommended value comes from a direct crosswalk from CPT code 38746 (Thoracic lymphadenectomy by thoracotomy). We compared the RUC-recommended work RVU for this procedure to other add-on codes with 30 minutes of intraservice time and found that the recommended work RVU of 4.12 would overestimate the overall intensity of this service relative to those with similar times. In reviewing the range of these codes, we believed that a more appropriate crosswalk is to CPT code 61797 (Stereotactic radiosurgery (particle beam, gamma ray, or linear accelerator)) at a work RVU of 3.48. We believed that this value is more accurate when compared to other add-on procedures with 30 minutes of intraservice time across the PFS. As a result, we proposed a work RVU of 3.48 for CPT code 36909 based on a direct crosswalk from CPT code 61797.

We proposed to use the RUC-recommended direct PE inputs for these nine codes with several refinements. We did not propose to include the recommended additional preservice clinical labor for CPT codes 36904, 36905, and 36906. The preservice work description is identical for all six of the global 0-day codes in this family; there is no justification given in the RUC recommendations as to why the second three codes need additional clinical labor time beyond the minimal preservice clinical labor assigned to the first three codes. We do not believe that the additional staff time would be typical. Patient care already would have been coordinated ahead of time in the typical case, and the need for unscheduled dialysis or other unusual circumstances would be discussed prior to the day of the procedure. We therefore proposed to refine the preservice clinical labor for CPT codes 36904, 36905, and 36906 to match the preservice clinical labor of CPT codes 36901, 36902, and 36903.

We proposed to refine the L037D clinical labor for “Prepare and position patient/monitor patient/set up IV” from 5 minutes to 3 minutes for CPT codes 36901-36906. The RUC recommendation included a written justification for additional clinical labor time beyond the standard 2 minutes for this activity, stating that the extra time is needed to prepare the patient's arm for the procedure. We agreed that extra time may be needed for this activity as compared to the default standard of 2 minutes; however, we proposed to assign 1 extra minute for preparing the patient's arm, resulting in a total of 3 minutes for this task. We did not believe that 3 extra minutes would be typically needed for arm positioning.

We proposed to remove the “kit, for percutaneous thrombolytic device (Trerotola)” supply (SA015) from CPT codes 36904, 36905, and 36906. We believed that this thrombolytic device kit and the “catheter, thrombectomy-Fogarty” (SD032) provide essentially the same supply, and the use of only one of them would be typical in these procedures. We believed that each of these supplies can be used individually for thrombectomy procedures. We proposed to remove the SA015 supply and retain the SD032 supply, and we solicited additional comment and information regarding the use of these two supplies.

We also proposed to remove the recommended supply item “covered stent (VIABAHN, Gore)” (SD254) and replace it with the “stent, vascular, deployment system, Cordis SMART” (SA103) for CPT codes 36903 and 36906. The Cordis SMART vascular stent was previously used in the past for CPT code 37238, which is the deleted code for transcatheter placement of an intravascular stent that CPT codes 36903 and 36906 are replacing. We did not have a stated rationale as to the need for this supply substitution, and therefore, we did not believe it would be appropriate to replace the current items with a significantly higher-priced item without additional information.

We also proposed to refine the quantity of the “Hemostatic patch” (SG095) from 2 to 1 for CPT codes 36904, 36905, and 36906. This supply was not included in any of the deleted base codes out of which the new codes are being constructed, and while we agreed that the use of a single hemostatic patch has become common clinical practice, we did not agree that CPT codes 36904-36906 would typically require a second patch. As a result, we proposed to refine the SG095 supply quantity from 2 to 1 for CPT codes 36904-36906, which also matches the supply quantity for CPT codes 36901-36903.

Included in the RUC recommendation for the Dialysis Circuit family of codes were a series of invoices for a “ChloraPrep applicator (26 ml)” supply. We solicited comments regarding whether the Betadine solution has been replaced by a Chloraprep solution in the typical case for these procedures. We also solicited comments regarding whether the “ChloraPrep applicator (26 ml)” detailed on the submitted invoices is the same supply as the SH098 “chlorhexidine 4.0% (Hibiclens)” applicator currently in the direct PE database.

Finally, we also solicited comments about the use of guidewires for these procedures. We requested feedback about which guidewires would be typically used for these procedures, and which guidewires are no longer clinically necessary.

The following is a summary of the comments we received regarding our proposed valuation of the Dialysis Circuit codes. Due to the large number of comments we received for this code family, we will first summarize the comments related to general code valuation, followed by the comments related to specific work RVUs, and finally the comments related to direct PE inputs.

Comment: Several commenters stated that the cumulative impact of reimbursement reductions for the Dialysis Circuit family of codes in physician work and practice expense would be quite dramatic. The commenters compared the total RVU of the old codes against the total RVU of the newly created codes and found a decrease of roughly 20-30 percent. Commenters expressed concern that if the proposed rates were to be implemented, many outpatient access centers that focus on providing care for ESRD patients might no longer be able to operate.

Response: We share the concern of the commenters in maintaining access to care for Medicare beneficiaries. We believe that improved payment accuracy under the PFS generally facilitates access to reasonable and necessary physicians' services.

We note that a change in overall RVUs for particular services, regardless of the magnitude of the change, may reflect improved accuracy. For example, comparing the summed total RVU of CPT codes 36147, 36148, 36870, and 37238 against the total RVU of CPT code 36906 is an accurate method to describe the services taking place under the coding schema effective for 2016 and 2017, respectively. Through the bundling of these frequently reported services, it is reasonable to expect that the new coding system will achieve savings via elimination of duplicative assumption of the resources involved in furnishing particular servicers. For example, a practitioner would not be carrying out the full preservice work four separate times for CPT codes 36147, 36148, 36870, and 37238, but preservice times were assigned to each of the codes under the old coding. We believe the new coding assigns a more accurate preservice time and thus reflects efficiencies in resource costs that existed regardless of how the services were previously reported.

Comment: Several commenters objected to the crosswalk codes used by CMS for proposed work valuation. Commenters stated that comparing the Dialysis Circuit codes to colonoscopy or endoscopic retrograde cholangiopancreatography (ERCP) codes was inappropriate, as it undervalued the technical skill and judgment necessary to furnish the services. In other words, the crosswalks chosen by CMS were invalid due to the differences in the procedures in question, with the Dialysis Circuit codes being more intensive procedures than the CMS crosswalks.

Response: We disagree with the commenters that the choice of crosswalk codes is inappropriate for work valuation. We believe that, generally speaking, codes with similar intensity and time values are broadly comparable across the PFS, as the fee schedule is based upon a relative value system. For the Dialysis Circuit codes in particular, we provided a specific rationale for each crosswalk detailing why we believed it to be an appropriate selection. Regarding the statement from the commenters that colonoscopy codes, such as CPT code 44388, are inappropriate for use as crosswalks in this family of codes, we note that the RUC-recommended work RVU for CPT code 36901 was based upon a direct crosswalk to the work RVU of a colonoscopy code (CPT code 45378). We continue to believe that the crosswalks for this family of codes are appropriate choices, since they share highly similar intensity and time values with the reviewed codes.

Comment: Some commenters disagreed with the use of time ratios for work valuation. These commenters stated that the use of direct crosswalks based only on intraservice time comparison or ratios of intraservice time inappropriately discounted the variation in technical skill, judgment, and risk inherent to these procedures.

Response: We continue to believe that the use of these methodologies, including the use of time ratios, is an appropriate process for identifying potential values for particular codes, especially when the recommended work RVUs do not appear to account for significant changes in time. As we stated earlier in our discussion on this topic in this final rule, we use time ratios to identify potential work RVUs and consider these work RVUs as potential options relative to the values developed through other methodologies for code valuation. We continue to believe our valuation for the Dialysis Circuit codes accurately captures the reduction in physician work caused by the efficiencies gained in both time and intensity through the bundling together of frequently reported services.

Comment: One commenter disagreed with the use of CMS comparisons between the RUC-recommended work RVUs for the Dialysis Circuit codes and the work RVU for other codes with similar time values in the rest of the fee schedule, particularly for CPT code 36901. The commenter stated that whether or not CPT code 36901 had the highest work RVU among other 0-day global codes with 25 minutes of intra-service time was irrelevant. The commenter pointed out that some code must be the highest value because the RBRVS represents a range of services of varying intensity. The commenter stated that CMS' reasoning undervalued the importance of work intensity in favor of the more easily quantifiable time variable, which was clinically inaccurate and contradictory to the principles of the relative value system.

Response: We disagree with the commenter about the invalidity of comparing newly created codes to existing codes with similar time values on the PFS. While it is true that there must be a highest value for any particular subset of codes, we believe the best approach in establishing work RVUs for codes is to compare the service to other services with similar times and identify codes with similar overall intensities. As we wrote in the proposed rule with regards to CPT code 36901, we have reservations with RUC-recommended values for newly “bundled” codes that appear not to recognize the full resource overlap between predecessor codes. Since the recommended values would establish a new highest value when compared to other services with similar time, we believe it likely that the recommended value for the new code does not reflect the efficiencies gained through bundling. We believe that these comparisons to other codes with similar time values and intensities are an important tool in helping to maintain relativity across the fee schedule.

Comment: A commenter disagreed with the CMS valuation for these codes based on a clinical rationale pertaining to how the services are defined. The commenter stated that the dialysis access circuit is defined as originating in the artery adjacent to the arterial anastomosis and including all venous outflow (whether single or multiple veins) to the axillary-subclavian vein junction. While several different arteries and veins may be included in this definition, from a functional perspective it is a single “vessel”. The commenter stated that because of this greater propensity for multiple lesions in these procedures, it is appropriate to define the access vessel as CPT has done and allow reporting of only a single angioplasty or stent in that entire conduit. However, the commenter reported that the survey built on the “typical patient” (51 percent of the cases) was unable to recognize the additional work of additional angioplasty or stent for the Dialysis Circuit family of codes, even though multiple or arterial lesions occur with significant frequency. Because the coding structure of the Dialysis Circuit family does not include a code for “additional vessels”, the valuation of the codes needs to incorporate the resource cost of patient cases where multiple or arterial lesions occur. The commenter contended that this problem with the survey methodology affected the work intensity of these codes, and justifies a higher intensity for these procedures.

Response: We share the commenter's concerns with the survey data collected by the RUC. This is why we have long employed different approaches to identify potential values for work RVUs, such as time ratios, building blocks, and crosswalks to key reference or similar codes, in addition to the recommended survey data. We also note that our methodology generally values services based on assumptions regarding the typical case, not occasional complications that may require additional work when they occur. For the particular case of the Dialysis Circuit family of codes, we do not agree with the commenter that the single “vessel” classification of these procedures supports a higher intensity compared to other related codes. These codes have been defined by CPT in a similar fashion to the lower extremity revascularization codes, in which the code is only billed a single time regardless of the number of lesions or number of stents placed. Due to the similarity with these existing codes located elsewhere in the PFS, we do not believe that it would be appropriate to value the Dialysis Circuit codes differently.

Comment: Several commenters suggested that there was compelling evidence for the higher RUC-recommended work RVUs because the vignette developed by the CPT Editorial Panel does not accurately reflect the typical ESRD patient. Commenters stated that the vignette for the Dialysis Circuit codes significantly underestimated the age of the typical patient, and may have led survey respondents to report less time. According to commenters, the frail and elderly ESRD patients that constitute the typical patients for these procedures are much sicker than the typical patient in other codes on the PFS, and this serves to justify valuing these codes at a higher intensity.

Response: We appreciate the submission of additional information regarding the patient population for these codes. We recognize that some services may require additional work due to an unusually difficult patient population. However, we do not agree at this time that the Dialysis Circuit family of codes has a uniquely different patient population that justifies an increase in valuation over other comparable codes on the PFS. We note that for CPT code 36901, the RUC recommended a work RVU of 3.36 based on a direct crosswalk to CPT code 45378, a flexible colonoscopy code. Our proposed work RVU of 2.82 for the same code was based on a direct crosswalk to CPT code 44388, which is another colonoscopy code. The patient population for these two crosswalk codes is similar, and both codes share similar time and intensity. We believe that our crosswalk code is a more appropriate choice given the time values and the efficiencies gained from bundling. However, based on this recommended crosswalk code, we believe that the RUC considers the patient population for CPT code 45378 to be appropriate for comparison to CPT code 36901, and that the reviewed code does not possess an unusually resource-intensive patient population. This same pattern holds true for the other codes in the Dialysis Circuit family, which were valued using similar comparisons to established codes with typical patient populations.

Comment: One commenter suggested that the difficulties posed by the patient population for the Dialysis Circuit codes were not sufficiently reflected in the RUC recommendations. The commenter stated that the patients receiving dialysis circuit services are extremely sick, and every step in the process of caring for those patients is more complex than those involved in caring for the average Medicare patient. The commenter stated that CMS underestimated the amount of time required to perform specific tasks and assumes that those tasks can be performed by individuals with lower levels of training and credentials than are used in typical practice. The commenter requested a series of direct PE refinements to this family of codes, many of which went above the original RUC recommendations, including clinical labor times significantly above the usual standards and using clinical labor staffing types outside the normal range. The commenter stated an intention to present data to support the recommendations at a later date.

Response: We appreciate the additional information provided by the commenter about this family of codes. We emphasize that we do not believe that the RUC need be the exclusive source of information used in valuation of PFS services, and we are supportive of the submission of additional data that can aid in the process of determining the resources that are typically used to furnish these services. Because we did not receive data from the commenter to support these increases above the RUC recommendations, we are not incorporating these changes into the Dialysis Circuit codes at this time. However, we urge interested stakeholders to consider submitting robust data regarding costs for these and other services.

We are also seeking information on how to reconcile situations where we have multiple sets of recommendations from the RUC and from other PFS stakeholders, both for this specific case and for the situation more broadly, given the need to maintain relativity among PFS services.

The following comments address the proposed work valuation of individual codes in the family.

Comment: A commenter contended that the proposed work RVU of 2.82 undervalues CPT code 36901. The commenter stated that compelling evidence regarding CPT's inaccurate description of the typical ESRD patient as 45 years old led to lower survey times and hence the “new highest value” problem mentioned by CMS. The commenter recommended that CMS should finalize the RUC work RVU of 3.36, or barring that, should finalize a work RVU of 3.02 based on a direct crosswalk to CPT code 36200. The commenter stated that this code is very similar clinically in work and intensity to CPT code 36901.

Response: We summarized and responded to the general issues surrounding patient populations above. We disagree with the commenter that CPT code 36200 is a more appropriate choice for a crosswalk code for CPT code 36901. CPT code 36200 has 5 additional minutes of intraservice time (30 minutes as compared to 25 minutes) and 25 additional minutes of total time (91 minutes as compared to 66 minutes). In addition to this substantial difference in time values, the intensity of CPT code 36200 is also significantly lower than CPT code 36901. If we were to adopt the recommended crosswalk to a work RVU of 3.02, the intensity of CPT code 36901 would be 50 percent higher than the intensity CPT code 36200. Since we are statutorily obligated to base our valuation on time and intensity, we believe that this makes CPT code 36200 an inferior choice for a crosswalk code when compared to our choice of CPT code 44388, which shares very similar time and intensity with CPT code 36901.

Comment: A commenter stated that CPT code 36902 should have a higher increment in work RVU from CPT code 36901 because it included work unable to be accounted for in a survey on the typical patient. The commenter indicated that according to published literature, more than one stenosis is present requiring angioplasty in 20-30 percent of dialysis access cases. A higher increment in work RVU from CPT code 36901 to 36902 would reflect the work of additional angioplasty on separate stenoses and arterial angioplasty that occurs in some cases, but cannot be reflected in a “typical” 51 percent case vignette. The commenter requested that CMS adopt the RUC-recommended work RVU for CPT code 36902.

Response: We generally establish RVUs for services based on the typical case. If a particular patient case requires treatment outside the defined dialysis circuit code descriptor, then additional catheter placement and imaging may be reported, assuming that all of the proper requirements for separate billing are met. We do not believe that it would be appropriate to increase the work RVU for CPT code 36902 based on these non-typical situations.

Comment: A commenter stated that the CMS proposed work RVU of 5.85 undervalues the work involved in the services described by CPT code 36903, based on the belief that the CPT patient vignette does not reflect the typical patient and that “additional vessel” angioplasty or stenting work is included in CPT code 36903 but was not able to be captured in a survey utilizing the “typical” patient.

Response: We addressed these issues in previous comment responses. We continue to believe that the proposed work RVU for CPT code 36903 is accurate.

Comment: A commenter disagreed with the proposed work RVUs for CPT codes 36904 and 36905. The commenter suggested that CMS should use time ratios from the base code in the family, CPT code 36901, starting from a work RVU of 3.02 instead of the proposed work RVU of 2.82. The commenter suggested that this would produce work RVUs for CPT codes 36904 and 36905 almost identical to the RUC-recommended values, which the commenter urged CMS to adopt.

Response: We agree with the commenter that the use of time ratios is one potential method to use in the process of determining code valuation. However, since we stated previously that we believe our proposed work RVU of 2.82 is more accurate for CPT code 36901 than the commenter's suggestion of 3.02, we do not believe that applying the same time ratios provides a rationale for adopting the RUC-recommended work RVUs for CPT codes 36904 and 36905.

Comment: A commenter disagreed with CMS' proposed work RVU of 9.88 for CPT code 36906 based upon the RUC-recommended increment of 7.06 from CPT code 36901. The commenter stated that the RUC value was well supported as the 25th percentile survey result and the survey times for the code were adversely impacted by CPT errors in the code descriptor and RUC survey limitations.

Response: We do not agree that the RUC's work valuation for CPT code 36906 maintains relativity within the fee schedule. We believe that the increment between CPT code 36901 and 36906 maintains relativity within the Dialysis Circuit family of codes, which is why we proposed to use it for valuation. However, we believe that the recommended work RVU for CPT code 36906 insufficiently accounted for the efficiencies in resource use achieved through bundling together its predecessor codes. We continue to believe that the proposed work RVU of 9.88, bracketed between crosswalks to CPT codes 31546 and 61623, provides the most accurate valuation for this service.

Comment: A commenter disagreed with the proposed work RVU of 2.48 for CPT code 36907, and stated that the work RVU should be identical to CPT code 37249 at a value of 2.97. The commenter stated these two services are clinically identical, and the CMS contention that CPT code 36907 would typically be a clinically less intense procedure is not correct. According to the commenter, the intensity involved in both of these add-on codes is the work and risk of crossing the central venous stenosis and performing intervention within the thorax where complications could be severe. The commenter stated that there is no difference in this work intensity based upon the direction of approach—from the dialysis access or from a native (femoral) vein. Both require advancing a long wire from the access site through the stenosis, superior and inferior vena cava, and right atrium, which is needed no matter which direction one is approaching the lesion. As a result, the commenter suggested that CPT code 36907 should have the same work RVU as CPT code 37249.

Response: While we agree with the commenter that these two services are clinically similar procedures, we do not agree with the commenter that the work between the two is identical. In particular, we believe that the difference in the intraservice time (25 minutes for CPT code 36907 against 30 minutes for CPT code 37249) should be accounted for in the work valuation, as the former code takes 20 percent less time to perform. We note as well that under our proposed valuation, these two codes have exactly the same intensity, with the difference in the work value occurring solely as a result of the decreased time required to perform CPT code 36907. Since time is one of the resources we are obligated to use for code valuation, we believe that the proposed values for these two codes are more accurate than setting both of them to the same work RVU.

Comment: One commenter supported the proposed work RVUs of 3.73 for CPT code 36908 and 3.48 for CPT code 36909.

Response: We appreciate the support from the commenter.

The following comments address the proposed direct PE inputs for the Dialysis Circuit family of codes.

Comment: Several commenters urged CMS to accept the recommended additional preservice clinical labor for CPT codes 36904, 36905, and 36906. Commenters stated that the patient presentation and the requisite preservice clinical labor is inherently different for CPT codes 36904-36906 when compared with CPT codes 36901-36903. Commenters indicated that the latter group are elective procedures, which are scheduled and planned well in advance of the procedure and performed on days that do not conflict with the patient's dialysis schedule. In contrast, the former group are urgent procedures typically done when a patient presents to their dialysis treatment with a thrombosed access. According to the commenters, the urgent nature of these procedures, the need for additional preoperative testing because of missed dialysis, and the need for arranging unscheduled dialysis treatment requires additional preservice time for the procedural staff.

Response: We disagree with the commenters. We continue to note that the preservice work description is identical for all six of the 0-day global codes in this family. Generally speaking, we also typically provide less preservice clinical labor time for emergent procedures, not more preservice clinical labor time, as there is no time for these tasks to be performed. We continue to believe that all six of these codes are most accurately valued by sharing the same preservice clinical labor times.

Comment: Several commenters stated that the recommended 5 minutes of clinical labor for “Prepare and position patient/monitor patient/set up IV” were reasonable because these cases are done on the upper extremity using portable c-arm fluoroscopy. According to commenters, the additional time includes prepping and positioning the arm, applying appropriate shielding to the patient's torso, positioning the c-arm unit, and then positioning other radiation shielding devices. Commenters stated that each of these activities requires more time in the arm, which typically must be extended to the side to be accessible for access and imaging; this is different from procedures done in the long plane of the body including the torso and legs. The commenters stated that 5 minutes is a more accurate reflection of the required clinical labor time than the proposed 3 minutes.

Response: We continue to believe that additional time may be needed for this activity as compared to the default standard of 2 minutes. However, we maintain that the commenter's request for 3 additional minutes (for a total of 5 minutes) would not typically be required for arm positioning, as this additional clinical labor time is generally not included in similar procedures. We do not agree that the additional tasks described by the commenters would require the requested 5 minutes of clinical labor time, and we are maintaining our proposed value of 3 minutes.

Comment: Several commenters opposed the CMS proposal to remove the “kit, for percutaneous thrombolytic device (Trerotola)” supply (SA015) from the RUC recommended supplies for CPT code 36904, 36905, and 36906, under the belief that only one device would typically be used in these procedures. Commenters indicated that this understanding was incorrect. According to the commenters, a mechanical thrombectomy device and a Fogarty thrombectomy balloon serve different purposes and both are necessary to perform a dialysis access thrombectomy. Commenters provided lengthy clinical rationales to support their point of view, which can be summarized as follows: “The Fogarty balloon is small and highly compliant allowing it to be pulled through the artery and into the access without damaging the vessels. The thrombectomy device cannot be used safely for this function. This device is larger so risks pushing the fibrin plug into the artery if passed across the arterial anastomosis from the access—risking distal arterial embolization. The device is also much more rigid being made from metal and with irregular shape that risks damaging the endothelium of the artery causing arterial injury.” As a result, commenters requested that the listed devices “catheter, thrombectomy-Fogarty” (SD032) and “kit, for percutaneous thrombolytic device (Trerotola)” supply (SA015) both remain in the supply list for these codes.

Response: We appreciated the detailed presentation of additional clinical information regarding the use of the percutaneous thrombolytic device kit from the commenters. After review of the comments and the contents of the kit, we believe that its inclusion in these three procedures is appropriate. According to the device literature, the kit contains a rotor for macerating the clot, a catheter for removing the clot, and a sheath for introducing the device. We will therefore restore the SA015 supply to CPT codes 36904, 36905, and 36906. However, we are removing the Fogarty catheter (SD032) and 1 of the 2 vascular sheaths (SD136), as these are contained within the kit. The literature for the percutaneous thrombolytic device kit clearly stipulates that there is no need for additional catheters to remove the clot, which makes the Fogarty catheter a duplicative supply which can be removed.

Comment: Several commenters disagreed with the CMS proposal to remove the recommended supply item “covered stent (VIABAHN, Gore)” (SD254) and replace it with the “stent, vascular, deployment system, Cordis SMART” (SA103) for CPT codes 36903 and 36906. Commenters stated that covered stents are the only stent devices that are FDA approved and supported by evidence from randomized controlled trials for use in dialysis access procedures. They are typically used in recurrent or elastic stenosis in dialysis access and have become the standard of care for these interventions. One commenter stated that Braid Forbes Health Research analyzed stent use in CPT code 37238 using CMS OPPS claims data, and found that the covered stent (VIABAHN, Gore), was used 67.5 percent of the time and the SA103, stent, vascular, deployment system, Cordis SMART, was used 32.5 percent of the time. Commenters stressed that bare metal stents, such as the Cordis SMART, are not indicated for use in the Dialysis Circuit procedures.

Response: We appreciate the submission of this additional clinical information regarding the use of stents for these procedures. After consideration of the comments, we are restoring the covered stent (VIABAHN, Gore) (SD254) to CPT codes 36903 and 36906 as originally recommended. Because we are including the SD254 covered stent, we are not adding the stent, vascular, deployment system, Cordis SMART (SA103) supply to these procedures.

Comment: Several commenters disagreed with the CMS proposal to reduce the quantity of the Hemostatic patch (SG095) from 2 to 1 for CPT codes 36904, 36905, and 36906. Commenters stated that two hemostatic patches are necessary in these procedures because they require two separate cannulations and sheaths. At the end of the case, both sheath sites are removed and covered with a hemostatic patch which aids in preventing bleeding and maintaining sterility. The commenters stressed that because there are two access sites, two hemostatic patches are required, one to cover each site.

Response: We appreciate the additional clinical information submitted by the commenters. In response to this information, we are finalizing inclusion of the second Hemostatic patch (SG095) to CPT codes 36904, 36905, and 36906, as recommended by the RUC.

Comment: In response to the CMS solicitation of feedback regarding the Chloraprep applicator (26 ml) supply, commenters indicated that Chloraprep solution has replaced Betadine solution when performing sterile preparation of the dialysis access circuit due to its greater efficacy as preoperative skin prep. Commenters indicated that this supply was most accurately represented by the submitted invoice. Another commenter stated that studies have shown that preparation of central venous sites with a 2% aqueous chlorhexidine gluconate (in 70% alcohol) is superior for skin site preparation to either 10% povidone-iodine or 70% alcohol alone, and that in 2002, the CDC recommended that 2% chlorhexidine be used for skin antisepsis prior to catheter insertion. One commenter recommended that CMS replace the Betadine povidone soln (SJ041) with two units of swab, patient prep, 3.0 ml (Chloraprep) supply (SJ088) in the inputs for CPT codes 36901-36906.

Response: We appreciate the submission of additional clinical information regarding the Chloraprep supply from the commenters. We agree with the recommended supply substitution, and we are therefore removing 60 ml of the Betadine solution (SJ041) and replacing it with two units of the swab, patient prep, 3.0 ml (Chloraprep) supply (SJ088) for CPT codes 36901-36906. We will add the Chloraprep applicator (26 ml) supply to the direct PE input database at a price of $8.48 based on an average of the three submitted invoices; it is not currently assigned to any codes. We also agree that it is a distinct supply from the “chlorhexidine 4.0% (Hibiclens)” (SH098) supply already located in the direct PE database.

Comment: Several commenters provided additional information regarding the use of guidewires in these procedures. Commenters stated that the three wires used in the Dialysis Circuit codes are the minimum required for these interventions and frequently additional wires would be needed in more complicated cases or in cases in which more than one access must be used. Commenters stated that the guidewires submitted are the bare minimum needed for the typical case.

Response: We appreciate the additional information from the commenters regarding the use of guidewires. We proposed to use the RUC-recommended quantities for these supplies, and we are not finalizing any changes.

Comment: One commenter stated that vascular procedures involving fluoroscopy or radiography require the use of a radio-opaque ruler (SD249) to accurately size or locate tributaries and lesions beneath the skin. The commenter indicated that some of the base procedure codes (CPT codes 36903 and 36906) include this supply, while it is missing from CPT codes 36902 and 36905 and should be included.

Response: Based upon recommendations from the RUC and specialties, we believe that the use of this supply is typical in stent procedures such as CPT codes 36903 and 36906. It was included in CPT code 37238, which is a predecessor code for these two procedures. However, the radio-opaque ruler does not appear to be typical in the other dialysis codes and we do not believe that it would be typically required in the non-stent procedures, as it was not included in any of the other predecessor codes.

Comment: One commenter requested that CMS include additional miscellaneous supplies that were missing or underrepresented in the cost inputs. These supplies were not included in the RUC recommendations for these codes. The commenter also requested increasing the quantity of each category of gloves to 3 and the quantity of gowns to 3 for each of the base codes (CPT codes 36901-36906) to more accurately reflect the typical use of these items in the dialysis circuit procedures.

Response: We believe the supplies as recommended are typical for these procedures. We also believe the proposed number of gloves and gowns would be sufficient for the typical case; we currently do not have any data to suggest that there is a need for additional gloves or gowns in these procedures. The remainder of the additional miscellaneous items appear to be new supplies with no included invoices. Many of these new items may have analogous supplies already present in our direct PE database. For the others, we will consider pricing them if invoices are submitted as part of our normal process for updating supply and equipment costs.

After consideration of comments received, we are finalizing the work RVUs for the Dialysis Circuit codes as proposed. We are also finalizing the proposed direct PE inputs, with the refinements detailed above.

(17) Open and Percutaneous Transluminal Angioplasty (CPT Codes 37246, 37247, 37248, and 37249)

In January 2015, a CPT/RUC workgroup identified the following CPT codes as being frequently reported together in various combinations: 35475 (Transluminal balloon angioplasty, percutaneous; brachiocephalic trunk or branches, each vessel), 35476 (Transluminal balloon angioplasty, percutaneous; venous), 36147 (Introduction of needle and/or catheter, arteriovenous shunt created for dialysis (graft/fistula); initial access with complete radiological evaluation of dialysis access, including fluoroscopy, image documentation and report), 36148 (Introduction of needle and/or catheter, arteriovenous shunt created for dialysis (graft/fistula); additional access for therapeutic intervention), 37236 (Transcatheter placement of an intravascular stent(s) (except lower extremity artery(s) for occlusive disease, cervical carotid, extracranial vertebral or intrathoracic carotid, intracranial, or coronary), open or percutaneous, including radiological supervision and interpretation and including all angioplasty within the same vessel, when performed; initial artery), 37238 (Transcatheter placement of an intravascular stent(s), open or percutaneous, including radiological supervision and interpretation and including angioplasty within the same vessel, when performed; initial vein), 75791 (Angiography, arteriovenous shunt (eg, dialysis patient fistula/graft), complete evaluation of dialysis access, including fluoroscopy, image documentation and report (includes injections of contrast and all necessary imaging from the arterial anastomosis and adjacent artery through entire venous outflow including the inferior or superior vena cava), radiological supervision and interpretation), 75962 (Transluminal balloon angioplasty, peripheral artery other than renal, or other visceral artery, iliac or lower extremity, radiological supervision and interpretation), and 75968 (Transluminal balloon angioplasty, each additional visceral artery, radiological supervision and interpretation).

At the October 2015 CPT Editorial Panel meeting, the panel approved the creation of four new codes and deletion of 13 existing codes used to describe bundled percutaneous transluminal angioplasty services. The Open and Percutaneous Transluminal Angioplasty family of codes overlaps with the Dialysis Circuit family of codes (CPT codes 36901-36909), as they are both being constructed from the same set of frequently reported together codes. We reviewed these two families of codes concurrently to maintain relativity between these clinically similar procedures based upon the same collection of deleted codes. After consideration of these materials, we proposed to accept the RUC-recommended work RVU for CPT codes 37246, 37247, 37248, and 37249.

For the clinical labor direct PE inputs, we proposed to use the RUC-recommend inputs with several refinements. Our proposed inputs refined the recommended clinical labor time for “Prepare and position patient/monitor patient/set up IV” from 5 minutes to 3 minutes for CPT codes 37246 and 37248. The RUC recommendation included a written justification for additional clinical labor time beyond the standard 2 minutes for this activity, stating that the extra time was needed to move leads out of X-ray field, check that X-ray is not obstructed and that there is no risk of collision of X-ray equipment with patient. As we wrote for the same clinical labor activity in the Dialysis Circuit family, we agreed that extra time might be needed for this activity as compared to the default standard of 2 minutes; however, we assigned 1 extra minute for the additional positioning tasks, resulting in a total of 3 minutes for this task. We did not believe that 3 extra minutes would be typically needed for preparation of the X-ray. The equipment times for the angiography room (EL011) and the PACS workstation (ED050) were also refined to reflect this change in clinical labor.

We proposed to remove the “drape, sterile, femoral” supply (SB009) and replace it with a “drape, sterile, fenestrated 16in x 29in” supply (SB011) for CPT codes 37246 and 37248. The two base codes out of which these new codes are being constructed, CPT codes 35471 and 35476, both made use of the SB011 fenestrated sterile drape supply, and there was no rationale provided for the switch to the SB009 femoral sterile drape in the two new codes. We solicited comment on the use of sterile drapes for these procedures, and what rationale there was to support the use of the SB009 femoral sterile drape as typical for these new procedures.

The following is a summary of the comments we received regarding our proposed valuation of the Open and Percutaneous Transluminal Angioplasty codes.

Comment: One commenter disagreed with the CMS proposed value of 3 minutes for the “Prepare and position patient/monitor patient/set up IV” clinical labor task. The commenter stated that the recommended 5 minutes of time was needed to move leads out of X-ray field, check that X-ray is not obstructed and that there is no risk of collision of X-ray equipment with patient. The commenter also indicated that the patient's arm needs to be positioned on an arm board, and requested time for this activity.

Recommended: We continue to believe that additional time may be needed for this activity as compared to the default standard of 2 minutes. However, we maintain that the commenter's request for 3 additional minutes (for a total of 5 minutes) would not typically be required for preparing the X-ray and conducting arm positioning. We do not agree that the additional tasks described by the commenters would require the requested 5 minutes of clinical labor time, and we are maintaining our proposed value of 3 minutes.

Comment: Several commenters objected to the proposed replacement of the “drape, sterile, femoral” supply (SB009) with the “drape, sterile, fenestrated 16in x 29in” supply (SB011) for CPT codes 37246 and 37248. Commenters stated that the vast majority of these new procedures will be performed from a femoral or jugular approach and will utilize a standard femoral drape. According to the commenters, the fenestrated drape provides a limited sterile field (16x29in) which does not allow room for sterile manipulation of wires and catheters as they extend away from the entry into the vascular system. With the creation of the new dialysis access circuit CPT code family, commenters indicated that the use of extremity access and fenestrated drapes would become much less typical for the new angioplasty code set.

Response: We appreciate the presentation of additional clinical information from the commenters regarding the sterile drape most appropriate for these procedures. As a result, we are finalizing inclusion of the sterile femoral drape supply (SB009) to CPT codes 37246 and 37248. We will therefore not be adding the fenestrated drape supply (SB011) to these procedures.

After consideration of comments received, we are finalizing the proposed work RVUs for the four codes in the family. We are also finalizing the proposed direct PE inputs, with the refinement to the sterile femoral drape detailed above.

(18) Esophagogastric Fundoplasty Trans-Oral Approach (CPT Code 43210)

For CY 2016, the CPT Editorial Panel established CPT code 43210 to describe trans-oral esophagogastric fundoplasty. The RUC recommended a work RVU of 9.00 and for CY 2016, we established an interim final work RVU of 7.75 for CPT code 43210. We noted that a work RVU of 7.75, which corresponds to the 25th percentile of the survey, more accurately reflected the resources used in furnishing this service.

Comment on the CY 2016 PFS final rule with comment period: Commenters urged CMS to accept the RUC-recommended work RVU of 9.00 for CPT code 43210. The commenters believed that the RUC-recommended value compared well with the key reference service, CPT code 43276 (Endoscopic retrograde cholangiopancreatography (ERCP); with removal and exchange of stent(s), biliary or pancreatic duct, including pre- and post-dilation and guide wire passage, when performed, including sphincterotomy, when performed, each stent exchanged), which has a work RVU of 8.94 and an intraservice time of 60 minutes. Commenters believed that due to similar intraservice times and intensities, that CPT code 43210 should be valued nearly identically to CPT code 43276. Some commenters also stated that to maintain relativity within the upper GI code families, CPT code 43210 should not have a lower work RVU than CPT code 43276 since the majority of survey participants indicated that CPT code 43210 is more complex than CPT code 43276. Additionally, one commenter noted that an esophagogastroduodenoscopy (EGD) is used twice during this service, before and after fundoplication. The commenter stated that because this is a multi-stage procedure, other EGD codes are not comparable. The commenter also pointed out that this technology has a small number of users and urged CMS to accept the RUC-recommended work RVU of 9.00 until there is additional utilization, and to consider reviewing this code again in subsequent years.

Response in the CY 2017 PFS proposed rule: We referred this code to the CY 2016 multi-specialty refinement panel for further review, which recommended we accept the RUC-recommended value of 9.00 work RVUs. There are four ERCP codes with 60 minutes of intraservice time, three of which have work RVUs of less than 7.00 and only one of the four codes has a work RVU higher than 7.75 RVUs (8.94). Based on our estimate of overall work for this service, we continue to believe that the 25th percentile of the survey more accurately reflects the relative resource costs associated with this service. Therefore, for CY 2017, we proposed a work RVU of 7.75 for CPT code 43210.

The following is a summary of the comments we received regarding our proposed valuation of CPT code 43210:

Comment: Commenters indicated that the survey results were limited since this is a new technology. Commenters requested that CMS finalize the RUC-recommended work RVU of 9.00, with the understanding that the service will be reviewed again in the near future with more robust survey data as the technology continues to be adopted. Commenters disagreed with CMS' comparison to other EGD codes for purposes of establishing the work RVU, due to differences in the inherent clinical procedural steps involved with this code, including that EGD is used more than once (pre- and post-fundoplication) to ensure successful completion of the procedure.

Response: While it may be true that multiple EGDs may be performed during this procedure, the surveyees are familiar with the service and we assume included this information in their proposed time and work recommendations. However, the values recommended by the survey and the RUC are not consistent with other codes with similar times and intensities. We noted in the CY 2016 interim final rule that CPT code 43240 (Drainage of cyst of the esophagus, stomach, and/or upper small bowel using an endoscope) has 10 minutes more intraservice time and a work RVU of 7.25. Therefore, we are finalizing for CY 2017 a work RVU of 7.75 for CPT code 43210.

(19) Esophageal Sphincter Augmentation (CPT Codes 43284 and 43285)

In October 2015, the CPT Editorial Panel created two new codes to describe laparoscopic implantation and removal of a magnetic bead sphincter augmentation device used for treatment of gastroesophageal reflux disease (GERD). The RUC noted that the specialty societies conducted a targeted survey of the 145 physicians who have been trained to furnish these services and who are the only physicians who have performed these procedures. They noted that only 18 non-conflicted survey responses were received despite efforts to follow up and that nine physicians had no experience in the past 12 months with the procedure. The RUC agreed with the specialty society that the expertise of those responding was sufficient to consider the survey; however, neither the RUC nor the specialty society used the survey results as the primary basis for their recommended value.

For CPT code 43284, the RUC recommended a work RVU of 10.13. We compared this code to CPT code 43180 (Esophagoscopy, rigid, transoral with diverticulectomy of hypopharynx or cervical esophagus (e.g., Zenker's diverticulum), with cricopharyngeal myotomy, includes use of telescope or operating microscope and repair, when performed), which has a work RVU of 9.03 and has identical intraservice time and similar total time. We stated in the proposed rule that we believe the overall intensity of these procedures is similar; therefore, we proposed a work RVU of 9.03 for CPT code 43284.

For CPT code 43285, the RUC recommended a work RVU of 10.47. We used the increment between the RUC-recommended work RVU for this code and CPT code 43284 (0.34 RVUs) to develop our proposed work RVU of 9.37 for CPT code 43285.

Comment: We received many comments on our proposal from various stakeholders including practitioners, manufacturers, the RUC, and medical specialty societies representing various surgical specialties. For CPT code 43284, commenters indicated that CMS' proposed crosswalk from CPT code 43180 was inadequate with regard to time and complexity of the services. Commenters stated that CPT code 43180 has 10 minutes less immediate post-service time and one less post-operative visit. Some commenters stated that it appears that the difference between the specialty society median survey total time for 43284 and the total time for CMS' proposed crosswalk from CPT code 43180 was too great to discount. Commenters also disagreed that CPT code 43284 and CMS' proposed crosswalk from CPT code 43180 had similar complexity considering that one of the procedures was performed on a natural orifice with endoscopy versus a procedure with a surgical incision. Commenters indicated that management of surgical patients with incisions necessitates a more thorough evaluation of the body than an endoscopic procedure.

For CPT code 43285, commenters noted that although CPT code 47562 (the RUC-recommended crosswalk) requires more intraservice time than the aggregate survey median time for CPT code 43285, the median intraservice time may be understated because of the number of people without experience, and suggested that the total time for CPT codes 43285 and 47562 is nearly identical and both require similar work and intensity. Commenters stated that only 18 non-conflicted survey responses were received despite the efforts of the specialty societies, and that nine physicians had no experience with the procedure in the past 12 months. Commenters also noted that the RUC recommendations used the specialty society survey times, but provided a crosswalk for work RVU valuation.

Many commenters expressed additional concerns about the specialty society survey data, indicating that the survey median and 25th percentile work RVUs were inconsistent with the total physician work for services reported with CPT codes 43284 and 43285. Commenters stated that to accept the results of the survey is to essentially state that the opinions of inexperienced surgeons is adequate to determine the value of a surgical procedure and lacked input from surgeons experienced in performing the procedure. Commenters suggested that CMS maintain carrier pricing for services reported with CPT codes 43284 and 43285 while the specialty societies conduct new surveys that include data from surgeons experienced with the procedures. Some commenters suggested that the work of CPT codes 43284 and/or 43285 is more similar to fundoplication procedures reported with CPT code 43280 (a work RVU of 18.10). Other commenters suggested valuations for these procedures ranging from 14 to 17 work RVUs, stating that the services reported with CPT codes 43284 and 43285 were slightly less complicated than fundoplication procedures, but more complex than the valuations reflected in the survey results, RUC recommendations, and CMS proposed values.

Response: We appreciate the feedback received from stakeholders regarding valuation of these services. After considering the comments received, for CY 2017, we are finalizing the RUC-recommended values for CPT codes 43284 (a work RVU of 10.13) and 43285 (a work RVU of 10.47). We recognize commenters' concerns regarding the specialty society survey data and believe these codes may be potentially misvalued. We look forward to receiving feedback from interested parties and specialty societies regarding accurate valuation of these services for consideration during future rulemaking.

(20) Percutaneous Biliary Procedures Bundling (CPT Codes 47531, 47532, 47533, 47534, 47535, 47536, 47537, 47538, 47539, 47540, 47541, 47542, 47543, and 47544)

This group of fourteen codes was reviewed by the RUC at the April 2015 meeting. We established interim final values for this group of codes during the CY 2016 PFS rulemaking cycle, and subsequently received updated RUC recommendations from the October 2015 meeting for the CY 2017 PFS rulemaking cycle. Our proposals for these codes incorporated both the updated RUC recommendations, as well as public comments received as part of the interim final status of these procedures.

We received several comments regarding the CMS refinements to the work values for this family of codes in the CY 2016 final rule with comment period. The relevance of many of these comments has been diminished by the new series of RUC recommendations for work values that we received as a result of the October 2015 meeting. Given that we proposed the updated RUC-recommended work RVUs for CPT codes 47531, 47532, 47533, 47534, 47535, 47536, 47537, 47538, 47539, 47540, 47542, 47543, and 47544, we solicited additional comments relative to these proposed values. We agreed that the second round of physician surveys conducted for the October 2015 RUC meeting more accurately captured the work and time required to perform these procedures. The one exception was CPT code 47541; the survey times for this procedure were identical as conducted for the April and October 2015 RUC meetings, yet the RUC recommendation increased from a work RVU of 5.61 in April to a work RVU of 7.00 in October. Given that the time values for the procedure remained unchanged between the two surveys, we do not understand why the work RVU would have increased by nearly 1.50 in the intervening months. Since this code also has an identical intraservice time (60 minutes) and total time (121 minutes) as CPT code 47533, we do not agree that it should be valued at a substantially higher rate compared to a medically similar procedure within the same code family. We therefore proposed to crosswalk the work value of CPT code 47541 to the work value of CPT code 47533, and we proposed a work RVU of 5.63 for both procedures.

We also note that many of the codes in the Percutaneous Biliary Procedures family were previously included in Appendix G, and were valued under the assumption that moderate sedation was typically performed on the patient. As part of the changes for services previously valued with moderate sedation as inherent, we are removing a portion of the work RVU and preservice work time from CPT codes 47532, 47533, 47534, 47535, 47536, 47538, 47539, 47540, and 47541. For example, we proposed a work value for CPT code 47541 with a 0.25 reduction from 5.63 to 5.38, and a 10 minute reduction in its preservice work time from 33 minutes to 23 minutes, to reflect the work that will now be reported separately using the new moderate sedation codes. CPT codes 47542, 47533, and 47544 also were valued with moderate sedation; however, as add-on codes, they are not subject to alterations in their work RVUs or work times since the moderate sedation code with work RVUs and work time (99152) will only be billed once for each base-code and not additionally with the add-on codes. These changes are reflected in Appendix B and the work time file posted to the web; see section II.D for more details.

For the direct PE inputs, we did not propose to include the recommended L051A clinical labor for “Sedate/apply anesthesia” and the L037D for “Assist Physician in Performing Procedure” for CPT codes 47531 and 47537. As we wrote in the CY 2016 final rule with comment period (80 FR 71053), we believe that this clinical labor describes activities associated with moderate sedation, and moderate sedation is not typical for these procedures. We also proposed to refine the L037D clinical labor for “Clean room/equipment by physician staff” from 6 minutes to 3 minutes for all of the codes in this family. Three minutes is the standard for this clinical labor activity, and we continued to maintain that the need for additional clinical labor time for this cleaning activity would not be typical for these procedures.

Comment on the CY 2016 PFS final rule with comment period: One commenter disagreed with our refinement to replace supply item “catheter, balloon, PTA” (SD152) with supply item “catheter, balloon ureteral (Dowd)” (SD150). The commenter stated that a Dowd catheter is designed and FDA approved for use in the prostatic urethra by retrograde placement through the penile urethra, and it is not designed for use in an antegrade ureteral dilation procedure. The commenter stated that this replacement is inappropriate. The updated RUC recommendations for this family of codes also restored the balloon PTA catheter.

Response in the CY 2017 PFS proposed rule: We proposed again to replace the recommended supply item “catheter, balloon, PTA” (SD152) with supply item “catheter, balloon ureteral (Dowd)” (SD150). We believed that the use of this ureteral balloon catheter, which is specifically designed for catheter and image guidance procedures, would be more typical than the use of a PTA balloon catheter. While we recognize that the Dowd catheter is not FDA approved, it is our understanding that the PTA balloon catheter has also not been FDA approved for use in these procedures. We were uncertain if the commenter was requesting that we should no longer include catheters that lack FDA approval in the direct PE database; this would preclude the use of most of the catheters in our direct PE database. We solicited additional comment on the use of FDA approved catheters; in the meantime, we continued our long-standing practice of using the catheters in the direct PE database without explicit regard to FDA approval in particular procedures.

We also proposed to remove the recommended supply item “stone basket” (SD315) from CPT code 47543 and add it to CPT code 47544. Based on the code descriptors, we believed that the stone basket was intended to be included in CPT code 47544 and was erroneously listed under CPT code 47543. We solicited comments from the public to help clarify this issue.

We noted again that many of the codes in the Percutaneous Biliary Procedures family were previously included in Appendix G, and as part of the change in moderate sedation reporting, we removed some of the recommended direct PE inputs related to moderate sedation from CPT codes 47532, 47533, 47534, 47535, 47536, 47538, 47539, 47540, and 47541. We removed the L051A clinical labor time for “Sedate/apply anesthesia”, “Assist Physician in Performing Procedure (CS)”, and “Monitor pt. following moderate sedation”. We also removed the conscious sedation pack (SA044) supply, and some or all of the equipment time for the stretcher (EF018), the mobile instrument table (EF027), the 3-channel ECG (EQ011), and the IV infusion pump (EQ032). These changes are reflected in the public use files posted to the web; see section II.D for more details.

The following is a summary of the comments we received regarding our proposed valuation of the Percutaneous Biliary Procedures codes.

Comment: Several commenters disagreed with the proposed work RVU of 5.45 for CPT code 47541. Commenters stated that although CPT codes 47541 and 47533 share similar time values, the patient population for CPT code 47541 is more complex with post-surgical anatomy and atypical problems. Therefore, the commenters stated that the direct crosswalk creates a sharp rank order anomaly within the family, and requested that CMS adopt the RUC-recommended work RVU.

Response: We agree with the commenters that the proposed work RVU for CPT code 47541 has the potential to create an anomalous relationship between the services in this family of codes. After considering the comments, we are finalizing a work RVU of 6.75 for CPT code 47541, which is the RUC-recommended work RVU of 7.00 after removing 0.25 RVUs to account for the fact that moderate sedation will now be billed separately for this service.

Comment: One commenter requested 2 minutes for the clinical labor task “Sedate/apply anesthesia” and 15 minutes for the clinical labor task “Assist Physician in Performing Procedure” for CPT codes 47531 and 47537. The commenter agreed with CMS that moderate sedation was not typical for either procedure, but stated that the 2 minutes was for the RN to administer the pre-procedure prophylactic antibiotics and the 15 minutes for assisting the physician was unrelated to moderate sedation.

Response: We disagree with the commenter that the clinical labor time for these tasks would be typical for CPT codes 47531 and 47537. For the 2 minutes of apply anesthesia time, we do not agree that this clinical labor time should be assigned when the clinical staff is performing an entirely different activity. We have not assigned clinical labor time in this way in the past, and the request for 2 minutes related to administering pre-procedure prophylactic antibiotics was never discussed in the recommendations for these procedures.

For the 15 minutes of assist physician time, the commenter did not provide a justification for why an additional staff member would be needed or what the staff member would be doing. CPT codes 47531 and 47537 already contain two clinical staff members, one technician to assist the physician and another technician to acquire images, plus a circulator. The other codes in the Percutaneous Biliary Procedures family previously had a third RN clinical staff member to administer the sedation to the patient, before moderate sedation was split off into its own separate procedure codes. However, CPT codes 47531 and 47537 do not typically require sedation, and we do not agree that this additional clinical staff member would be required to perform the procedures.

Comment: Several commenters again objected to the proposed replacement of the recommended supply item “catheter, balloon, PTA” (SD152) with supply item “catheter, balloon ureteral (Dowd)” (SD150). Commenters stated that this would not reflect the practice patterns of the Interventional Radiology community, as it is atypical and even quite rare to use ureteral balloon dilatation catheters in the biliary tree. The commenters provided information regarding the size of uretal balloon catheters, indicating that the maximum diameter is 8mm (Bard) or 7mm (Cook Medical). According to commenters, these sizes are frequently inadequate to treat the wide variety of pathologies in the biliary tree where often balloon sizes up to 12 mm are required. As a result, the commenters stated that the change of the balloon catheter supply item does not accurately represents the actual supplies utilized in real practice, nor does the Dowd ureteral balloon catheter satisfy the clinical need performed during the procedure.

Response: We appreciate the additional clinical information supplied by the commenters regarding the current use of balloon catheters. However, although commenters stated that Bard catheters and Cook Medical catheters are frequently too small to treat some of the wide variety of pathologies that occur in the biliary tree, commenters did not indicate what size balloon catheter would be typically used for these particular procedures in the Percutaneous Biliary Procedures, or provide a specific rationale for why the catheter we proposed (the Dowd ureteral balloon catheter) would not be appropriate for these procedures. We note again that we are required to assess resources based on the typical case, and the commenters did not provide data to indicate that the proposed Dowd catheter would be inadequate in the typical case for these procedures in question, only that it may be insufficient for certain pathologies in the biliary tree. We continue to believe that the Dowd ureteral balloon catheter, which is specifically designed for catheter and image guidance procedures, would be more typical than the use of a PTA balloon catheter.

Comment: One commenter indicated that the stone basket supply (SD315) had indeed been incorrectly assigned to CPT code 47543, and thanked CMS for moving it to CPT code 47544 where it was intended.

Response: We appreciate the response from the commenter.

After consideration of comments received, we are finalizing our proposed work RVUs for the Percutaneous Biliary Procedures family of codes, with the one change to a work RVU of 6.75 for CPT code 47541. We are finalizing our proposed direct PE inputs without refinement.

(21) Percutaneous Image Guided Sclerotherapy (CPT Code 49185)

For CY 2016, we established an interim final work RVU of 2.35 for CPT code 49185 based on a crosswalk from CPT code 62305 (Myelography via lumbar injection, including radiological supervision and interpretation; 2 or more regions (e.g., lumbar/thoracic, cervical/thoracic, lumbar/cervical, lumbar/thoracic/cervical)); which we believed accurately reflected the time and intensity involved in furnishing services reported with CPT code 49185. We also requested stakeholder input on the price of sclerosing solution (supply item SH062) as the volume of the solution in this procedure (300 mL) is much higher than other CPT codes utilizing sclerosing solution (between 1 and 10 mL).

Comment on the CY 2016 PFS final rule with comment period: In response to the CY 2016 PFS final rule with comment period (80 FR 71054), commenters disagreed with CMS' crosswalk from CPT code 62305. Commenters suggested that the RUC's recommended crosswalk from CPT code 31622 (Bronchoscopy, rigid or flexible, including fluoroscopic guidance, when performed; diagnostic, with cell washing, when performed (separate procedure)) was a more appropriate comparison due to the similarity of the services. Commenters requested that CPT code 49185 be referred to the refinement panel. The requests did not meet the requirements related to new clinical information for referral to the refinement panel. We continue to believe that for CPT code 49185 a crosswalk from the value of CPT code 62305 is accurate due to similarities in overall work.

Commenters also stated that the procedure reported with CPT code 49185 required a separate clinical labor staff type. The commenter noted that, due to the inclusion of this additional individual, the L037D clinical labor and additional gloves were appropriate to include in the procedure. The commenter did not provide any evidence for this claim.

Response in the CY 2017 PFS proposed rule: We continue to believe that this additional use of clinical staff would not be typical for CPT code 49185. This procedure does not involve moderate sedation, and therefore, we do not believe that there would be a typical need for a third staff member. Additionally, we did not receive any information regarding the sclerosing solution (supply item SH062) that supports maintaining an input of 300 mL, which far exceeds the volume associated with other CPT codes.

Therefore, for CY 2017, we proposed a work RVU of 2.35 for CPT code 49185. We sought stakeholder feedback regarding why a different work RVU or crosswalk would more accurately reflect the resources involved in furnishing this service. We also proposed to maintain our direct PE refinements from the CY 2016 PFS final rule with comment period, but proposed to refine the direct practice expense inputs for the sclerosing solution (supply item SH062) from 300 mL to 10 mL, which is the highest level associated with other CPT codes utilizing sclerosing solution.

The following is a summary of the comments we received regarding our proposed valuation of CPT code 49185.

Comment: Commenters requested that CMS use the RUC-recommended crosswalk from CPT code 31622 instead of the CMS-proposed crosswalk from CPT code 62305. Commenters stated that CMS' crosswalk undervalues the services, the RUC-recommended crosswalk has analougous clinical activities during the procedure, as well as a similar risk, and the intensity of work involved for services reported with CMS' comparison code is less than during sclerotherapy. Commenters suggested that the sclerotherapy procedure includes inherent risks and challenges that are not adequately accounted for in CMS' proposed crosswalk.

Response: We disagree with commenters that the RUC's recommended crosswalk from CPT code 31622 has analogous clinical activities compared to CMS' proposed crosswalk from CPT code 62305. CMS' crosswalk code refers to a procedure with injection, drainage, and aspiration, which has more clinical similarity to CPT code 49185 than the RUC's recommended crosswalk from 31622, which is used to report a broncoscopy procedure. We continue to believe that a work RVU of 2.35 is an appropriate valuation for services reported using CPT code 49185 and we maintain that CPT code 62305 is an accurate crosswalk, since CPT codes 49185 and 62305 have similar service times. Therefore, for CY 2017, we are finalizing a work RVU of 2.35 for CPT code 49185.

Comment: Commenters disagreed with CMS' proposal to include a direct PE input of 10 mL of sclerosing solution (supply item SH062) and requested that CMS accept the RUC's recommendation to include 300 ml of sclerosing solution as part of the direct PE inputs for this procedure. One commenter indicated that other services that utilize sclerosing solution are used to describe injection of sclerosant into vascular structures which tend to be relatively small in size, and therefore, use a much smaller volume. Another commenter stated that for this procedure, the sclerosing solution is injected and drained three separate times, equating to 100 mL per injection, and that use of lesser volumes of sclerosant or less than three administrations of the sclerosant during the procedure would allow for more frequent recurrence necessitating additional procedures.

Response: We appreciate the commenters' feedback regarding the direct PE inputs for CPT code 49185. We inadvertently included the RUC-recommended quantity of 300 mL for the sclerosing solution (supply item SH062) in developing the proposed rates for this code. For CY 2017, we are finalizing the RUC-recommended direct PE inputs, including 300 mL of sclerosing solution. We welcome stakeholder feedback regarding the appropriate PE inputs for this procedure for consideration for CY 2018, including volume and pricing of the sclerosing agent.

(22) Genitourinary Procedures (CPT Codes 50606, 50705, and 50706)

In the CY 2016 PFS final rule with comment period, we established as interim final the RUC-recommended work RVUs for all three codes. We did not receive any comments on the work values for these codes, and we proposed to maintain all three at their current work RVUs.

The RUC recommended the inclusion of “room, angiography” (EL011) for this family of codes. As we discussed in the CY 2016 PFS final rule with comment period, we did not believe that an angiography room would be used in the typical case for these procedures, and we therefore replaced the recommended equipment item “room, angiography” with equipment item “room, radiographic-fluoroscopic” (EL014) for all three codes on an interim final basis. We also stated our belief that since the predecessor procedure codes generally did not include an angiography room and we did not have a reason to believe that the procedure would have shifted to an angiography room in the course of this coding change, we did not believe that the use of an angiography room would be typical for these procedures.

Comment on the CY 2016 PFS final rule with comment period: Several commenters disagreed with the CMS substitution of the fluoroscopic room in place of the angiography room. The commenters stated that all three of these procedures were previously reported using CPT code 53899 (Unlisted procedure, urinary system) which does not have any PE inputs, and the RUC recommendations included as a reference CPT code 50387 (Removal and replacement of externally accessible transnephric ureteral stent), which includes an angiography room. The commenters suggested that CPT code 50387 was an example of a predecessor code that included the use of an angiography room, along with other codes that are being bundled together to create the new Genitourinary codes.

Response in the CY 2017 PFS proposed rule: We did not agree with the commenters' implication that because CPT code 50387 was an appropriate reference code for use in valuation, that it necessarily would have previously been used to describe services that are now reported under CPT codes 50606, 50705, or 50706. Our perspective was consistent with the RUC-recommended utilization crosswalk for the three new codes, which did not suggest that the services were previously reported using 50706. We did not believe that use of one particular code for reference in developing values for another necessarily meant that the all of the same equipment would be used for both services.

We did not believe that these codes described the same clinical work either. CPT code 50387 is for the “Removal and replacement of externally accessible transnephric ureteral stent” while CPT code 50606 describes an “Endoluminal biopsy of ureter and/or renal pelvis”, CPT code 50705 refers to “Ureteral embolization or occlusion”, and CPT code 50706 details “Balloon dilation, ureteral stricture.” Additionally, the codes do not have the same global periods, which makes comparisons between CPT code 50387 and CPT codes 506060, 50705, and 50706 even more difficult. We noted that while the commenter stated that CPT code 50387 was provided as a reference for these procedures, 50387 is not listed as a reference for any of these three codes, or mentioned at all in the codes' respective summary of recommendations. However, we acknowledged that among the procedures that are provided as references, many of them included the use of an angiography room, such as CPT code 36227 (Selective catheter placement, external carotid artery) and CPT code 37233 (Revascularization, endovascular, open or percutaneous, tibial/peroneal artery, unilateral, each additional vessel). Therefore, we agreed that the use of the angiography room in these procedures, or at least some of its component parts, might be warranted.

Comment on the CY 2016 PFS final rule with comment period: A commenter stated that the substitution of the fluoroscopic room for the angiography room was clinically unjustified. The commenter stated that the angiography room was needed for these procedures to carry out 3-axis rotational imaging (so as to avoid rolling the patient), ensure sterility, and avoid unacceptable radiation exposure to physicians, their staff, and their patients. The commenter indicated that the only piece of equipment listed in the angiography room that would not be typically utilized for these procedures is the Provis Injector. All of the other items were used for these Genitourinary procedures. The commenter urged CMS to restore the angiography room to these procedures.

Response in the CY 2017 PFS proposed rule: We agreed that it is important to provide equipment that is medically reasonable and necessary. Our concern with the use of the angiography room for these codes was that we did not believe all of the equipment would be typically necessary to furnish the procedure. For example, the commenter agreed that the Provis Injector would not be required for these Genitourinary codes. Therefore, we proposed to remove the angiography room from these three procedures and add in its place the component parts that make up the room. Table 17 detailed these components:

We included all of the above components except the Provis Injector, as commenters indicated that its use would not be typical for these procedures. We welcomed additional comments regarding if these or other components were typically used in these Genitourinary procedures. We lacked pricing information for these components; we therefore proposed to include each of these components in the direct PE input database at a price of $0.00 and we solicited invoices from the public for their costs to be able to price these items for use in developing final PE RVUs for CY 2017.

We also noted that we believed that this issue illustrated a potentially broad problem with our use of equipment “rooms” in the direct PE input database. For most services, we only include equipment items that are used and unavailable for other uses due to their use during the services described by a particular code. However, for items included in equipment “rooms,” we allocate costs regardless of whether the individual items that comprise the room are actually used in the particular service.

To maintain relativity among different kinds of procedures, we were interested in obtaining more information specifying the exact resources used in furnishing services described by different codes. We hoped to address this subject in greater detail in future rulemaking.

The following is a summary of the comments we received regarding our proposed valuation of the Genitourinary codes:

Comment: Many commenters objected to the removal of the angiography room from these codes and its replacement with the component parts of the room. Commenters stated that it was misguided to unbundle the components of the angiography room when one equipment item within the room is not utilized. They indicated that there are numerous cases where an equipment room is used despite the fact that not every item in the room is needed for a service, because in practice the rooms are configured for the most typical type of procedure performed within the room and it would not be efficient or realistic to remove items from a room when a less typical service is needed. For the specific case of the Provis Injector equipment, commenters stated it could not be used elsewhere and there was no way to create a separate angiography room for nonvascular procedures that did not require the injector.

Commenters did not generally agree with the CMS proposal to price all of the components of the angiography room at $0.00 pending invoices from the public regarding their individual cost. Commenters stated that the resource cost of the angiography room components was clearly not $0.00, since the equipment in total costs over $1.3 million. Commenters stated that it was not realistic to submit 21 separate invoices during the 60 day comment period, and furthermore that the components of the angiography room are typically not sold separately.

Response: We appreciate the feedback from commenters regarding the difficulties involved in pricing the components for the angiography room. We have longstanding issues with the equipment rooms as they are currently constituted, due to our belief that all of the components of the room may not typically be used in performing the procedure in question. We continue to believe that these three codes do not make use of all of the components of the angiography room, and we believe that this code family serves as a clear example of the problems in relativity associated with the use of “rooms” as equipment items for a limited set of services under the PFS. However, we agree with the commenters that it is not likely that the components of the angiography room do not have a price. Therefore, while we continue to seek invoices for more detailed pricing information, we are restoring the angiography room (EL011) equipment to these three codes, with an equipment time of 47 minutes for CPT code 50606, 62 minutes for CPT code 50705, and 62 minutes for CPT code 50706, in each case consistent with the equipment time in CY 2016. We intend to continue to consider the use of equipment “rooms” more broadly for future rulemaking.

After consideration of comments received, we are finalizing our work values for the three Genitourinary codes as proposed. We are finalizing the proposed direct PE inputs as well, with the changes to the angiography room as detailed above.

(23) Electromyography Studies (CPT Code 51784)

We identified CPT code 51784 as potentially misvalued through a screen of high expenditure services by specialty. This family also includes CPT code 51785 (Needle electromyography studies (EMG) of anal or urethral sphincter, any technique) but was not included in this survey. Both services have 0-day global periods. The RUC recommended a work RVU of 0.75 for CPT code 51784. We believe that this service is more accurately valued without a global period, since that is more consistent with other diagnostic services, and specifically, with all the other diagnostic electromyography services. We proposed to eliminate the global period and proposed the RUC-recommended work RVU of 0.75 for CY 2017. We also proposed to change the global period for CPT code 51785 from 0-day to no global period, to be consistent with the global period for CPT code 51784. Additionally, we proposed to add CPT code 51785 to the list of potentially misvalued codes to update the value of the service considering the change in global period, and to maintain consistency with CPT code 51784.

Comment: A commenter supported CMS' proposal to accept the RUC-recommended work value. The commenter requested that CMS indicate any global period changes and requests for codes as part of the family when CMS initially nominates a code or reviews the RUC level of interest (LOI) prior to distribution.

Another commenter, while supporting our acceptance of the RUC-recommended work RVU for CPT code 51784, did not support adding CPT code 51785 to the potentially misvalued code list as that code was addressed recently when the new CPT codes were created for urodynamic testing procedures.

Response: We appreciate commenters' perspectives. We note that CPT code 51785 has not been valued since January 2003, at the same RUC meeting wherein CPT code 51784 was valued. We encourage stakeholders to submit the entire code family when submitting codes for inclusion on the list of potentially misvalued codes.

Comment: One commenter stated that there is no difference in the work value of CPT code 51784 whether it has a 0-day global period versus an XXX global period, and should not be considered as potentially misvalued.

Response: We note that CPT code 51784 was identified as potentially misvalued through a screen of high expenditure services by specialty. In the standard process of code valuation, CMS decided to change the global period to XXX, indicating no global period, so that the code is more closely aligned with other similar services.

Comment: One commenter did not agree that CMS should accept the RUC-recommended work values, stating that the RUC-recommended work RVU underestimates the work involved in furnishing this service.

Response: We remind commenters and stakeholders that disagree with CMS values, including those based on RUC recommendations, that in addition to submitting comments on our proposed rules, they may also nominate codes as potentially misvalued through the public nomination process. After consideration of comments, we are finalizing the work and global period changes as proposed.

(24) Cystourethroscopy (CPT Code 52000)

In the CY 2016 PFS final rule with comment period, CMS identified CPT code 52000 through the screen for high expenditure services. We stated in the CY 2017 proposed rule that the RUC-recommended work RVU of 1.75 for CPT code 52000 is higher than the work RVUs for all 0-day global codes with 10 minutes of intraservice time and we did not believe that the overall intensity of this service was greater than all of the other codes. Instead, we proposed that this code compares favorably to CPT code 58100 (Endometrial sampling (biopsy) with or without endocervical sampling (biopsy), without cervical dilation, any method (separate procedure)), which has a work RVU of 1.53, and has identical intraservice time and similar total time. Therefore, we proposed a work RVU of 1.53 for CPT code 52000, using a direct crosswalk to CPT code 58100.

Comment: Commenters requested that CMS finalize the RUC-recommended work RVU of 1.75 for this procedure. Commenters stated that the RUC-recommended crosswalk codes were a more accurate comparison of physician work, time, and intensity for procedures reported with CPT code 52000.

Response: The RUC-recommended work RVU of 1.75 is higher than the work RVUs associated with all other codes with 0-day global periods and 10 minutes of intraservice time, and we continue to believe that the work and intensity of this service is similar to other CPT codes with 10 minutes of intraservice time. Therefore, we are finalizing a work RVU of 1.53 for CPT code 52000.

(25) Biopsy of Prostate (CPT Code 55700)

In the CY 2016 PFS final rule with comment period, CMS identified CPT code 55700 as potentially misvalued based on the high expenditure by specialty screen.

The RUC subsequently reviewed this code for physician work and practice expense and recommended a work RVU of 2.50 based on the 25th percentile of the survey. We believed the RUC-recommended work RVU overestimates the work involved in furnishing this service given the reduction in total service time; specifically, the reduction in preservice and postservice times. The RUC recommendation also appears overvalued when compared to similar 0-day global services with 15 minutes of intraservice time and comparable total times. To develop a proposed work RVU, we crosswalked the work RVUs for this code from CPT code 69801 (Labyrinthotomy, with perfusion of vestibuloactive drug(s), transcanal), noting similar levels of intensity, similar total times, and identical intraservice times. Therefore, we proposed a work RVU of 2.06 for CPT code 55700.

Comments: A few commenters, including the RUC, noted the RUC compared CPT code 55700 to other 0-day global services with 15 minutes of intraservice time and stated that the RUC-recommended value was appropriate. The RUC noted that the overall work of the surveyed code was similar to services: CPT code 93503 (Insertion and placement of flow directed catheter (eg, Swan-Ganz) for monitoring purposes) (work RVU = 2.91, intraservice time of 15 minutes) and CPT code 36556 (Insertion of non-tunneled centrally inserted central venous catheter; age 5 years or older) (work RVU = 2.50, intraservice time of 15 minutes). The RUC determined that these services required the same intra-service time, comparable physician work and intensity and recommended CMS accept the RUC-recommended work RVU of 2.50. Additionally, the RUC continued to urge specialty societies to submit invoices for new equipment.

Response: We appreciate additional information offered by the commenters. After consideration of comments received, we agree with the additional information provided by commenters and are finalizing the RUC-recommended work RVU of 2.50.

(26) Laparoscopic Radical Prostatectomy (CPT Code 55866)

In the CY 2016 PFS final rule with comment period, we established an interim final work RVU of 21.36 for CPT code 55866 based on a direct crosswalk to CPT code 55840 (Prostatectomy, retropubic radical, with or without nerve sparing). We stated that we believed these codes were medically similar procedures with nearly identical time values, and we did not believe that the difference in intensity between CPT code 55840 and CPT code 55866 was significant enough to warrant the RUC-recommended difference of 5.50 work RVUs. We also compared CPT code 55866 to the work RVU of 25.18 for CPT code 55845, and stated our belief that, in general, a laparoscopic procedure would not require greater resources than an open procedure.

Comment on the CY 2016 PFS final rule with comment period: Several commenters disagreed with the statement that a laparoscopic procedure, such as CPT code 55866, would generally require fewer resources than an open procedure, such as CPT code 55840. Commenters stated that developing the skill necessary to perform a minimally invasive laparoscopic surgery requires a greater degree of experience and specialized training than that required to perform an open prostatectomy. Commenters indicated that this level of practitioner skill should be reflected in the work RVU for the procedure, as intensity is based in part upon skill, mental effort, and psychological stress.

Response in the CY 2017 PFS proposed rule: We agreed with the commenters that skill and technique, as well as mental effort and psychological stress on the part of the practitioner contribute to the overall intensity of the furnishing a given service, and therefore, are one of the two components in determining code-level work RVUs. However, we did not believe that relative increases in requisite skill or technique can be considered alone. Although the development of new technology (such as robotic assistance) may create a greater burden of knowledge on the part of the practitioner, it can also make procedures faster, safer, and easier to perform. This means that there may be reductions in time for such a procedure (which is the other component of the work RVU), but also that the mental effort and psychological stress for a given procedure may be mitigated by the improvements in safety. Therefore, we did not agree that a newer procedure that includes additional technology and requires greater training would inherently be valued at a higher rate than an older and potentially more invasive procedure.

Comment on the CY 2016 PFS final rule with comment period: A commenter stated that CPT code 55866 describes two very different procedures in one code. The descriptor for the code states “includes robotic assistance when performed”, and the procedure is performed differently depending on whether or not the robotic assistance is included. The commenter indicated that the vast majority of radical prostatectomies are performed with the robot, and although the outcomes are the same in both cases, the procedures are completely different.

Response in the CY 2017 PFS proposed rule: We agreed with the commenter that the descriptor includes the possibility for confusion, especially on the part of the survey respondents. Valuing this code based on the typical case is difficult when the procedure differs depending on the inclusion or exclusion of robotic assistance. We suggested that valuation might be improved if the CPT Editorial Panel were to consider further revisions to this code to describe the two cases of laparoscopic radical prostatectomy: With and without robotic assistance.

Comment on the CY 2016 PFS final rule with comment period: One commenter stated that the application of the phase-in transition for facility-only codes like CPT code 55866 would have a particularly egregious impact in the second year of the transition. The commenter urged CMS to ensure that its implementation of the phase-in transition does not undermine the protections created by the statute.

Response in the CY 2017 PFS proposed rule: Please see sections II.G and II.H for a discussion of the phase-in transition and its implementation in its second year.

Comment on the CY 2016 PFS final rule with comment period: Several commenters requested that CMS refer CPT code 55866 to the refinement panel for review. At the refinement panel, the presenters brought up new evidence in the form of a study published in 2016 describing discharge data for radical laparoscopic prostatectomies. The presenters stated that there were many more people included in this study as opposed to the 30 respondents in the survey data, and that on average the robotic procedure took 90 minutes longer than the open procedure. The additional time needed to perform the procedure, as indicated by this new study's results, was presented as a new rationale as to why CMS should accept the RUC-recommended work RVU.

Response in the CY 2017 PFS proposed rule: CPT code 55866 was referred to the CY 2016 Multi-Specialty Refinement Panel per the request of commenters. The outcome of the refinement panel was a median work RVU of 26.80, the same value as the RUC recommended in the previous rulemaking cycle. After consideration of the comments and the results of the refinement panel, we proposed for CY 2017 to maintain the interim final work RVU of 21.36 for CPT code 55866. We were interested in the results of the study mentioned at the refinement panel, and we stated that we would consider incorporating this data into the valuation of this code, including, if appropriate, adjustments to the work times used in PFS ratesetting. We also solicited that the study be submitted through the public comment process so that we could allow it proper consideration along with other information submitted by the public, rather than using the results of a single study to propose valuations. We were also curious about the time values regarding the duration of CPT code 55866. One of the members of the refinement panel stated that on average the robotic procedure took 90 minutes longer than the open procedure. This was not what was indicated by the survey data from the RUC recommendations, which had the two procedures valued at virtually identical times (same intraservice time, 6 minutes difference total time). We therefore solicited comment on whether the times included in this study were more accurate than the time reflected in the RUC surveys.

The following is a summary of the comments we received regarding our proposed valuation of CPT code 55866:

Comment: One commenter agreed that the code descriptor for CPT code 55866 might have caused confusion by the RUC survey respondents. The commenter stated that they were encouraged by the CMS comments that the valuation might be improved if the CPT Editorial Panel were to consider further revisions to this code to describe a laparoscopic radical prostatectomy with and without robotic assistance. The commenter requested a strong statement from CMS urging the CPT Editorial Panel to create two unique codes: One for laparoscopic radical prostatectomy and one for robotic radical prostatectomy.

Response: We believe that there are potential problems with CPT code 55866 as it is currently described and with the corresponding RUC recommendation. Commenters presented data suggesting that there are significant differences between the robotic and non-robotic versions of the procedure in the length of time required to perform the operation. However, the same data also suggests that the non-robotic version of the laparoscopic radical prostatectomy has become comparatively rare. Given the information presented by commenters, we believe that valuation might be improved with further revisions to this code. However, we note that we do not direct the work of the CPT Editorial Panel, and we also note the comparative rarity of the non-robotic version of the procedure.

Comment: Several commenters referenced a study entitled “Robot-assisted versus Open Radical Prostatectomy: A Contemporary Analysis of an All-payer Discharge Database” by J.L. Leow, S.L. Chang, and colleagues. This study was published in February 2016, and it detailed how university investigators analyzed more than 600,000 men undergoing radical prostatectomy in the United States from 2003-2013, which showed that the robotic approach took on average 90 minutes longer than an open radical prostatectomy. Commenters noted how this contrasted to the RUC survey data that had only 32 respondents and recommended an intraservice time equal to an open radical prostatectomy (180 minutes). The commenters presented the study data in favor of demonstrating how the robotic approach to radical prostatectomy detailed in CPT code 55866 takes significantly more time to perform than the open approach detailed in the CMS crosswalk code 55840. Commenters recommended that CMS adopt the RUC-recommended work RVU of 26.80 based on this new clinical evidence contained in the study.

Response: We appreciate the submission of this additional clinical information from the commenters. We have had longstanding interest in using robust data sources regarding the resource costs of PFS services, and we believe that the use of such additional outside data sources can improve the accuracy of the valuation of services. However, we do note that the cited study was not specifically designed to measure intraoperative times and did not use the same “skin to skin” definition of intraservice time typically used in the development of times included in PFS ratesetting.

In this case of the particular comment, we note the potential logical dissonance of the commenter urging us to adopt the RUC-recommended work value derived from the RUC survey by citing alternative data that calls into question the accuracy of the time data from the same RUC survey. In other words, we are troubled with the idea that we should consider survey data as valid for work while rejecting its validity for time, given that time is one of the two elements of overall work.

Despite these concerns, we agree that the study presents additional data indicating that there is a significant difference between the open and robotic-assisted forms of laparoscopic radical prostatectomy, and that the robotic form described by CPT code 55866 likely takes a longer time to perform. Based on this presentation of additional clinical evidence, we agree with the commenters that the recommended work RVU of 26.80 is a more appropriate value for this procedure.

After consideration of comments received, we are finalizing a work RVU of 26.80 for CPT code 55866.

(27) Hysteroscopy (CPT Codes 58555, 58558, 58559, 58560, 58561, 58562, and 58563)

During CY 2016 PFS rulemaking, we identified CPT code 58558 as a potentially misvalued code via the high expenditure specialty screen. CPT codes 58559-58563 were also included in the RUC's January 2016 review of this family of codes.

For CPT code 58555, the RUC recommended a work RVU of 3.07. We proposed that the 25th percentile survey result, a work RVU of 2.65, accurately reflects the resources involved in furnishing this service. We stated that this value is bracketed by two crosswalk codes, CPT code 43191 (Esophagoscopy, rigid, transoral; diagnostic, including collection of specimen(s) by brushing or washing when performed (separate procedure)), which has a work RVU of 2.49, and CPT code 31295 (Nasal/sinus endoscopy, surgical; with dilation of maxillary sinus ostium (for example, balloon dilation), transnasal or via canine fossa), which has a work RVU of 2.70. CPT codes 43191 and 31295 have identical intraservice times and similar total times when compared with CPT code 58555.

For CPT code 58558, the RUC recommended a work RVU of 4.37. However, we believed that a direct crosswalk from CPT code 36221 (Non-selective catheter placement, thoracic aorta, with angiography of the extracranial carotid, vertebral, and/or intracranial vessels, unilateral or bilateral, and all associated radiological supervision and interpretation, includes angiography of the cervicocerebral arch, when performed), which has a work RVU of 4.17, and identical intraservice time, and similar total time, more accurately reflects the time and intensity of furnishing this service. Our proposed work RVU was additionally supported by using an increment between this code and the base code for this family, CPT code 58555. The increment between the RUC-recommended values for these two codes is 1.3. That increment added to the proposed work RVU of 2.65 for the base code, CPT code 58555, results in a work RVU of 3.95. Therefore, we proposed a work RVU of 4.17 RVUs for CPT code 58558.

For CPT code 58559, the RUC recommended a work RVU of 5.54. However, we believed that a direct crosswalk from CPT code 52315 (Cystourethroscopy, with removal of foreign body, calculus, or ureteral stent from urethra or bladder (separate procedure); complicated), which has a work RVU of 5.20, a similar intraservice time, and similar total time as compared with CPT code 58559 more accurately reflects the time and intensity of furnishing this service. This proposed value was additionally supported by using an increment between CPT code 58559 and the base code for this family, CPT code 58555. The increment between the RUC recommended values for the two codes is 2.47. That increment added to the proposed value for the base code, CPT code 58555, would result in a work RVU of 5.12. Therefore, we proposed a work RVU of 5.20 for CPT code 58559.

For CPT code 58560, the RUC recommended a work RVU of 6.15. We stated in the proposed rule that we believe that a direct crosswalk from CPT code 52351 (Cystourethroscopy, with ureteroscopy and/or pyeloscopy; diagnostic), which has a work RVU of 5.75 and which has more intraservice time and very similar total time, more accurately reflects the time and intensity of furnishing this service. Our proposal further supported this value by using an increment between CPT code 58560 and the base code for this family, CPT code 58555. We stated that the increment between the RUC recommended values for the two codes is 3.08. That increment added to the proposed value for the base code, CPT code 58555, would result in a work RVU of 5.73. Therefore, we proposed a work RVU of 5.75 for CPT code 58560.

For CPT code 58561, the RUC recommended a work RVU of 7.00. We stated in the proposed rule that we believe that a direct crosswalk from CPT code 35475 (Transluminal balloon angioplasty, percutaneous; brachiocephalic trunk or branches, each vessel), which has a work RVU of 6.60 and which has similar intraservice and total times, more accurately reflected the time and intensity of furnishing this service. We also noted that our proposal was further supported by using an increment between CPT code 58561 and the base code for this family, CPT code 58555. The increment between the RUC recommended values for the two codes is 3.93. That increment added to the proposed value for the base code, CPT code 58555, would result in a work RVU of 6.58. Therefore, we proposed a work RVU of 6.60 for CPT code 58561.

For CPT code 58562, the RUC recommended a work RVU of 4.17. However, we believed that a direct crosswalk of the work RVUs for CPT code 15277 (Application of skin substitute graft to face, scalp, eyelids, mouth, neck, ears, orbits, genitalia, hands, feet, and/or multiple digits, total wound surface area greater than or equal to 100 sq cm; first 100 sq cm wound surface area, or 1% of body area of infants and children), which has a work RVU of 4.00 and which has identical intraservice time and similar total time, more accurately reflects the time and intensity of furnishing this service. The RUC also used this code as one of its supporting codes for its recommendation. This value is additionally supported by using an increment between CPT code 58562 and the base code for this family, CPT code 58555. The increment between the RUC recommended values for the two codes is 1.10. That increment added to the proposed value for the base code, CPT code 58555, results in a work RVU of 3.75. Therefore, we proposed a work RVU of 4.00 for CPT code 58562.

For CPT code 58563, the RUC recommended a work RVU of 4.62. However, we believed that a direct crosswalk of the work RVUs for CPT code 33962 (Extracorporeal membrane oxygenation (ECMO)/extracorporeal life support (ECLS) provided by physician; reposition peripheral (arterial and/or venous) cannula(e), open, 6 years and older (includes fluoroscopic guidance, when performed)), which has a work RVU of 4.47 and that has identical intraservice time and similar total time, more accurately reflects the resources involved in furnishing this service. This value is additionally supported by using an increment between CPT code 58563 and the base code for this family, CPT code 58555. The increment between the RUC recommended values for the two codes is 1.55. That increment added to the proposed value for the base code, CPT code 58555), results in a work RVU of 4.20. We note that CPT code 58563 has the same instraservice time and the same total time as CPT code 58558; however, we agreed that the intensity would be slightly higher for this service. Therefore, we proposed a work RVU of 4.47 for CPT code 58562.

The RUC submitted invoices for two new equipment items used in furnishing CPT code 58558, the hysteroscopic fluid management system and the hysteroscopic resection system. We proposed to use these invoice prices for the hysteroscopic fluid management system, which totaled $14,698.38. The hysteroscopic resection system included the price of the hysteroscope, as well as other items necessary for tissue removal. However, we generally price endoscopes separately and not as a part of a system. To maintain consistency, we proposed not to include the hysteroscope from the Resection System. Instead, we proposed to update the equipment item “endoscope, rigid, hysteroscopy” (ES009) with the invoice price, $6,207.50. We did not propose to include the sterilization tray from the hysteroscopic resection system because we believe this tray has generally been characterized as an indirect practice expense. For the hysteroscopic resection system, we proposed to include the hysteroscopic tissue remover ($18,375), the sheath ($1,097.25), and the calibration device ($300), and created a new equipment item code, priced at $19,857.50 in the proposed direct PE input database. We did not propose to include the calibration device since the submitted price was not documented with a paid invoice.

Comment: Commenters, including the RUC, disagreed with CMS' proposed refinements to the work RVUs for these procedures, and requested that CMS finalize the RUC-recommended work values for these codes. Commenters suggested that these procedures are more complex in cases where it is more difficult to find and feed the scopes through the cervix. Commenters suggested that it appeared as though CMS used a time to work ratio to value these services, stating further that, for example, CPT code 58555 requires a forced dilation of a natural orifice, very small in size and can be difficult to identify in a post-menopausal patient or a patient with prior cervical surgery. Commenters suggested that the CMS crosswalk codes are for a natural orifice that might not require any dilation or only a 10% dilation, and the orifice is consistently the same with little variation among patients.

Response: While we appreciate the commenters' feedback, we do not consider forced or difficult dilation as described by the commenter to be typical based on the RUC's clinical vignette and that the difficulty of forced dilation at the time of surgery can often be offset by preoperative cervical ripening. Therefore, we are finalizing the following work RVUs for each code in this family.

• CPT code 58555, 2.65 work RVUs;
• CPT code 58558, 4.17 work RVUs;
• CPT code 58559, 5.20 work RVUs;
• CPT code 58560, 5.75 work RVUs;
• CPT code 58561, 6.60 work RVUs;
• CPT code 58562, 4.00 work RVUs; and
• CPT code 58563, 4.47 work RVUs.

Comment: Regarding the direct PE inputs for CPT code 58558, one commenter requested that CMS add a procedure kit and update the prices for these supplies to reflect the cost of providing this procedure in the physician office setting. The commenter also submitted invoices related to other direct PE inputs for this code, including invoices for the incisor blade and the procedure kit, which the commenter indicated includes inflow tubing, outflow tubing, and the non-sterile components of jumper cables and a tissue trap.

Response: We appreciate the feedback we received regarding the direct PE inputs for CPT code 58558. We agree with the addition of the hysteroscopic procedure kit and are creating a new supply item “hysteroscopic fluid management tubing set” using a single invoice price of $320. Additionally, we note that we inadvertently did not remove the existing direct PE inputs related to suction, which we proposed to replace with the hysteroscopic fluid management system. Therefore, we are removing direct PE inputs for the following items:

• Supply item SD009: Canister, suction;
• Supply item SD031: Catheter, suction; and
• Equipment item EQ235: Suction machine (Gomco).

The commenter also included an additional invoice for the incision instrument. Based on this new information, we are renaming this new supply item, “hysteroscopic tissue removal device,” with a final price of $629.00, which is the simple average of the two invoice prices we have received for this supply item ($599 and $659 respectively). Additionally, we note that our proposed summary price for the hysteroscopic resection system was added incorrectly. The correct price is $19,772.25. We are also modifying the equipment title to ensure clarity of items included in the hysteroscopic resection system (control unit, footpiece, handpiece, sheath and calibration device).

(28) Intracranial Endovascular Intervention (CPT Codes 61645, 61650, and 61651)

For CY 2016, we established an interim final work RVU of 15.00 for CPT code 61645, 10.00 for CPT code 61650 and 4.25 for CPT code 61651. The RUC-recommended values for CPT codes 61645, 61650 and 61651 were 17.00, 12.00 and 5.50, respectively. We valued CPT code 61645 by applying the ratio between the RUC-recommended reference code, CPT code 37231 (revascularization, endovascular, open or percutaneous, tibial, peroneal artery, unilateral, initial vessel; with transluminal stent placement(s) and atherectomy, includes angioplasty within the same vessel, when performed), to the work and time for CPT code 61645. We valued CPT code 61650 based on a crosswalk to CPT code 37221 (revascularization, endovascular, open or percutaneous, iliac artery, unilateral, initial vessel; with transluminal stent placement(s), includes angioplasty within the same vessel, when performed), due to similar intensity and intraservice time. We valued CPT code 61651 based on a crosswalk to CPT code 37223 (revascularization, endovascular, open or percutaneous, iliac artery, each additional ipsilateral iliac vessel; with transluminal stent placement(s), includes angioplasty within the same vessel, when performed (list separately in addition to the code for primary procedure)), due to similar intraservice time and intensity.

Both CPT codes 61645 and 61650 included postservice work time associated with a level 3 inpatient hospital visit. In the CY 2016 PFS final rule with comment period, we stated that we believe that for the typical patient, these services would be considered hospital outpatient services, not inpatient services. As a result, the intraservice time of the hospital observation care service was valued in the immediate postservice time. We refined the work time for CPT code 61645 by removing 55 minutes of work time associated with CPT code 99233, and added 30 minutes of time to the immediate postservice time. Therefore, the total time for CPT code 61645 was reduced to 241 minutes and the immediate postservice time increased to 83 minutes. We also removed the inpatient visit from CPT code 61650, which reduced the total time to 206 minutes and increased the postservice time to 75 minutes.

Comment on the CY 2016 PFS final rule with comment period: Commenters disagreed with our categorization of these codes as typically outpatient. Commenters stated that according to Medicare claims data, the predecessor codes were performed primarily on an inpatient basis. Additionally, commenters pointed out that the new codes would typically be performed on acute stroke patients. Commenters also said as the new codes are inpatient-only, the CMS reductions in work and time based on the assumption of outpatient status are flawed and suggested we accept the RUC-recommended values. Commenters also requested that these codes be referred to the refinement panel.

Response in the CY 2017 PFS proposed rule: For CY 2016, we valued CPT codes 61645, 61650, and 61651 based on comparisons to CPT codes 37231, 37221, and 37223, respectively. We continue to believe that these crosswalks are appropriate comparisons based on intensity and intraservice time, and because no persuasive information was presented at the refinement panel that indicated that these comparisons are not accurate. Therefore, for CY 2017, we proposed work RVUs of 15.00 for CPT code 61645, 10.00 for CPT code 61650, and 4.25 for CPT code 61651. We also proposed time inputs based on our refinements of the RUC recommendations, including removing the time associated with a hospital inpatient visit (CPT code 99233) from the intraservice work time, and adding 30 minutes to the immediate postservice time for both CPT codes 61645 and 61650.

We do not believe that 0-day global codes should include post-operative visits; rather, if global codes require post-operative visits, they are more appropriately assigned 10- or 90-day global periods based on our current criteria. Our policy has been to remove the visit from the post-operative period and the associated minutes from the total time while adding 30 minutes to the immediate postservice period without necessarily making an adjustment to the work RVU (see the CY 2010 PFS proposed rule, 74 FR 33557; also see the CY 2011 PFS proposed rule, 75 FR 40072). We solicited comment on the inclusion of post-operative visits in valuation of codes with 0-day global periods. Both CPT codes 61645 and 61650 are assigned 0-day global periods, and the refinements we proposed reflected changes to more appropriately value these codes with 0-day global periods.

The following is a summary of the comments we received regarding our proposed valuations for the intracranial endovascular intervention family:

Comment: Commenters, including the RUC, requested that CMS finalize the RUC-recommended work RVUs for CPT codes 61645, 61650 and 61651. The RUC suggested that evaluating the actual physician work performed in the inpatient setting is more accurate than applying a crosswalk to a CPT code that is performed predominantly in the outpatient setting. As examples, the RUC noted that CPT code 61645 would not be performed in the outpatient setting, and CPT codes 61650 and 61651 would be performed in the intensive care unit. For CPT codes 61645 and 61650, commenters also expressed concern about CMS' proposed refinements to remove the time associated with a postservice visit from each code and subsequently adding 30 minutes to the immediate postservice period for each of these codes. The RUC suggested that these CMS refinements artificially reduced the total work time for CPT codes 61645 and 61650.

Response: We continue to believe that our crosswalks for each of these codes accurately reflect the physician work involved in these procedures due to similarities in intensity and intraservice time. For example, our proposed work RVU of 15.00 for CPT code 61645 would be the highest work value among comparable codes with similar intraservice times. We note that we identified three CPT codes with similar intraservice times (CPT codes 33955, 33956, and 33988) that had higher work RVUs than our proposed work RVU of 15.00, but these three CPT codes are used to report extracorporeal membrane oxygenation or extracorporeal life support services (ECMO/ECLS) procedures, which we do not believe are comparable to the CPT codes in this family.

Regarding physician time for CPT codes 61645 and 61650, as we discussed in the proposed rule, we do not believe that 0-day global codes should include post-operative visits; rather, if global codes require post-operative visits, they are more appropriately assigned 10- or 90-day global periods based on our current criteria. Our policy has been to remove the visit from the post-operative period and the associated minutes from the total time while adding 30 minutes to the immediate postservice period without necessarily making an adjustment to the work RVU (see the CY 2010 PFS proposed rule, 74 FR 33557; also see the CY 2011 PFS proposed rule, 75 FR 40072).

Therefore, for CY 2017, we are finalizing a work RVU of 15.00 for CPT code 61645, a work RVU of 10.00 for CPT code 61650, and a work RVU of 4.25 for CPT code 61651.

(29) Epidural Injections (CPT Codes 62320, 62321, 62322, 62323, 62324, 62325, 62326, and 62327)

We proposed the RUC-recommended work RVU for all eight of the codes in this family.

We proposed to remove the 10-12mL syringes (SC051) and the RK epidural needle (SC038) from all eight of the codes in this family. We stated that these supplies were duplicative, as they are included in the epidural tray (SA064). As an alternative, we raised the possibility of removing the epidural tray and replacing it with the individual supply components used in each procedure; we solicited public comment on either the inclusion of the epidural tray or its individual components for this family of codes.

The following is a summary of the comments we received regarding our proposed valuation of the Epidural Injection codes:

Comment: A few commenters expressed their support for the proposed work values.

Response: We appreciate the support from the commenters.

Comment: Several commenters disagreed with the proposed removal of the 10-12mL syringes (SC051) and the RK epidural needle (SC038) due to the CMS belief that they are duplicative of the supplies in the epidural tray (SA064). Commenters stated that although there are three syringes listed in the epidural tray, none of the syringes in the tray are the 10-12mL syringe. In addition, none of the needles currently listed in the epidural tray (SA064) are an epidural needle. As a result, commenters indicated that there was no reason to replace the epidural tray with its individual components.

Response: We appreciate this clarification from the commenters regarding the components that make up the epidural tray. Taking this information into account, we are restoring the 10-12mL syringes (SC051) and the RK epidural needle (SC038) to all eight of the codes in this family.

After consideration of comments received, we are finalizing the proposed work RVUs for the Epidural Injection codes. We are also finalizing the proposed direct PE inputs, with the addition of the 10-12mL syringes and the RK epidural needle detailed above.

(30) Endoscopic Decompression of Spinal Cord (CPT Code 62380)

For CY 2016, the CPT Editorial Panel created CPT code 62380 to describe the endoscopic decompression of neural elements. The RUC recommended a work RVU of 10.47 based on a crosswalk to CPT code 47562 (Laparoscopy, surgical; cholecystectomy) with a higher intraservice time than reflected in the survey data. Since we believe CPT codes 62380 and 47562 are similar in intensity, we believe using the same work RVU as the crosswalk code overestimates the work involved in furnishing CPT code 62380. Reference CPT code 49507 (Repair initial inguinal hernia, age 5 years or older; incarcerated or strangulated) has a work RVU of 9.09 and has similar intensity and an identical intraservice time compared to CPT code 62380. Therefore, we proposed a work RVU of 9.09 for CPT code 62380.

Comment: Some commenters reiterated that the RUC-recommended direct crosswalk to CPT code 47562 is appropriate since this code has a similar physician time, and the IWPUT of the RUC-recommended work RVU is 0.085, a comparable valuation when compared with other spinal decompression procedures. The RUC agreed that the intensity of CPT code 62380 was greater, which offsets the 10 minute difference in intraservice time between the two codes. The RUC indicated that the difference in intensity between these procedures is based on CPT code 62380 involving decompression about neural elements and the spinal cord, where the opportunity for complications and for loss of function is high. One commenter indicated that CMS' proposed work RVU would fall below the minimum survey results.

A few commenters expressed concerns about the structure of the CPT code descriptors and RUC-recommended valuations. Commenters suggested that the CPT Editorial Panel and the RUC did not take certain indications into account such as differences between the physician work required for endoscopic tubular microdiscectomy compared to lumbar spinal stenosis decompression and posterior cervical posterior laminoforaminotomy. Commenters indicated that the specialty society survey data was inadequate due to the inexperience of the survey respondents, with others suggesting that the survey times were not reflective of some practitioners' experience or patient complexity.

The commenters indicated that the current RUC recommendations for full endoscopic tubular endoscopic surgery are based on limited experience among survey respondents with lumbar microdiscectomy, and insufficient experience with lumbar spinal stenosis decompression and posterior cervical foraminotomy without fusion and are invalid for these indications. Commenters requested that the current CPT codes and valuations for full endoscopic lumbar spinal stenosis decompression and posterior cervical foraminotomy without fusion remain unchanged until further RUC survey data are examined. Some commenters suggested alternative crosswalks including CPT code 61548 (Hypophysectomy or excision of pituitary tumor, transnasal or transseptal approach, nonstereotactic) with a work RVU of 23.37, CPT code 63030 (Laminotomy (hemilaminectomy), with decompression of nerve root(s), including partial facetectomy, foraminotomy and/or excision of herniated intervertebral disc; 1 interspace, lumbar) with a work RVU of 13.18, and CPT code 63056 (Transpedicular approach with decompression of spinal cord, equina and/or nerve root(s) (e.g., herniated intervertebral disc), single segment; lumbar (including transfacet, or lateral extraforaminal approach) (e.g., far lateral herniated intervertebral disc)) with a work RVU of 21.86.

Response: As discussed above, commenters raised multiple concerns about the accuracy of the survey results, the RUC's recommended valuation of this service, and our subsequent proposed refinements. Therefore, at this time, we are finalizing contractor pricing for CPT code 62380. We note that the summary of recommendations (SOR) included with the RUC recommendations indicated that the expert panel reviewing the survey data for this procedure believed the survey median and 25th percentile work RVU were inconsistent with the physician work as it related to other major open spine procedures. Subsequently, the RUC recommended a work RVU of 10.47 based on a crosswalk from CPT code 47562 (Laparoscopy, surgical; cholecystectomy). The RUC noted that procedures reported with CPT code 62380 have ten minutes less intraoperative time compared to the RUC's recommended crosswalk from CPT code 62380, but suggested that the physician work of endoscopic decompression in the small disc interspace near the spinal nerve roots of the cauda equina is more complex and will require more post-discharge office work for required imaging to confirm stabilization and for physical therapy orders and monitoring.

We note that based on the RUC's utilization crosswalk, services that will be reported in CY 2017 with CPT code 62380 are currently reported using either CPT code 22899 (Unlisted procedure, spine) or CPT code 0275T (Percutaneous laminotomy/laminectomy (interlaminar approach) for decompression of neural elements, (with or without ligamentous resection, discectomy, facetectomy and/or foraminotomy), any method, under indirect guidance (e.g., fluoroscopic, CT), with or without the use of an endoscope, single or multiple levels, unilateral or bilateral; lumbar)), which are both contractor priced for CY 2016. We welcome feedback from interested parties and specialty societies regarding valuation of this service for consideration in future rulemaking.

(31) Paravertebral Block Injection (CPT Codes 64461, 64462, and 64463)

In CY 2015, the CPT Editorial Panel created three new codes to describe paravertebral block injections at single or multiple levels, as well as for continuous infusion for the administration of local anesthetic for post-operative pain control and thoracic and abdominal wall analgesia. For the CY 2016 PFS final rule with comment period, we established the RUC-recommended work RVUs of 1.75 and 1.10 as interim final for CPT codes 64461 and 64462, respectively. For CPT code 64463, we utilized a direct crosswalk from three other injection codes (CPT codes 64416 (Injection, anesthetic agent; brachial plexus, continuous infusion by catheter (including catheter placement), 64446 (Injection, anesthetic agent; sciatic nerve, continuous infusion by catheter (including catheter placement), and 64449 (Injection, anesthetic agent; lumbar plexus, posterior approach, continuous infusion by catheter (including catheter placement)), which all had a work RVU of 1.81, as we believed this crosswalk more accurately reflected the work involved in furnishing this service.

Comment on the CY 2016 PFS final rule with comment period: We received comments from the RUC stating CPT code 64463 was more comparable to CPT code 64483 (Injection(s), anesthetic agent and/or steroid, transforaminal epidural, with imaging guidance (fluoroscopy or CT); lumbar or sacral, single), which has a work RVU of 1.90 and requires the same physician work and time to perform. The RUC recommended we accept a work RVU of 1.90, which is the 25th percentile of the survey. Another commenter stated that our interim final work RVU for CPT code 64463 was inappropriate since imaging guidance is not part of our comparison codes. The commenter advocated for us to accept the survey respondent's selection of CPT code 64483 as the most appropriate comparison code and assign a work RVU of 1.90.

Response in the CY 2017 PFS proposed rule: After reviewing and considering the comments, we stated we continued to believe that CPT codes 64416, 64446, and 64449, all of which have 20 minutes of intraservice time, are better crosswalks to CPT code 64463, which also has 20 minutes of intraservice time and a similar total time. In contrast, the crosswalk code recommended by commenters, CPT 64483, only has 15 minutes of intraservice time. Therefore, for CY 2017 we proposed a work RVU of 1.81 for CPT code 64463.

The following is a summary of the comments we received regarding our proposed valuations for the Paravertebral Block Injection family:

Comment: One commenter stated that CMS based its decision on an inappropriate comparison of CPT code 64463 with codes that describe continuous peripheral nerve blocks that do not include imaging guidance. The commenter stated that the imaging component included in CPT code 64463 was justification for at least the 0.09 difference between the RUC recommendation and the CMS proposed value. The commenter offered CPT code 47000 (Biopsy of liver, needle; percutaneous), which has identical intraservice time and a work RVU of 1.90 as a comparator code.

Response: We appreciate the additional information offered by the commenters and we agree with the commenter's statement that the image guidance component of this service was justification for the 0.09 difference between the RUC recommendation and the CMS proposed value. After review and consideration of the comments, we are finalizing the RUC-recommended work RVUs of 1.75, 1.10 and 1.90 for CPTs code 64461, 64462 and 64463, respectively for CY 2017.

(32) Implantation of Neuroelectrodes (CPT Codes 64553 and 64555)

The RUC identified CPT codes 64553 and 64555 as a site of service anomaly during the CY 2016 PFS rulemaking cycle. In the Medicare claims data, these services were typically reported in the nonfacility setting, yet the survey data were predicated on a facility-based procedure. We agreed with the RUC that these two codes should be referred to the CPT Editorial Panel to better define the services, in particular to investigate the possibility of establishing one code to describe temporary or testing implantation and another code to describe permanent implantation. We maintained the CY 2015 work RVUs and direct PE inputs for these two codes on an interim basis until receiving updated recommendations from the CPT Editorial Panel and the RUC.

Comment on the CY 2016 PFS final rule with comment period: A commenter requested that CMS allow practitioners to bill the MACs separately for a percutaneous electrode kit (SA022) for CPT code 64555. The commenter stated that without allowing for a separate payment for the percutaneous electrode kit, the payment for the procedure would be insufficient to cover the physician's costs.

Response in the CY 2017 PFS proposed rule: We agreed that CPT codes 64553 and 64555 as currently constructed were potentially misvalued codes, which is why we maintained the CY 2015 work RVUs and direct PE inputs on an interim basis. We believe that the disposable supplies furnished incident to the procedure are paid through the nonfacility PE RVUs. The percutaneous electrode kit (SA022) was not previously included in the direct PE inputs for either of these two services, and since we proposed to maintain current direct PE inputs pending additional recommendations, we do not agree that disposable supplies should be separately payable. We proposed to maintain the interim final work RVUs and direct PE inputs for these two codes, and we looked forward to reviewing recommendations regarding these procedures again for future rulemaking.

Additionally, we were alerted to a discrepancy regarding the times for these codes in the CY 2016 work time file. Our proposed CY 2017 work time file addressed this discrepancy by reflecting the RUC recommended times of 155 minutes for CPT code 64553 and 140 minutes for CPT code 64555.

The following is a summary of the comments we received regarding our proposed valuation of the Implantation of Neuroelectrodes codes:

Comment: One commenter responded to the CMS request for information about whether there was a need for separate codes for temporary/testing and permanent placement for neuroelectrodes. The commenter stated that it did not support the creation of new separate codes at this time. The commenter stressed that the current codes account for the work of both temporary/testing and permanent placement, making the creation of new codes unwarranted.

Response: We appreciate the submission of this information from the commenter. We did not receive any comments addressing the proposed valuation of these codes.

After consideration of comments, we are finalizing the proposed work RVUs and proposed direct PE inputs for CPT codes 64553 and 64555.

(33) Ocular Reconstruction Transplant (CPT Code 65780)

In CY 2015, the RUC identified CPT code 65780 as potentially misvalued through a misvalued code screen for 90-day global services that included more than 6 office visits. The RUC recommended a direct work RVU crosswalk from CPT code 27829 (Open treatment of distal tibiofibular joint (syndesmosis) disruption, includes internal fixation, when performed). After examining comparable codes, we determined the RUC-recommended work RVU of 8.80 for CPT code 65780 would likely overstate the work involved in the procedure given the change in intraservice and total times compared to the previous values. We believed that the ratio of the total times (230/316) applied to the work RVU (10.73) more accurately reflected the work involved in this procedure. Therefore, we established an interim final work RVU of 7.81 for CPT code 65780.

Comment on the CY 2016 PFS final rule with comment period: The RUC and other commenters disagreed with our interim final values based on objections to our use of time ratios in developing work RVUs for PFS services.

Response in the CY 2017 PFS proposed rule: We stated that we appreciate the commenters' concerns and responded to these concerns about our methodology in section II.L of the CY 2017 proposed rule. After review of the comments, we continued to consider the work RVU of 7.81 to accurately represent the work involved in CPT code 65780. We believed this service was similar in overall intensity to CPT code 27766 (Open treatment of medial malleolus fracture, includes internal fixation, when performed) that has a work RVU of 7.89 and a total time that more closely approximates that of CPT code 65780.

In the CY 2017 proposed rule, we proposed a work RVU of 7.81 for CPT code 65780.

We did not receive any comments in response to our proposed valuation on CPT code 65780; therefore, we are finalizing a work RVU of 7.81 as proposed.

(34) Trabeculoplasty by Laser Surgery (CPT Code 65855)

In CY 2015, the RUC identified CPT code 65855 as potentially misvalued through the review of 10-day global services with more than 1.5 postoperative visits. The RUC noted that the code was changed from a 90-day to a 10-day global period when it was last valued in 2000. However, the descriptor was not updated to reflect that change. CPT code 65855 describes multiple laser applications to the trabecular meshwork through a contact lens to reduce intraocular pressure. The current practice is to perform only one treatment session during a 10-day period and then wait for the effect on the intraocular pressure. The descriptor for CPT code 65855 has been revised and removes the language “1 or more sessions” to clarify this change in practice.

The RUC recommended a work RVU of 3.00 for CPT code 65855. While the RUC-recommended value represents a reduction from the CY 2015 work RVU of 3.99, we stated that significant reductions in the intraservice time, the total time, and the change in the office visits represent a more significant change in the work resources involved in furnishing the typical service. The intraservice and total times were decreased by approximately 33 percent while the elimination of two postoperative visits (CPT code 99212) alone would reduce the overall work RVU by at least 24 percent under the reverse building block method. However, the RUC-recommended work RVU only represents a 25 percent reduction relative to the previous value. To identify potential work RVUs for this service, we calculated an intraservice time ratio between the CY 2015 intraservice time, 15 minutes, and the RUC-recommended intraservice time, 10 minutes, and applied this ratio to the current work RVU of 3.99 to arrive at a work RVU of 2.66 for CPT code 65855, which we established as interim final for CY 2016.

Comment on the CY 2016 PFS final rule with comment period: A few commenters, including the RUC, provided explanations as to how the RUC recommendation had already accounted for the reduction in physician intraservice time and post-operative visits. Some commenters disagreed with CMS' interim final values based on objections to CMS' use of time ratios in developing work RVUs for PFS services.

Response in the CY 2017 PFS proposed rule: We stated that we appreciated the commenters' concerns regarding the time ratio methodologies and responded to those concerns about our methodology in section II.H.2 of the CY 2017 proposed rule. After considering the explanations provided by commenters through public comments describing the RUC's methodologies in more detail, we agreed that the proposed value did not accurately reflect the physician work involved in furnishing the service.

In the CY 2017 proposed rule, we proposed the RUC-recommended work RVU value of 3.00 for CPT code 65855.

We did not receive any comments in response to our proposed valuation on CPT code 65855; therefore, we are finalizing a work RVU of 3.00 as proposed.

Comment: A few commenters stated their support of CMS' decision to propose the RUC-recommended value for CY 2017 and strongly urged us to finalize the proposal.

Response: Thank you for your comments. For CY 2017 we are finalizing the RUC-recommended work RVU of 3.00 for CPT code 65855.

(35) Glaucoma Surgery (CPT Codes 66170 and 66172)

The RUC identified CPT codes 66170 and 66172 as potentially misvalued through a screen for 90-day global codes that included more than six office visits. We believed the RUC-recommended work RVU of 13.94 for CPT code 66170 did not accurately account for the reductions in time. Specifically, the survey results indicated reductions of 25 percent in intraservice time and 28 percent in total time. These reductions suggested that the RUC-recommended work RVU for CPT code 66170 overstated the work involved in furnishing the service, since the recommended value only represented a reduction of approximately seven percent. We believed that applying the intraservice time ratio, the ratio between the CY 2015 intraservice time, 60 minutes, and the RUC-recommended intraservice time, 45 minutes, applied to the current work RVU, 15.02, resulted in a more appropriate work RVU of 11.27. Therefore, for CY 2016, we established an interim final work RVU of 11.27 for CPT code 66170.

For CPT code 66172, the RUC recommended a work RVU of 14.81. After comparing the RUC-recommended work RVU for this code to the work RVU for similar codes (for example, CPT code 44900 (Incision and drainage of appendiceal abscess, open) and CPT code 52647 (Laser coagulation of prostate, including control of postoperative bleeding, complete (vasectomy, meatotomy, cystourethroscopy, urethral calibration and/or dilation, and internal urethrotomy are included if performed))), we believed the RUC-recommended work RVU of 14.81 overstated the work involved in this procedure. For the same reasons and following the same valuation methodology utilized above, we applied the intraservice time ratio between the CY 2015 intraservice time and the survey intraservice time, 60/90, to the CY 2015 work RVU of 18.86. This resulted in a work RVU of 12.57 for CPT code 66172. Therefore, for CY 2016, we established an interim final work RVU of 12.57 for CPT code 66172.

Comment on the CY 2016 PFS final rule with comment period: Several commenters, including the RUC, objected with our interim final values based on objections to our use of time ratios in developing work RVUs for PFS services. Commenters also requested CMS refer CPT codes 66170 and 66172 to the refinement panel.

Response in the CY 2017 PFS proposed rule: We acknowledged commenters' concerns regarding the time ratio methodologies and responded to those concerns in section II.H.2 of the CY 2017 proposed rule (81 FR 46162). CPT codes 66170 and 66172 were referred to the CY 2016 multi-specialty refinement panel per commenters' request. The outcome of the refinement panel was a median of 13.94 RVUs for CPT code 66170 and 14.84 RVUs for CPT code 66172. Due to the new information presented to the refinement panel regarding the level of intensity required to perform millimeter incisions in the eye, we agreed with the assessment of the refinement panel and proposed a work RVU of 13.94 for CPT code 66170 and 14.84 for CPT code 66172 for CY 2017.

The following is a summary of the comments we received regarding our proposed valuations for the Glaucoma Surgery family:

Comments: Several commenters stated their support of CMS' decision to propose the values recommended by the refinement panel for CPT codes 66170 and 66172. Some commenters, including the RUC, also brought to our attention discrepancies between our proposal for these codes in the CY 2017 proposed rule and the work RVUs posted in Addendum B on the CMS Web site.

Response: For CY 2017, we are finalizing a work RVU of 13.94 for CPT code 66170 and a work RVU of 14.84 for CPT code 66172. We appreciate commenters bringing this issue regarding conflicting information in the CY 2017 PFS proposed rule preamble text and the public use files published on the CMS Web site. We have corrected this discrepancy in this final rule and the associated public use files.

(36) Retinal Detachment Repair (CPT Codes 67101, 67105, 67107, 67108, 67110, and 67113)

For CY 2015, the CPT Editorial Panel made several changes to CPT codes 67101 and 67105. These changes include revising the code descriptors to exclude “diathermy” and “with or without drainage of subretinal fluid” and removing the reference to “1 or more sessions.” The recommended global period also changed from 90 days to 10 days. For CPT code 67101, we proposed the RUC recommended work RVU of 3.50, which was based on the 25th percentile of the survey. For CPT code 67105, the RUC recommended a work RVU of 3.84 based on the 25th percentile of the survey. The RUC also stated that CPT code 67105 was a more intense procedure, and therefore, it should have a higher work RVU than CPT code 67101. Currently, CPT code 67101 has a higher work RVU than CPT code 67105 and according to the surveys, the intraservice and total times remain higher for CPT code 67101. We do not understand why the RUC believes that CPT code 67105 is more work than CPT code 67101. Therefore, we did not propose the RUC-recommended work RVU of 3.50 for CPT code 67105. We did not find evidence that CPT code 67105 is more intense than CPT code 67101 and accordingly, proposed a lower work RVU for CPT code 67105. To value CPT code 67105, we used the RVU ratio between CPT codes 67101 and 67105. We divided the current work RVU of 8.53 for CPT code 67105, by the current work RVU of 8.80 for CPT code 67101 and multiplied the quotient by the RUC-recommended work RVU of 3.50 for CPT code 67101 to arrive at a work RVU of 3.39. Therefore, for CY 2017, we proposed a work RVU of 3.39 for CPT code 67105.

CPT codes 67107, 67108, 67110, and 67113 were identified through the Relative Assessment Workgroup process under the 90-day global post-operative visit screen in CY 2015. The RUC recommended a work RVU of 16.00 for CPT code 67107, which corresponded to the 25th percentile of the survey. While the RUC recommendation represented a five percent reduction from the current work RVU of 16.71, we believed the RUC recommendation still overvalued the service given the 15 percent reduction in intraservice time and 25 percent reduction in total time.

We used the intraservice time ratio between the existing and new time values to identify an interim final work RVU of 14.06. We believed this value accurately reflected the work involved in this service and was comparable to other codes that have the same global period and similar intraservice time and total time. For CY 2016, we established an interim final work RVU of 14.06 for CPT code 67107. For CPT code 67108, the RUC recommended a work RVU of 17.13 based on the 25th percentile of the survey, which reflected a 25 percent reduction from the current work RVU. The survey results reflected a 53 percent reduction in intraservice time and a 42 percent reduction in total time. We believed the RUC-recommended work RVU overestimated the work, given the significant reductions in intraservice time and total time and does not maintain relativity among the codes in this family. To determine the appropriate value for this code and maintain relativity within the family, we preserved the 1.13 work RVU increment recommended by the RUC between this code and CPT code 67107 and applied that increment to the interim final work RVU of 14.06 for CPT code 67107. Therefore, we established an interim final work RVU of 15.19 for CPT code 67108. For CPT code 67110, the RUC recommended maintaining the current work RVU of 10.25. To maintain appropriate relativity with the work RVUs established for the other services within this family, we used the RUC-recommended 5.75 work RVU differential between CPT code 67107 and CPT code 67110 to establish the CY 2016 interim final work RVU of 8.31 for CPT code 67110. For CPT code 67113, the RUC recommended and we established an interim final work RVU of 19.00 based on the 25th percentile of the survey.

Comment on the CY 2016 PFS final rule with comment period: We received several comments disagreeing with our interim final values based on objections to our use of time ratios in developing work RVUs for PFS services. Some commenters also stated that by using some RUC-recommended increments and rejecting others, we have not only established inconsistencies within the family of codes, but potentially opened up anomalies across a wide range of services. The RUC also expressed disagreement with using the recommended work RVU increments without using the recommended work RVU. Some commenters also stated the new IWPUT values for these three services are inappropriately low and pointed to the derived per minute intensity of 0.064 for CPT code 67110 as particularly problematic.

Response in the CY 2017 PFS proposed rule: We disagreed with the statement about inconsistencies as the codes in this family are valued relative to one another based on the times and level of physician work required for each code.

We also stated that generally we do not agree that a low IWPUT itself indicates overall misvaluation as the validity of the IWPUT as a measure of intensity depends on the accuracy of the assumptions regarding the number, level, and work RVUs attributable to visits for services in the post-operative global period for individual services.

We provided an example where a service with an unrealistic number or level of postoperative visits may have a very low derived intensity for the intra-service time. CPT codes 67107, 67108, and 67110 were referred to the CY 2016 multispecialty refinement panel per commenters' request. The outcome of the refinement panel was a median work RVU of 16.00, 17.13, and 10.25, respectively. After consideration of the comments and the results of the refinement panel, we proposed a work RVU of 16.00, 17.13, and 10.25 for CPT codes 67107, 67108, and 66110, respectively, for CY 2017.

The following is a summary of the comments we received regarding our proposed valuations for the Retinal Detachment Repair family:

Comments: A few commenters, including the RUC, noted that CPT codes 67101 and 67105 were last valued by the Harvard study. The RUC stated that during the Harvard studies, CPT code 67101 was valued higher due to greater total time. However, now photocoagulation is reported at vastly higher levels than the cryotherapy procedure, as it is considered to be a more effective treatment. A few commenters stated that given the changing nature of the service since the last valuation, the intensity of CPT code 67105 is now greater and urged CMS to accept the RUC-recommended values.

For CPT codes 67107, 67108, 67110, and 67113, several commenters supported CMS' decision to propose the values recommended by the refinement panel and urged CMS to finalize these proposed values. A few commenters, including the RUC, brought to our attention discrepancies between our proposal for these codes and the work RVUs posted in Addendum B on the CMS Web site.

Response: We note that, according to the surveys, the intraservice and total times were significantly higher for CPT code 67101 and note the specialty societies recommended a higher work RVU for CPT code 67101 prior to the RUC meeting. Although commenters state that photocoagulation (CPT code 67105) is typically billed more frequently than diathermy (CPT code 67101), we do not believe the utilization rate of a service in and of itself is reason enough to warrant an increase in RVUs. Therefore, for CY 2017, we are finalizing a work RVU of 3.50 and 3.39 for CPT codes 67101 and 67105, respectively.

We appreciate commenters bringing to our attention the issue regarding conflicting information in the CY 2017 PFS proposed rule preamble text and the public use files published on the CMS Web site. We have corrected this discrepancy in this final rule and the public use files.

For CY 2017, we are finalizing a work RVU of 16.00, 17.13, 10.25 and 19.00 for CPT codes 67107, 67108, 66110 and 67113, respectively, in agreement with the refinement panel recommendations.

(37) Fetal MRI (74712 and 74713)

For CY 2016, we established the RUC-recommended work RVU of 3.00 as interim final for CPT code 74712. We established an interim final work RVU of 1.78 for CPT code 74713 based on a refinement of the RUC-recommended work RVU of 1.85 using the ratio of work to time for both codes. This proposed value also corresponds to the 25th percentile survey result.

Comment on the CY 2016 PFS final rule with comment period: Commenters stated that the work RVU of 1.78 for CPT code 74713 did not reflect the higher intensity inherent in the procedure's typical patient. The commenter explained that the typical patient is pregnant with twins and has a higher likelihood of complications related to congenital anomalies, as well as of ischemic brain injury with twin gestations. The commenter further stated that twin gestations are more difficult to image. Commenters requested that CPT code 74713 be referred to the multispecialty refinement panel.

Response in the CY 2017 PFS proposed rule: CPT code 74713 was referred to the CY 2016 multispecialty refinement panel. After considering the comments and the results of the refinement panel, we agreed with commenters that an RVU of 1.78 underestimated the work for CPT code 74713.

In the CY 2017 proposed rule, we proposed a work RVU of 1.85 for the service for CY 2017.

We did not receive any comments in response to our proposed valuation on CPT code 74713; therefore, we are finalizing the proposed work RVU.

(38) Abdominal Aortic Ultrasound Screening (CPT Code 76706)

For CY 2017, the CPT Editorial Panel created a new code, CPT code 76706, to describe abdominal aortic ultrasound screening, currently described by HCPCS code G0389. The specialties that surveyed CPT code 76706 for the RUC were vascular surgery and radiology, and the direct PE inputs recommended by the RUC included an ultrasound room. Based on an analysis of Medicare claims data, the dominant specialties furnishing the service are family practice and internal medicine. We believe that these specialties may more typically use a portable ultrasound device rather than an ultrasound room. Therefore, we proposed to accept the RUC-recommended work RVU of 0.55, and the RUC-recommended PE inputs for this service, but we solicited comment regarding whether or not it would be more accurate to substitute a portable ultrasound device or possibly a hand-held device for an ultrasound room for CPT code 76706. We note that while the phase-in of significant reductions in RVUs ordinarily would not apply to new codes, we believe that it would be appropriate to consider this change from a G-code to a CPT code to be fundamentally similar to an editorial coding change since the service is not described differently, and therefore, we proposed to apply the phase-in to this service by comparing the previous value of the G-code to the value for the new CPT code.

Comment: One commenter stated that this service should be furnished by a physician or surgeon that specializes in vascular disease. The commenter noted that CMS should assign inputs based on which specialties would more appropriately furnish a given service. Another commenter disagreed with our statement in the CY 2017 proposed rule that the dominant specialties furnishing this service are family practice and internal medicine. The commenter stated that these specialties are more likely to make use of a portable ultrasound device rather than an ultrasound room. One commenter says that this service is underutilized, and CMS should implement policies which support screening.

Response: We appreciate the commenters' perspectives. We note that, in evaluating codes in the Medicare Physician Fee Schedule (MPFS), we price codes based on the typical service. Our review of the Medicare claims data indicates that the combined utilization for the technical component of this service and the service billed globally is typically billed under the PFS by family practice and internal medicine, which is why we solicited comment on whether the PE inputs for this service should be revised.

Comment: A commenter supported our decision to apply the phase-in to this code.

Response: We thank the commenter for the support.

Comment: A commenter agreed with CMS that family practice physicians typically use a portable ultrasound device rather than an ultrasound room. The commenter stated that CMS should continue to include an ultrasound room as a direct PE input, unless other specialties furnishing the service indicate that they do not typically make use of an ultrasound room.

One commenter states that abdominal aortic aneurysm screenings are performed on nonportable machines in either ambulatory or hospital settings, and therefore, an ultrasound room is appropriate.

Response: We thank the commenters, and we will take this information regarding the appropriate PE inputs for this service into consideration for future rulemaking. While the specialty mix of the practitioners furnishing services can be helpful in identifying typical PE inputs, we continue to seek definitive information regarding the most appropriate PE inputs for this code. For CY 2017, we are finalizing the RUC-recommended work and PE inputs, as proposed.

(39) Fluoroscopic Guidance (CPT Codes 77001, 77002, and 77003)

In the CY 2015 PFS final rule with comment period, CMS indicated that while CPT codes 77002 and 77003 had been previously classified as stand-alone codes without global periods, we believe their vignettes and CPT Manual parentheticals are consistent with an add-on code as has been established for CPT code 77001. Therefore, the global periods for CPT codes 77002 and 77003 now reflect an add-on code global period with modifications to the vignettes and parentheticals.

For CPT code 77001, we proposed the RUC-recommended work RVU of 0.38. We stated that the RUC-recommended work RVUs for CPT codes 77002 and 77003 did not appear to account for the significant decrease in total times for these codes relative to the current total times. We noted that these three codes describe remarkably similar services and have identical intraservice and total times. Based on the identical times and notable similarity for all three of these codes, we proposed a work RVU of 0.38 for all three codes.

The following is a summary of the comments we received regarding our proposed valuation of the Fluoroscopic Guidance codes:

Comment: A few commenters disagreed with the change in the global period for CPT codes 77002 and 77003 to reflect their status as add-on codes. The commenters stated that this would imply that the imaging-related preservice and postservice activities inherent to these image guidance codes are captured by the base codes with which they are reported, which simply is not the case. The commenters provided an example of how reporting of radiation specific information, such as fluoroscopy time, is not included in the postservice activities of the base codes.

Response: CPT codes 77002 and 77003 were surveyed under the assumption that they would be classified as add-on codes, and the RUC recommendations for both work RVUs and direct PE inputs reflect this status. We do not believe that it would be appropriate to assign these codes a different global period after they were surveyed and valued with the understanding that they would be classified as add-on codes.

Comment: Many commenters disagreed with the proposed work RVU of 0.38 for CPT codes 77002 and 77003. Commenters stated that these two codes should not share the same work RVU as CPT code 77001, on the basis that the physician work, intensity and complexity of codes 77002 and 77003 are greater than the first code in the family. Commenters stated that the intensity and complexity increases in parts of the body where there are additional anatomy considerations, such as superficial and deep structures to consider with CPT code 77002, as well as additional neuro and spinal structures to consider when performing CPT code 77003. One commenter suggested that there was clinical data indicating that CPT codes 77002 and 77003 take longer to perform than CPT code 77001, in contradiction of the RUC survey data that assigned all three codes identical time values. The commenter stated that this was likely due to the greater complexity and procedural variability of the latter two codes. Another commenter recognized that these codes describe similar services but stressed that they do not describe identical services, which was especially important for CPT code 77003 as it pertains to spinal procedure and carries more risk than the other two codes.

Response: We recognize the concerns raised by the commenters in assigning the same work RVU of 0.38 to the three codes in the Fluoroscopic Guidance family. We note that even in cases where we assign the same work RVU, we do not believe that the services are identical, only that they share the same overall resources in work as measured in RVUs. We also appreciate the reference to additional clinical data from one commenter suggesting that CPT codes 77002 and 77003 take longer to perform than CPT code 77001. We have longstanding concerns about using survey data alone for code valuation, and we are always interested in investigating additional sources of information to assist in this process. We encourage future commenters to submit this data as part of their public comment so that it can be used by CMS as part of the code valuation process. Based on the submission of this additional data, we believe that the CPT codes 77002 and 77003 are more accurately valued at a higher RVU than CPT code 77001.

After consideration of comments received, we are finalizing the RUC-recommended work RVUs for all three codes in the family, which is an increase from the proposed work RVU of 0.38 to a work RVU 0.54 for CPT code 77002 and to 0.60 for CPT code 77003. We are finalizing the proposed work RVU of 0.38 for CPT code 77001 without change.

(40) Mammography—Computer Aided Detection Bundling (CPT Codes 77065, 77066 and 77067)

Section 104 of the Medicare, Medicaid, and SCHIP Benefits Improvement and Protection Act of 2000 (BIPA) (Pub. L. 106-554) required us to create separate codes with higher payment amounts for digital mammography compared to film mammography, which was the technology considered to be typical at the time. In addition, the statute required additional payment to be made when computer-aided detection (CAD) was used.

In CY 2002, we began valuing digital mammography services using three G-codes, G0202, G0204, and G0206 to describe screening mammography, unilateral diagnostic mammography, and bilateral diagnostic mammography, respectively. CMS implemented the requirements of BIPA section 104(d)(1), which applied to tests furnished in 2001, by using the work RVUs of the parallel CPT codes, but establishing a fixed PE RVU rather than using PE RVUs developed under the standard PE methodology. The fixed amount of PE RVUs for these codes has generally remained unchanged since implementation of the G-codes that specifically described digital imaging.

Most mammography services under Medicare have since been billed with these G-codes when digital mammography was used, and with CPT codes 77055, 77056, and 77057 when film mammography was used. The use of CAD has been reported with CPT codes 77051 and 77052. For CY 2017, the CPT Editorial Panel deleted CPT codes 77051, 77052, 77055, 77056, 77057 and created three new CPT codes, 77065, 77066, and 77067, to describe mammography services bundled with CAD. For CY 2017, the RUC recommended work RVUs of 0.81 for CPT code 77065, 1.00 for CPT code 77066, and 0.76 for CPT code 77067, as well as new PE inputs for use in developing resource-based PE RVUs based on our standard methodologies. The RUC recommended these inputs and only one medical specialty society provided us with a set of single invoices to price the equipment used in furnishing these services.

We reviewed these coding changes and proposed changes to valuation for these codes for CY 2017. The revised CPT coding mitigates the need for both separate G-codes and the CAD add-on codes. Based upon these coding changes and the RUC-recommended input values, overall Medicare payment for mammography services would be drastically reduced. This is particularly true for the technical component of these services, which could possibly be reduced up to 50 percent relative to the PE RVUs currently used for payment for these services.

Based on our initial review of the recommended inputs for the new codes, we believed that these changes would likely result in values more closely related to the relative resources involved in furnishing these services. However, we recognized that these services, particularly the preventive screenings, are of particular importance to the Medicare program and the health of Medicare beneficiaries. We were concerned that making drastic changes in coding and payment for these services could be disruptive in ways that could adversely impact beneficiary access to necessary services. We also recognized that unlike almost any other high-volume PFS service, the RVUs used for payment for many years have not been developed through the generally applicable PFS methodologies, and instead reflect the statutory directive under section 104 of the BIPA. Similarly, we recognized that the changes in both coding and valuation are significant changes for those who provide these services. Therefore, instead of proposing to simultaneously adopt the revised CPT coding and drastic reductions in overall payment rates, we believed it was advisable to propose to adopt the new coding, including the elimination of separate billing for CAD, for CY 2017 without proposing immediate implementation of the recommended resource inputs. We anticipated that we would consider the recommended inputs, including the pricing of the required equipment, as carefully as possible prior to proposing revised PE values through subsequent rulemaking.

Therefore, for CPT codes 77065, 77066, and 77067, we proposed to accept the RUC-recommended work RVUs, but to crosswalk the PE RVUs for the technical component of the current corresponding G-codes, as we sought further pricing information for these equipment items.

Since the publication of the proposed rule, we have determined that for several reasons related to claims processing systems, Medicare claims systems will be unable to process claims using CPT codes 77065, 77066, and 77067 for CY 2017. However, given the parallel structure of these new CPT codes, 77065, 77066, and 77067 to existing G-codes G0206, G0204, and G0202, we anticipate that the claims systems will be fully capable of processing the appropriate payment policies and prices discussed below for CPT codes 77065, 77066, and 77067 by using the existing G-codes. Therefore, for CY 2017, we will operationalize the new coding rules, including adoption of the new code descriptors for CPT codes 77065, 77066, and 77067 through use of the three current G-codes. For the purposes of discussion below, we discuss policies and payment rates for these three codes using the CPT numbers. Therefore, in the preamble discussion below, references to the G-codes refer to the descriptors, policies, and rates for CY 2016 and references to the new CPT codes refer to the 2017 descriptors, policies and rates that will be implemented through revisions to the current G-codes. We anticipate being able to adopt the CPT coding for CY 2018.

In addition to soliciting comment on this proposal, we also solicited input on rates for these services in the commercial market to help us understand the potential impacts of any future proposed revisions to PFS payment rates.

Finally, we noted that by proposing to adopt the new coding for CY 2017, any subsequent significant reduction in RVUs (greater than 20 percent) for the codes would be subject to the statutory phase-in under section 1848(c)(7).

To help us examine the resource inputs for these services, we solicited public comment on the list of items recommended as equipment inputs for mammography services. We also invited commenters to provide any invoices that would help with future pricing of these items.

We also received specialty society recommendations for a new Equipment Item, a physician PACS mammography workstation. We note that we discuss physician PACS workstation in section II.A of this rule. The items that comprise the physician PACS mammography workstation are listed in Table 19. We requested public comment as to the appropriateness of this list and if some items are indirect expenses or belong in other codes. We also invited commenters to provide any invoices that would help with future pricing of these items.

We also note that for CY 2015, the CPT Editorial Panel created CPT codes 77061, 77062, and 77063 to describe unilateral, bilateral, and screening digital breast tomosynthesis, respectively. CPT code 77063 is an add-on code to CPT code 77057, the CPT code for screening mammography. To be consistent with our use of G-codes for digital mammography, we did not implement two of these three CPT codes for Medicare purposes. We only adopted CPT code 77063 as an add-on code to HCPCS code G0202. Instead of adopting stand-alone CPT codes 77061 and 77062, we created a new code, G0279 Diagnostic digital breast tomosynthesis, as an add-on code to the diagnostic digital mammography HCPCS codes G0204 and G0206 and assigned it values based on CPT code 77063. Pending revaluation of the mammography codes using direct PE inputs, we proposed in CY 2017 to maintain the current coding structure for digital breast tomosynthesis with the technical change that HCPCS code G0279 be reported with CPT codes 77065 or 77066 as the replacement codes for HCPCS codes G0204 and G0206.

Comment: Many commenters expressed support for our decision to prevent a drastic reduction in payment for the technical component of these services by maintaining the PE RVUs from CMS' digital mammography coding. A few commenters expressed concern that shifting to our standard resource-based PE valuation methodology in future rulemaking would drastically reduce payments. Some commenters agreed that CMS does not have sufficient pricing data to value digital mammography. One commenter stated that the RUC-recommended direct PE inputs do not need to be re-considered, as they include pricing data provided by the specialty that most frequently furnishes the service.

Response: We will continue to carefully consider the potential negative impact that our valuation of these services will have on beneficiary access as we evaluate all relevant sources of data in future rulemaking, including data provided by the RUC.

Comment: A commenter did not support our intention to seek more pricing information in the commercial market, stating that commercial payers are generally more responsive to market incentives to reduce rather than increase prices.

Response: We refer readers to the CY 2010 PFS final rule with comment period (74 FR 61743 through 61748) that describes CMS' methodology in evaluating practice expense. We would consider a variety of different data sources, pending their availability and applicability. We believe that having more information regarding pricing in the commercial market may help us to contextualize recommended pricing, as well as potential impact of significant changes in payment.

Comment: One commenter expressed concern that, despite our maintenance of PE RVUs and our acceptance of RUC-recommended work RVUs, these services will still see significant payment reductions.

Response: We are accepting the RUC-recommended work RVUs, which equal the sum of the base code work RVUs for mammography and for CAD. The work RVUs for the new mammography coding are therefore not changing from their current values. Furthermore, as we are retaining the PE RVUs from the digital mammography G-codes in the new coding, the practice expense valuation is not changing. Therefore, payment amounts for mammography services will not see significant reductions for CY 2017. We expect to revalue these services through our standard code valuation process in future rulemaking.

Comment: One commenter said that CMS should accept the RUC-recommended direct PE inputs.

Response: As noted earlier, we did not propose the RUC-recommended inputs for these three codes for several reasons, including our concerns that drastic changes in coding and payment for these services could be disruptive in ways that could adversely affect beneficiary access to necessary services, and that unlike almost any other high-volume PFS service, the RVUs used for payment for many years have not been developed through the generally applicable PFS methodologies. Therefore, instead of proposing to simultaneously adopt the revised CPT coding and drastic reductions in overall payment rates, we believed it was advisable to propose to adopt the new coding, including the elimination of separate billing for CAD, for CY 2017 without proposing immediate implementation of the recommended resource inputs.

Comment: One commenter requested clarification regarding if the PE RVUs were valued using the RUC-recommended direct PE inputs, as these inputs were posted in Public Use Files (PUFs) for the CY 2017 Proposed Rule.

Response: We thank the commenter for pointing out that direct PE inputs were posted for these codes. These inputs were inadvertently included in the Public Use Files. We reiterate that we are not implementing PE inputs for these services, and we are instead crosswalking the PE RVUs from the digital mammography HCPCS codes G0202, G0204, and G0206, as doing so prevents a drastic reduction in payments. We included potential direct PE inputs in the text of the CY 2017 proposed rule to facilitate public comment and information in anticipation of developing updated PE RVUs for these services in future rulemaking.

Comment: A commenter stated that this coding violates statutory requirements set forth by BIPA that required the agency to: (1) create separate codes with higher payment amounts for digital mammography compared to film mammography and (2) pay separately when computer-aided detection (CAD) was used.

Response: The BIPA requirements specifically refer to screening and diagnostic mammography furnished during the period beginning on April 1, 2001, and ending on December 31, 2001. CMS chose to retain the payment rates for the technical component following this period.

Comment: A number of commenters volunteered to help CMS in pricing direct PE inputs for these services.

Response: We thank the commenters and seek as much information as possible regarding appropriate establishment of direct PE inputs for these services.

Comment: A commenter stated that the potential reductions to the technical component that we are avoiding would have been based on flawed methodology, particularly stating that the PE per hour values used in PE ratesetting methodology is inaccurate as it is based on the Physician Practice Expense Information Survey (PPIS) from 2007-2008, which the commenter considers to be flawed. The commenter also stated that the interest rate applied to high cost capital equipment such as imaging is inappropriately low, and that the equipment utilization rate assumption is inappropriately high.

Response: We note that the 90 percent equipment utilization rate only applies to diagnostic imaging services with equipment priced at $1 million dollars or more. The most recent recommended inputs for these services do not include imaging equipment priced at $1 million dollars or more, so the 90 percent equipment utilization would not apply. However, we would address any application of a different utilization rate through notice and comment rulemaking when valuing the codes under our standard PE methodology. As always, we welcome information about the validity of the assumptions we make in calculation of direct and indirect costs in terms of PE. We previously noted our interest in improving PE calculations through incorporation of alternative data sources and we continue to seek information from interested stakeholders as to the kinds of data sources that might be available.

For CY 2017, we are finalizing the proposed work RVUs and PE RVUs associated with CPT codes 77067, 77066 and 77065 for use with HCPCS codes G0202, G0204, and G0206, respectively.

(41) Radiation Treatment Devices (CPT Codes 77332, 77333, and 77334)

We identified CPT codes 77332, 77333, and 77334 through the high expenditures by specialty screen. These services represent an incremental increase of complexity from the simple to the intermediate to the complex in design of radiation treatment devices. The RUC recommended no change from the current work RVUs of 0.54 for CPT code 77332, 0.84 for CPT code 77333 and 1.24 for CPT code 77334. We believed the recommended work RVUs overstate the work involved in furnishing these services, as they do not sufficiently reflect the degree to which the RUC concurrently recommended a decrease in intraservice or total time. For CPT code 77332, we believed the RUC recommendation to maintain its current value despite a 34 percent decrease in total time appeared to ignore the change in time. Therefore, we proposed a value for this code based on a crosswalk from the value from CPT code 93287 (Peri-procedural device evaluation (in person) and programming of device system parameters before or after a surgery, procedure, or test with analysis, review and report by a physician or other qualified health care professional; single, dual, or multiple lead implantable defibrillator system), due to its identical intraservice time, similar total time, and similar level of intensity. We therefore proposed a work RVU of 0.45 for CPT code 77332. We further supported this valuation with CPT code 97760 (Orthotic(s) management and training (including assessment and fitting when not otherwise reported) upper extremity(s), lower extremity(s) and/or trunk, each 15 minutes), which has similar physician time and intensity measurements and a work RVU of 0.45. As these codes are designed to reflect an incremental increase in work value from simple, to intermediate, and complex device designs, we used an incremental difference methodology to value CPT codes 77333 and 77334. We proposed a work RVU of 0.75 for CPT code 77333, maintaining its recommended increment from CPT code 77332. For CPT code 77334, we proposed a work RVU of 1.15, which would maintain its increment from CPT code 77332.

Comment: Several commenters did not support CMS' use of CPT code 93287 as a crosswalk code to value CPT code 77332, as it is not a radiology service.

Response: We appreciate the commenters' concern about using a non-radiology service to assist in our valuation of this code family. We note that it is fundamental to the validity of the relative value system that codes furnished by different kinds of physicians remain valid relative to each other. We commonly value codes by use of crosswalks to other codes that are similar in terms of time and intensity, and this may extend across different mixes of specialties furnishing each service on the MPFS.

Comment: One commenter did not support CMS' pointing to the RUC's recommendation of a reduction of total time without a commensurate reduction in work RVU, as the current time is a CMS/Other source time, which is not derived from a survey and was assigned over 20 years ago.

Response: We utilize a variety of methodologies and approaches in developing work RVUs, and we believe that the total time value for this service is one of several appropriate criteria that can be used to estimate the overall time and intensity. We believe that the intraservice and total times listed for this service are valid elements in allowing us to determine an appropriate work RVU. Furthermore, we note that the current times assigned to this code have been used to allocate indirect PE to services furnished by the same specialties, and use of this value is consistent with code valuation methodology.

Comment: One commenter asked for clarification regarding if CMS is comparing the total time for CPT code 93287 to the current physician time of 77332 or to the survey time on which the RUC recommendation was based. One commenter stated that CMS' characterization of the intraservice time of crosswalk CPT code 93287 as identical to CPT code 77332 is incorrect; the intraservice time for 77332 is 15 minutes, and the intraservice time of CPT code 93287 is 13.5 minutes.

Response: We thank the commenter for bringing this to our attention; our previous statement that the intraservice time of CPT code 93287 is identical to the RUC-recommended intraservice time is incorrect. The RUC-recommended intraservice time of 15 minutes is similar, but not identical to the intraservice time of CPT code 93287 which is 13.5 minutes. We continue to believe that a work RVU of 0.45 is appropriate because we continue to believe the overall work for these services is approximately the same as 97760. As further support for our proposed value, we refer to 93016 (Cardiovascular stress test using maximal or submaximal treadmill or bicycle exercise, continuous electrocardiographic monitoring, and/or pharmacological stress; supervision only, without interpretation and report) which has an intraservice time that is identical to the RUC-recommended intraservice time for 77332, as well as a similar total time.

Comment: One commenter stated that these codes have XXX global periods, and therefore, do not have standard pre or post service packages. These standard pre and post services packages did not exist at the time that this service was valued, thus the convention of eliminating pre-service time and applying minimal post-service time to services with XXX global periods was not applied at that time.

Response: We appreciate the commenter's concerns about standard time packages not being applied to these codes. We continue to believe, however, that use of the RUC-recommended time to value the work RVU in this case is appropriate because we believe that time values are a critical element of establishing work RVUs.

Comment: A few commenters stated that CMS' proposed reduction in the work RVUs for CPT codes 77333 and 77334 based on an incremental relationship with CPT code 77332 is arbitrary, and that a reduction to the work RVU for CPT code 77332 does not automatically justify a reduction to the other two family codes.

A commenter supported the use of incremental valuation methodology in theory but did not believe it is appropriately applied to these codes, because the commenter believes that the valuation of CPT code 77332, upon which the increments are based, is incorrect.

Response: We refer readers to a discussion of the methodology for establishing work RVUs in section II.L.2 of this final rule. As outlined there, we frequently use an incremental methodology to identify potential work RVUs for particular codes. We note that we are maintaining the RUC-recommended incremental relationship between these three codes. This code family is structured to represent simple, intermediate, and complex procedures, and we seek to maintain that structure for this code family. Therefore, we are finalizing the work RVUs as proposed. We provide the information in Table 20 to illustrate our valuation of CPT code 77332 and its value relative to our crosswalk codes:

(42) Special Radiation Treatment (CPT Code 77470)

We identified CPT code 77470 through the high expenditures by specialty screen. We proposed the RUC-recommended work RVU of 2.03. However, we believe the description of service and vignette describe different and unrelated treatments being performed by the physician and clinical staff for a typical patient, and this presents a disparity between the work RVUs and PE RVUs. We solicited comment on information that would clarify this apparent disparity to help determine appropriate PE inputs. In addition, we solicited comment to determine if creating two G-codes, one that describes the work portion of this service, and one that describes the PE portion, may be a potentially more accurate method of valuing and paying for the service or services described by this code.

Comment: Some commenters maintained that the clinical labor and physician work component are related and are necessarily reported together. Commenters did not approve of CMS suggestion of breaking the work and PE components of this service into two separate G-codes in future rulemaking, stating that the CPT descriptor is accurate and represents the typical patient. Some commenters sought greater explanation for why CMS believes that the work and PE portions of this service are unrelated; commenters question if it is because the vignettes offered for the work and PE describe treatments for two separate diagnoses. Commenters also questioned if CMS is assuming that the “devices” mentioned in the description of clinical labor activities overlap with Radiation Treatment Devices codes which are also being evaluated in this rule. A commenter stated that if CMS is suggesting that there should be multiple CPT codes for every possible diagnosis for the use of this code, then that suggestion is problematic.

Response: According to the description of work provided for this service, the physician performs cognitive work such as planning, consideration of test results, and therapeutic treatment contingency planning that is in addition to what he or she would typically be performing for most radiation treatments. Meanwhile, the radiation therapist handles the treatment devices, performs tasks such as positioning the patient, and helps facilitate the scan of the patient. We believe that this may describe activities that are fundamentally disconnected. To illustrate our concern, we offer the example that this is akin to a physician removing a mole from a patient's hand while the clinical staff places a cast on the patient's foot; we see no compelling clinical evidence to indicate that the two tasks are related. In addition, the disparate diagnoses described by the vignettes further calls into question the degree to which the work and PE components are interrelated. While we agree that there should not separate coding for each possible diagnosis for a particular service, in trying to accurately assess relative value, we believe that the work and PE components should be valued under unified assumptions about the typical service. We are finalizing the RUC-recommended work RVU and PE inputs as proposed; however, we continue to have serious concerns about the validity of this coding.

(43) Interstitial Radiation Source Codes (CPT Codes 77778 and 77790)

In the CY 2016 PFS final rule with comment period, we established an interim final value for CPT code 77790 without a work RVU, consistent with the RUC's recommendation. We did not use the RUC-recommended work RVU to establish the interim final values for CPT code 77778. We stated that the specialty society survey included a work time that was significantly higher than the RUC-recommended work time without a commensurate change in the work RVU. For CY 2016, we established the 25th percentile work RVU survey result of 8.00 as interim final for CPT code 77778 and 0 work RVUs for CPT code 77790.

Comment on the CY 2016 PFS final rule with comment period: Commenters agreed that the preservice survey times and the RUC-recommended survey times were inconsistent and explained that this inconsistency resulted from the RUC's use of preservice packages in developing recommendations. In addition, commenters stated that because the work associated with CPT code 77790 (including pre-time supervision, handling, and loading of radiation seeds into needles) was bundled into CPT code 77778, that the additional work should be reflected in the RVU for CPT code 77778. Commenters encouraged us to accept the RUC-recommended work RVU of 8.78 and requested that CPT code 77778 be referred to the refinement panel.

Response in the CY 2017 PFS proposed rule: We did not refer CPT code 77778 to the CY 2016 multispecialty refinement panel because commenters did not provide new clinical information. We continued to believe that, based on the reduction in total work time, an RVU of 8.00 accurately reflected the work involved in furnishing CPT code 77778.

In the CY 2017 proposed rule, we proposed a work RVU of 8.00 for CPT code 77778 and 0 work RVUs for CPT code 77790. We also sought comment on whether we should use time values based on preservice packages if the recommended work value was based on time values that were significantly different than those ultimately recommended.

The following is a summary of the comments we received regarding our proposed valuations for CPT codes 77778 and 77790:

Comment: Some commenters stated that CMS underestimates the additional work inherent in furnishing CPT code 77778, considering that it is being bundled with CPT code 77790.

Commenters did not agree with our decision not to accept the RUC-recommended work RVU of 8.78 and to propose for CY 2017 a work RVU of 8.00, considering the disparity between the survey total time and the RUC-recommended total time. According to the RUC, the survey respondents had accurately estimated the work RVU based on magnitude estimation while overestimating the relatively low intensity pre-service time involved in performing this service, and this explains the disparity between the survey time and the RUC-recommended total time. One commenter noted that the RUC significantly reduced the pre-time because it did not include work in supervising the ordering of the isotope. Several commenters stated that CMS routinely accepts and uses pre-service time packages as recommended by the RUC.

Response: We continue to question how the same survey respondents that significantly overestimated the total time based on the RUC's analysis could nonetheless accurately estimate the overall work. We are also concerned about the specialty society's perspective that the RUC does not consider the work of supervising the ordering of the isotope as part of the service, given the survey respondents clearly considered such work to be described by the code. We believe that it is important that a particular code clearly describes the work involved in furnishing a service. While we appreciate the usefulness of pre-time packages generally, for this particular code, we believe that in this case the drastic time difference from the survey time value to the RUC- recommended time value that the pre-time package produces is problematic, especially since there does not appear to be consensus regarding which services are included in the code, or which might be perceived to be separately reportable.

In general we are concerned with using recommended time values that are disconnected from recommended work RVUs, including in cases where the recommended work RVU may include elements of work that are not reflected in the assumptions in time, as appears to be the case for this code. We reiterate that we believe the statute directs us to establish work RVUs that reflect the relative resource costs in time and intensity, so we believe that there should be an identifiable relationship between time and work RVUs.

To align the time and work associated with this code, we proposed a reduction of the work RVU from 8.78 to 8.00 as we proposed. However, upon consideration of comments, we were persuaded that the RUC-recommended work RVUs for this service are appropriate, particularly because the work includes the supervision, handling, and loading of radiation seeds, and it reflects the bundling with CPT code 77790.

While we are not finalizing a change in the time associated with this code since we proposed to use the RUC recommended value based on the pre-service package, we seek additional information regarding the best approach to valuing work when there is a clear disconnect between assumptions regarding time described by a code and the time recommended by the RUC. We understand that pre-service time packages can be a helpful tool in assigning estimates of time to particular codes relative to others on the PFS and that these times may be significantly different than those derived from survey results. However, since the RUC has repeatedly stated that its recommendations reflect the typical resources involved in furnishing PFS services, we believe it would be important for us to be able to identify cases where the recommended time values reflect the application of particular policies rather than the best estimate of the actual time involved in furnishing procedures.

(44) Colon Transit Imaging (78264, 78265, 78266)

In establishing CY 2016 interim final values, we accepted the RUC recommended work RVUs for CPT codes 78265 and 78266. We believed that the RUC-recommended RVU of 0.80 overestimated the work involved in furnishing CPT code 78264 and as a result, we established an interim final work RVU of 0.74 based on a crosswalk to CPT code 78226 (hepatobiliary system imaging, including gallbladder when present), due to similar intraservice times and intensities.

Comment on the CY 2016 PFS final rule with comment period: Commenters did not support our interim final work RVU for CPT code 78264. Commenters disagreed with our assessment of CPT code 78264 as having a higher work RVU and shorter intraservice time relative to the other codes in the family. One commenter stated that a difference of two minutes in intraservice time was insignificant and should not be used as a rationale for revaluing. Another commenter stated that we should have maintained the RUC-recommended crosswalk of CPT code 78264 to CPT code 78227 (Hepatobiliary system imaging, including gallbladder when present; with pharmacologic intervention, including quantitative measurement(s) when performed) due to similarities in service, work and intensity. Based on these concerns, commenters requested that CPT code 78264 be referred to the refinement panel.

Response in the CY 2017 PFS proposed rule: CPT code 78264 was referred to the CY 2016 multi-specialty refinement panel for further review. We calculated the refinement panel results as the median of each vote. That result for CPT code 78264 was 0.79 RVUs.

In the CY 2017 proposed rule, we proposed a value of 0.79 for CPT code 78264.

The following is a summary of the comments we received regarding our proposed valuation of the Colon Transit Imaging codes:

Comment: A commenter recommended that we reexamine the data associated with these codes to ensure the accuracy of the final values.

Response: We thank the commenter for this input. We continue to believe that the proposed valuation on CPT code 78264 most accurately describes the work, time and intensity associated with this service; therefore, we are finalizing the work RVU as proposed.

(45) Cytopathology Fluids, Washings or Brushings and Cytopathology Smears, Screening, and Interpretation (CPT Codes 88104, 88106, 88108, 88112, 88160, 88161, and 88162)

In the CY 2016 PFS final rule with comment period, we made a series of refinements to the recommended direct PE inputs for this family of codes. We removed the equipment time for the solvent recycling system (EP038) and the associated clinical labor described by the tasks “Recycle xylene from stainer” and “Order, restock, and distribute specimen containers and or slides with requisition forms” due to our belief that these were forms of indirect PE. This refinement applied to all seven codes in the family. We also noticed what appeared to be an error in the quantity of non-sterile gloves (SB022), impermeable staff gowns (SB027), and eye shields (SM016) assigned to CPT codes 88108 and 88112. The recommended value of these supplies was a quantity of 0.2, which we believed was intended to be a quantity of 2. We therefore refined the value of these supplies to 2 for CPT codes 88108 and 88112.

Comment on the CY 2016 PFS final rule with comment period: Several commenters disagreed with our characterization of the solvent recycling system and its associated clinical labor tasks as indirect PE. Commenters stated that the solvent recycling system costs are direct expenses since they are based on the amount of recycled solvent allocated to each specimen, with solvents allocated to specific specimens based on batch size. They indicated that the related clinical labor tasks are direct PE as they are also based on the amount of recycled solvent allocated to each specimen. The time for these tasks varies based on the batch size, which varies by procedure.

Response in the CY 2017 PFS proposed rule: We maintained our previously stated belief that these are forms of indirect PE, as they are not allocated to any individual service. Under the established PE methodology, direct PE inputs are defined as clinical labor, medical supplies, or medical equipment that are individually allocable to a particular patient for a particular service. We continue to believe that a solvent recycling system would be in general use for a lab practice, and that the associated clinical labor tasks for ordering and restocking specimen containers can be more accurately described as administrative activities. We proposed to maintain these refinements from the previous rulemaking cycle for CPT codes 88104-88162.

Comment on the CY 2016 PFS final rule with comment period: A commenter indicated that we did not account for the batch size when considering the supply quantities for CPT codes 88108 and 88112. The commenter indicated that the practice expense inputs should be assumed to have a batch size of five for these two codes, and therefore, no edits should be made. The commenter requested that we restore the quantity of 0.2 for the gloves, gowns, and eye shields associated with these procedures. This did not apply to the other codes on the submitted spreadsheet, which had a batch size of one.

Response in the CY 2017 PFS proposed rule: We appreciated the assistance of the commenter in clarifying the batch size for these procedures. As a result, we proposed to refine the supply quantity of the non-sterile gloves (SB022), impermeable staff gowns (SB027), and eye shields (SM016) back to the RUC-recommended value of 0.2 for CPT codes 88108 and 88112.

The following is a summary of the comments we received regarding our proposed valuation of the Cytopathology Fluids and Cytopathology Smears codes:

Comment: A few commenters continued to disagree that the proposed refinements to the direct PE inputs were forms of indirect PE. Commenters stated that these tasks are direct expenses, as they are variable based on the volume of these services, with the clinical labor and equipment time directly attributable to the quantity of specimens typically provided from a typical laboratory. Commenters also stated that these activities were not captured in the questions asked on the indirect practice expense cost survey.

Response: We continue to believe that these are administrative tasks that are more accurately classified as forms of indirect PE because they are not allocable to an individual service. Whether these tasks are variable based on the volume of the services is unrelated to this classification. For example, some services may require additional time for administrative staff to record electronic health records or restock inventory than other services, but in all cases these are defined as indirect PE under the established methodology, as they are administrative tasks that are not allocated to any individual service. We disagree that the validity of the practice expense data rests on whether or not particular questions were asked on the survey. We note that we understand medical practice and technology often change over time and the PE survey data is used to capture the relative difference in practice expenses incurred by various specialties as opposed to representing a summation of all individual items that incur an expense. Therefore, we do not believe that inclusion or exclusion of particular items means that the underlying data are invalid for purposes of measuring relativity.

Comment: A commenter agreed with the changes to the RUC-recommended supply quantity of 0.2 for the non-sterile gloves (SB022), impermeable staff gowns (SB027), and eye shields (SM016) in CPT codes 88108 and 88112.

Response: We appreciate the support from the commenter.

After consideration of comments, we are finalizing the proposed direct PE inputs for CPT codes 88104, 88106, 88108, 88112, 88160, 88161, and 88162.

(46) Flow Cytometry Interpretation (CPT Codes 88184, 88185, 88187, 88188, and 88189)

The Flow Cytometry Interpretation family of codes is split into a pair of codes used to describe the technical component of flow cytometry (CPT codes 88184 and 88185) that do not have a work component, and a trio of codes (CPT codes 88187, 88188, and 88189) that do not have direct practice expense inputs, as they are professional component only services. CPT codes 88184 and 88185 were reviewed by the RUC in April 2014, and their CMS refined values were included in the CY 2016 PFS final rule with comment period. The full family of codes was reviewed again at the January 2016 RUC meeting, and new recommendations were submitted to CMS as part of the CY 2017 PFS rulemaking cycle.

We proposed the RUC-recommended work RVU of 0.74 for CPT code 88187, and the RUC-recommended work RVU of 1.70 for CPT code 88189. For CPT code 88188, we proposed a work RVU of 1.20 instead of the RUC-recommended work RVU of 1.40. We arrived at this value by noticing that there were no comparable codes with no global period in the RUC database with intraservice time and total time of 30 minutes that had a work RVU higher than 1.20. The RUC-recommended work RVU of 1.40 would go beyond the current maximum value and establish a new high, which is not consistent with our estimation of the overall intensity of this service relative to the others. As a result, we believe it is more accurate to crosswalk CPT code 88188 to the work value of the code with the current highest value, which is CPT code 88120 (Cytopathology, in situ hybridization (for example, FISH), urinary tract specimen with morphometric analysis, 3-5 molecular probes) at a work RVU of 1.20. We believe that CPT code 88120 is crosswalk comparable code since it shares the identical intraservice time and total time of 30 minutes with CPT code 88188.

We also noted that the survey increment between CPT codes 88187 and 88188 at the RUC-recommended 25th percentile was 0.40 (between work RVUs of 1.00 and 1.40), and this increment of 0.40 when added to CPT code 88187's work RVU of 0.74 would arrive at a value of 1.14. In addition, the total time for CPT code 88188 decreases from 43 minutes to 30 minutes, which is a ratio of 0.70, and when this time ratio is multiplied by CPT code 88188's previous work value of 1.69, the result would be a new work RVU of 1.18. With this information in mind, we proposed a work RVU of 1.20 for CPT code 88188 as a result of a direct crosswalk to CPT code 88120.

For CPT codes 88184 and 88185, which describe the technical component of flow cytometry, we proposed to use the RUC-recommended inputs with a series of refinements. However, we believe that the coding for these two procedures may inhibit accurate valuation. CPT code 88184 describes the first marker for flow cytometry, while CPT code 88185 is an add-on code that describes each additional marker. We believe that it may be more accurate to have a single CPT code that describes the technical component of flow cytometry on a per patient case basis, as these two procedures are always performed together and it is difficult to determine the clinical labor, supplies, and equipment used in the typical case under the current coding structure. We solicited comments regarding the public interest in consolidating these two procedures into a single code used to describe the technical component of flow cytometry.

Absent such a change in coding, we proposed to refine the clinical labor time for “Instrument start-up, quality control functions, calibration, centrifugation, maintaining specimen tracking, logs and labeling” from 15 minutes to 13 minutes for CPT code 88184. We maintained that 13 minutes for this activity, which is the current time value, would be typical for the procedure, as CPT code 88182 also uses 13 minutes for the identical clinical labor task. We also proposed to refine the L054A clinical labor for “Load specimen into flow cytometer, run specimen, monitor data acquisition, and data modeling, and unload flow cytometer” from 10 minutes to 7 minutes using the same rationale, a comparison to CPT code 88182.

We proposed to maintain the clinical labor for “Print out histograms, assemble materials with paperwork to pathologists Review histograms and gating with pathologist” for CPT code 88184 at 2 minutes, as opposed to the RUC-recommended 5 minutes. A clinical labor time of 2 minutes is standard for this activity; we disagree with the RUC rationale that reviewing histograms and gating with the pathologist in this procedure is not similar to other codes. We also note that the review of histograms with a pathologist is not even described by CPT code 88184, which again refers to the technical component of flow cytometry, not the professional component. We also proposed to refine the L033A clinical labor time for “Clean room/equipment following procedure” from 2 minutes to 1 minute for CPT code 88184. We have established 1 minute in previous rulemaking (80 FR 70902) as the standard time for this clinical labor activity in the laboratory setting.

We proposed to maintain our removal of the clinical labor time for “Enter data into laboratory information system, multiparameter analyses and field data entry, complete quality assurance documentation” for both CPT code 88182 and CPT code 88184. As we stated in the CY 2016 PFS final rule with comment period (80 FR 70979), we have not recognized the laboratory information system as an equipment item that can be allocated to an individual service. We continue to believe that this is a form of indirect PE, and therefore, we do not recognize the laboratory information system as a direct PE input, and we do not consider this task as typically performed by clinical labor on a per-service basis.

We proposed to maintain the quantity of the “lysing reagent” supply (SL089) at 2 ml for CPT code 88185, as opposed to the RUC-recommended quantity of 3 ml. In our discussions with pathology specialists who perform flow cytometry, we were informed that the use of 50-55 ml of the lysing reagent would be typical for an entire patient case. The RUC recommendation similarly suggested a quantity of 46 ml or 48 ml per patient case. We were also told that the most typical number of markers used for flow cytometry is 24, consisting of 1 service of CPT code 88184 and 23 services of CPT code 88185. An investigation of our claims data confirmed this information, indicating that 24 markers is the most frequent per patient case for flow cytometry, and the use of more than 20 markers is typical. We believe that this data supports our refinement of the lysing reagent from a quantity of 3 ml to a quantity of 2 ml for CPT code 88185, which is also the current value for the procedure and the RUC-recommended value from the previous set of recommendations. For the typical case of 24 markers, our value would produce a total lysing reagent quantity of 51 ml (5 ml from the single service of CPT code 88184 and 46 ml from the 23 services of CPT code 88185), which matches with the amount required for a total per patient case. If we were to adopt the RUC recommendation, the total lysing reagent quantity would be 74 ml, which is well in excess of what we believe to be typical for these procedures.

We also proposed to refine the quantity of the “antibody, flow cytometry” supply (SL186) from quantity 1.6 to quantity 1, which is also the current value for the supply and the RUC-recommended value from the previous set of recommendations. We do not agree that more than one antibody would be typically used for each marker. We are reaffirming the previous RUC recommendation, and maintaining the current quantity of 1 antibody for each marker.

We did not agree with the recommended additional time for the “printer, dye sublimation (photo, color)” equipment (ED031). We proposed to maintain the equipment time at 2 minutes for CPT code 88184, and at 1 minute for CPT code 88185. As we stated in the CY 2016 PFS final rule with comment period (80 FR 70979), we proposed to assign equipment time for the dye sublimation printer to match the clinical labor time for “Print out histograms, assemble materials with paperwork to pathologists.” We do not believe that it would be typical for the printer to be in use longer than it takes to accomplish this clinical labor task.

The following is a summary of the comments we received regarding our proposed valuation of the Flow Cytometry Interpretation codes. Due to the large number of comments we received for this code family, we will first summarize the comments related to the coding structure of CPT codes 88184 and 88185, followed by the comments related to specific work RVUs, and finally the comments related to the direct PE inputs.

Comment: Many commenters disagreed with the potential concept of consolidating CPT codes 88184 and 88185 into a single code used to describe the technical component of flow cytometry. Commenters stated that the resources required for the first marker and for each subsequent marker differ, and with flow cytometry, there is no “typical case.” Because the number of markers differ for different disease states, such as HIV, Lyme disease, and acute leukemias, the current coding structure is designed to reflect different valuations of the professional component codes, based on the number of markers that must be interpreted. Many commenters stressed that this makes one code for the technical component of flow cytometry infeasible, and strongly advised against it. One commenter was also concerned that a coding structure change may exacerbate the undervaluation of these services, which have been recently reviewed twice by the RUC and resulted in substantial decreases in the practice expense relative values.

A few commenters supported the possibility of combining CPT codes 88184 and 88185 into a single code. One commenter stated that the current coding structure does not incentivize the use of less reagents, and actually penalizes labs that appropriately test fewer markers. According to this commenter, moving to a single code structure would be consistent with the vast majority of lab tests, would simplify billing processes, and may make development of more cost-effective panels financially desirable. The commenter supported further examination of a single CPT code and urged that current payment rates should be frozen while such examination occurs. Another commenter suggested a slightly different coding structure, one which would collapse the codes into a series of case rate codes that reflect the procedures: screening, classification, and monitoring. There was support from one additional commenter for a three code proposal designed to track this workflow.

Response: We appreciate the detailed responses from the commenters about the proper coding structure used to describe the technical component of flow cytometry. We do not intend to finalize any recommendations regarding the coding structure at this time, but we will consider this information for future proposals regarding these services.

Comment: Many commenters made general comments about decreases to the proposed rates for either the professional or the technical component of the flow cytometry codes. Commenters stated that there was no justification for the reduction in payment rates, and that the decreases would hamper laboratories' ability to offer the flow cytometry services. One commenter stated that the payment cuts were not realistic and would result in flow cytometry not being financially feasible in the less expensive physician-office setting. Another commenter indicated that further reductions to these codes would result in an inability to maintain the level of professional services required to reduce medical errors.

Response: We share the concern of the commenters in ensuring that payment for Medicare services is based on an accurate assessment of the relative resource costs involved in furnishing the service. With regards to the technical component of flow cytometry, most of the decrease in code valuation is taking place due to a decrease in the quantity of the lysing reagent supply (SL089). The RUC has agreed that there was previously an excess of this supply in CPT codes 88184-88185, and has recommended a decrease of approximately 78 percent in this supply quantity, from 336 ml to 74 ml, in the typical case of 24 markers. Due to the resource-based nature of the RVU system, this substantial reduction in supply costs will be reflected in the RVUs for these procedures. We note that since CY 2016 the phase-in of significant reductions in RVUs has been in effect; if the total RVUs for a service for a year would otherwise be decreased by an estimated 20 percent or more as compared to the total RVUs for the previous year, those decreases are limited to a 19 percent reduction in total RVUs. We note that the phase-in mechanism allows reductions to be transitioned in over time rather than instituting large decreases in a single rule cycle. Please see section II.H for more information regarding the phase-in of significant RVU reductions.

Comment: Many commenters disagreed with the proposed work RVU of 1.20 for CPT code 88188. Several commenters took issue with the CMS statement that there were no comparable codes with no global period in the RUC database with intra-service time and total time of 30 minutes that had a work RVU higher than 1.20. These commenters indicated that there were at least 10 such codes valued over 1.20 RVUs in the 2016 RUC database (1 XXX and 9 ZZZ add-on global codes), ranging in work value from 1.38 to 2.40 RVUs, with a median of 1.67. The commenters suggested that these codes supported the higher RUC-recommended work RVU of 1.40 for CPT code 88188.

Response: We continue to believe that there are no comparable codes with the same global period with intraservice time and total time of 30 minutes that have a work RVU higher than 30 minutes. When we used the phrase “no global period” to refer to CPT code 88188, we were not referring to add-on codes with a global period of ZZZ. We have stated on numerous occasions that we believe the resources required to furnish add-on codes constitute a separate category, and we typically only compare add-on codes to other add-on codes. We do not believe that it is appropriate to compare the work RVU of add-on codes with 30 minutes of intraservice time to the work RVU of CPT code 88188, which is not an add-on code.

With regards to non-add on codes, Table 21 lists all 13 codes in the RUC database with 30 minutes of intraservice time, fewer than 40 minutes of total time, and a global period of XXX:

As we stated previously, there are no codes with a work RVU higher than 1.20, which is where we proposed to value CPT code 88188. We acknowledge that there are global XXX codes with 30 minutes of intraservice time that have a work RVU greater than 1.20. However, all of these codes have at least 40 minutes of total time, which is 33 percent higher at a minimum than the total time for CPT code 88188. We believe that a crosswalk to CPT code 88120, which shares the identical time values as CPT 88188, is a more appropriate choice than codes that have substantially higher total time. In the particular case of CPT code 88188, we continue to believe that establishing a new maximum work value above 1.20 would not be consistent with our estimation of the overall intensity of this service relative to the others on the PFS.

Comment: Some commenters disagreed with the proposed work RVU of CPT code 88188 based on the work increments between the codes in the family. These commenters stated that the original recommended work values had almost identical increments between the three services (0.60 between CPT codes 88187 and 88188, and 0.63 between CPT codes 88188 and 88189); however the median survey results indicated a much greater physician work increment between CPT codes 88188 and 88189. According to commenters, the final RUC recommendations were based on the expertise of the RUC to establish the work increment between CPT codes 88187 and 88188 (0.74) higher than the increment between CPT codes 88188 and 88189 (0.30). In other words, the recommended work increment between CPT code 88187 (work RVU = 0.74) and CPT code 88188 (work RVU = 1.40) was significantly larger than the work increment between CPT code 88188 (work RVU = 1.40) and CPT code 88189 (work RVU = 1.70). The commenters stated that the survey results and expert opinion justified this smaller increment between the final two codes, and the RUC-recommended work RVU of 1.40 for CPT code 88188.

Response: We disagree that the survey data justifies a smaller increment between the final two codes. While this is true for the 25th percentile survey results, the exact opposite is true for the survey median results, in which the increment between CPT codes 88187 and 88188 is 0.35 and the increment between CPT codes 88188 and 88189 is 0.70. In addition, in the current pre-reviewed version of these codes, the increment between CPT codes 88187 and 88188 is 0.33, while the increment between CPT codes 88188 and 88189 is 0.54. We believe that this suggests the survey data on the work increments is conflicting, not conclusive, and that the RUC-recommended increments are a departure from the previous incremental structure of this code family, in which the second two codes had a larger increment than the first two codes. We do not agree that the work increments at the survey 25th percentile are a sufficient justification for adopting the recommended work RVU for CPT code 88188 due to the additional data regarding work increments between these codes detailed above.

Comment: Several commenters stated that over the last decade, flow cytometric analyses have changed through new technological advances that have led to an increased interpretative sophistication. It is now typical for the physician to analyze substantially more data than in the past. According to commenters, with the advent of 5, 6, 8, and 10 color flow cytometry the intensity and complexity of these services has significantly increased. Commenters stated that this increased intensity and complexity is reflected in the RUC recommendation for this service, based on new physician work associated with technological changes, time, and intensity.

Response: We appreciate this additional information about the professional interpretation of flow cytometry from the commenters. However, we note that the RUC-recommended intensity of CPT codes 88187 and 88189 has actually decreased compared to the current pre-reviewed version of these codes. We believe that this indicates that the same new technological advances also allow practitioners to analyze data faster and with fewer errors, which is reflected in the decreased work RVUs and time values in the RUC recommendations. The only one of the three codes with a RUC-recommended increase in intensity is CPT code 88188. This increased intensity in the second code creates an anomalous relationship within the family, as the RUC-recommended intensity for CPT code 88188 is equal to the intensity for CPT code 88189, in contrast to the current pre-reviewed version of these codes where the three codes have a linear increase in intensity (IWPUT = 0.39, 0.43, 0.50). We do not understand why the professional interpretation of 9 to 15 markers would have an equal intensity to interpreting 16 or more markers. Logic would suggest that CPT code 88188 should have a lower intensity than CPT code 88189, which is indeed the case at our proposed work RVU of 1.20. The proposed value also re-establishes a linear increase in intensity between the three codes as additional markers are interpreted (IWPUT = 0.37, 0.40, 0.47). We believe that this intensity data offers additional support for our proposed work RVU.

Comment: One commenter disagreed with the CMS crosswalk to the work RVU of CPT code 88120, which the commenter suggested was completely different in step by step work effort, intensity, and complexity. The commenter stated that CPT code 88120 typically only involves identifying and quantifying a limited subset molecular probes (for example, FISH probes for chromosomes 3, 7, 17 and 9p21 loss), using two to four color signal enumeration to detect aneuploidy staining of nuclei on slides from isolated cell preparations, usually from morphologically well-characterized specimens. In contrast, the commenter stated that for CPT code 88188 the pathologist is required to integrate multi-parameter diagnostic information on different cell populations (both abnormal and normal), by assessing cell scatter (size and shape) along with signal intensity and pattern of staining of cell surface markers with antibody reagents using four to six (or more) color fluorescent antibody probes. The pathologist must perform successive, iterative analyses of 2- and 3-dimensional plots and histograms and re-gating of identified cell populations (based on size, shape, relative staining patterns, signal intensity, etc.) to characterize cell lineage and render a final diagnosis and interpretation. Due to this clinical rationale, the commenter indicated that the work and complexity of CPT code 88188 was substantially greater than CPT code 88120.

Response: We disagree with the commenter that CPT code 88120 is an inappropriate crosswalk code for CPT code 88188. These codes are both recently-reviewed pathology codes with identical intraservice time and total time values within the Cytopathology listing of the CPT manual. We also note that many of the activities listed by the commenter are not detailed in the intraservice work description for CPT code 88188, and may not be needed in the typical case.

The following comments address the proposed direct PE inputs for the Flow Cytometry family of codes.

Comment: Many commenters disagreed with the proposed time of 13 minutes for the clinical labor activity “Instrument start-up, quality control functions, calibration, centrifugation, maintaining specimen tracking, logs and labeling.” Commenters stated that the CMS comparison to CPT code 88182 was not appropriate, as that code uses older/simpler technology, with CPT code 88184 using 4-6 or more color channels while CPT code 88182 uses 1-2 channels. Commenters stressed that these clinical labor tasks are unique to this flow cytometry service, and they should not be assumed to take the identical time as other services. Other commenters stated that three instruments must be run consecutively, and the task includes quality control calibration, taking a minimum of 13 to 16 minutes in dedicated technical staff time. Another commenter indicated that the time required to complete these activities is continually increasing as more regulatory requirements are added, and that the recently added flow cytometry requirement for individual antibody lot/shipment testing increased this time exponentially.

Response: We disagree with the commenters that the identical clinical labor activity would take longer to perform for CPT code 88184 than it would for CPT code 88182. Both of these procedures use the same equipment to perform this task, a flow cytometer (EP014) and a centrifuge (EP007). We do not agree that there is additional clinical labor time required for using additional color channels in CPT code 88184, as the same equipment is being used to perform the same clinical labor task as in CPT code 88182. We did not receive data from the commenters suggesting that regulatory requirements are increasing the time required to perform this clinical labor task, nor was this reflected in the RUC recommendations, which continued to recommend the same unchanged time for this task.

Comment: Many commenters objected to the proposed clinical labor time of 7 minutes for the “Load specimen into flow cytometer, run specimen, monitor data acquisition, and data modeling, and unload flow cytometer” activity for CPT code 88184. Commenters stated that the CMS comparison to the clinical labor time used for this same activity in CPT code 88182 was not appropriate as CPT code 88184 uses 4-6 color channel instruments and up, while 88182 uses only 1-2 channels. According to the commenters, the time it takes for data capture, data modeling, data acquisition, and computational analysis is exponentially longer for CPT code 88184 than for CPT code 88182, since additional colors result in more complicated profiles which are more difficult and time consuming to evaluate. Another commenter stated that 7 minutes was wholly inadequate to perform all of these tasks, and that analysis of a specimen can take 12 to 15 minutes, depending on the complexity of the case.

Response: We continue to disagree with the commenters that the identical clinical labor activity would take longer to perform for CPT code 88184 than it would for CPT code 88182. As we stated in response to the previous comment, we do not agree that there is additional clinical labor time required for using additional color channels in CPT code 88184, as the same equipment is being used to perform the same clinical labor task as in CPT code 88182. For the same reason, we do not agree that this clinical labor activity takes 12 to 15 minutes to perform, since the identical task only requires 7 minutes for CPT code 88182.

Comment: Many commenters opposed the proposed value of 2 minutes for the clinical labor activity “Print out histograms, assemble materials with paperwork to pathologists Review histograms and gating with pathologist” for CPT code 88184. Commenters stated that it was not reasonable to expect that a flow cytometry technologist could print out histograms, assemble the documents and deliver them to a pathologist, and review the histograms with a pathologist, all in the span of a mere 120 seconds. Commenters were concerned that flow cytometry technologists cannot produce a high-quality product and ensure its accuracy and completeness for presentation to a pathologist in the proposed time. One commenter noted that although their specific procedure for these steps was largely electronic, their workflow analysis corroborated the RUC's conclusion because it showed that it took 5 minutes for staff to complete the equivalent activities. Several other commenters stated that if the time the cyotechnologist takes to determine exactly which histograms to print is subtracted, then they could agree with the proposed 2 minutes. Commenters also stated that printing is not performed all at one time, with 25-30 pages of information and data printed over a 5 minute time span, and one commenter indicated that the process was “largely electronic” with clinical staff not using the equipment for the full duration that it is in use.

Response: We appreciate the support from several of the commenters. In responding to the comments for this clinical labor activity and the equipment time for the dye sublimation printer (ED031), it became clear that the clinical labor time for printing was not the same as the equipment time that the printer was in use. Based on the information from the commenters that printing is not performed all at one time, we are assigning the full 5 minutes of equipment time for the dye sublimation printer; however, we are maintaining our proposed 2 minutes of clinical labor time for “Print out histograms, assemble materials with paperwork to pathologists Review histograms and gating with pathologist”, as commenters have informed us that the clinical staff do not use the equipment for the full duration that it is in use.

Comment: Several commenters disagreed with the proposed clinical labor time of 1 minute for “Clean room/equipment following procedure” for CPT code 88184. The commenters stated that this time is allocated over entire patient case, and that it is typical and critical to clean the equipment between patient cases. The commenters also supplied details about the cleaning process, regarding how the laboratory technician cleans the equipment and workspace by decontaminating the equipment and work bench surfaces, as well as carrying out waste management after the procedure.

Response: We appreciate the additional information from the commenters regarding the cleaning of the room. However, the commenters did not provide a rationale as to why CPT code 88184 requires additional clinical labor above the standard value of 1 minute for room cleaning in lab procedures. We continue to believe that the standard clinical labor time is the most accurate valuation for this clinical labor task.

Comment: Many commenters requested that CMS restore the clinical labor time for the “Enter data into laboratory information system, multiparameter analyses and field data entry, complete quality assurance documentation” activity. Commenters stated that this data entry is manually entered and must be performed for each individual patient case. Several commenters indicated that entering test-specific data takes between five and ten minutes, and entry of client information and demographics and specimen information takes additional time that cannot be short-changed. Commenters emphasized that these are extremely important tasks that require technical skills, and assigning zero minutes to this critical task was illogical for a service like flow cytometry. One commenter stated that the current RUC-recommended value of four minutes was already a gross underestimation of the time required to complete these activities for the majority of testing, and suggested that these activities commonly take more than ten minutes to perform.

Response: We agree with the commenters that entering patient data into information systems is an important task, and we agree that it would take more than zero minutes to perform. However, the commenters did not address our rationale for removing this clinical labor time from CPT codes 88184 and 88185, which is that this task is indirect PE. As we stated in the CY 2016 final rule with comment period (80 FR 70979), we have not recognized the laboratory information system as an equipment item that can be allocated to an individual service. We continue to believe that this is indirect PE, and therefore, we do not recognize the laboratory information system as a direct PE input, and we do not consider this task as typically performed by clinical labor on a per-service basis.

Comment: One commenter requested the inclusion of additional cytotechnologist time of 10 minutes for CPT code 88184 and 2 minutes for CPT code 88185, as well as an additional desktop computer with monitor (ED021) equipment times of 10 minutes for 88184 and 2 minutes for 88185. This additional time was intended to reflect the time spent using the flow cytometry analytics software (EQ380).

Response: We agree with the RUC recommendations that the clinical labor and equipment time associated with the flow cytometry analytics software is already accounted for in the recommended clinical labor inputs. As the recommendations indicate, this time is included as part of the clinical labor activities “Accession specimen”, “Instrument start-up, quality control functions”, “Load specimen into flow cytometer, run specimen” and “Print out histograms, assemble materials with paperwork to pathologists.”

Comment: Many commenters disagreed with the proposed supply quantity of 2 for the lysing reagent (SL089) in CPT code 88185. Commenters stated that although they acknowledged that the current Medicare data showed that a patient case of 24 markers is typical, this result ignored other relevant pieces of information. The commenters indicated that an analysis of the 2014 Medicare 5% Sample Carrier Database showed that over 50 percent of individual providers typically bill fewer than 20 markers per patient case, and that since these providers are generally smaller and see fewer annual cases, the proposed supply quantity of 2 would potentially drive these providers to consider ceasing their flow cytometry services. The commenter also stated that these codes are often billed as part of either the Hospital IPPS or OPPS, which should be factored into the typical number of markers billed per case. The commenter also stated that the most common professional component of flow cytometry, CPT code 88189, would be associated with patient cases that bill for fewer than 24 markers, from 16 to 24.

Response: We reiterate that we establish payment rates based on the typical case, which the commenters agreed was 24 total markers. We have historically established payment rates based on the typical service and do not believe that it would be appropriate or serve the purpose of relativity to deviate from that practice in this case. We also do not believe that the payment for these codes under the IPPS or OPPS is a directly relevant factor in defining the typical case under the Physician Fee Schedule. We believe that the patient population and typical case under the IPPS would not necessarily be the same as the typical case under the PFS. Finally, we agree that CPT code 88189 would be associated with patient cases that bill for fewer than 24 markers, as the code descriptor states that it refers to the performance of 16 or more markers. However, we do not believe that this affects the number of markers in the typical case, which the commenters agreed was 24 for the typical patient.

Comment: A commenter stated that that it opposed putting a number or cap on markers because there is a wide range of possible markers required to achieve patient diagnosis.

Response: We agree with the commenter, and we are not establishing a cap or determining a fixed number of markers to use for these procedures. As stated previously, however, we are required to establish payment rates based on the typical case, which our internal data and commenter feedback has agreed is 24 markers.

Comment: Other commenters disagreed with the CMS proposal for the lysing reagent based on the supply quantity needed to perform the procedure. A commenter stated that the 46-48 mL quantity detailed by CMS in the proposed rule was based on a RUC recommendation; however, the RUC's amount was based on an average of 16 markers, not 24 markers. Although the commenter agreed that 24 markers reflected a common case, the commenter stated that it was necessary to consider the amount of lysing agent for a 24 marker case, not to assume that the 46-48 mL amount based upon 16 markers also applies to 24 markers. Another commenter stated that a laboratory using ammonia chloride needs at least 2.5 ml of lysing reagent for each time that CPT code 88185 is performed.

Response: We did not base our proposal for this supply quantity upon the RUC recommendation. As we stated in the proposed rule, we were informed that the use of 50-55 ml of the lysing reagent would be typical for an entire patient case based on our discussions with pathology specialists who perform flow cytometry. For the typical case of 24 markers, our value would produce a total lysing reagent quantity of 51 ml (5 ml from the single service of CPT code 88184 and 46 ml from the 23 services of CPT code 88185), which matches with the amount required for a total per patient case. Since commenters agreed that 24 markers was the typical patient case, we continue to believe that our proposed quantity of 2 ml is the most accurate value for CPT code 88185.

Comment: Many commenters objected to the proposed supply quantity of 1 for the flow cytometry antibody (SL186) in CPT codes 88184 and 88185. Commenters stated that although it is standard practice to use a single antibody multiple times during the analysis, each antibody or marker can only be billed once per analysis. According to commenters, multiple use of such antibodies are not reportable or billable, but are critical to the overall analysis and interpretation of results and are part of the total cost for each procedure performed. Some commenters explained that the recommended quantity of 1.6 antibodies per billed marker was based on averaging together two separate analyses: a survey of 59 professionals performing flow cytometry that found 1.52 antibodies required per marker, and a customer survey that found 1.87 antibodies per marker. A different commenter stated that its member laboratories found that under the current four-color process, 1.36 antibodies per marker is necessary. Another commenter stated that while one antibody is generally used per marker, the required use of controls for many of these markers for analysis or quality control means that this value is greater than 1 antibody per marker reported.

Response: We appreciate the additional data presented regarding the clinical use of the flow cytometry antibody supply. However, we continue to have reservations regarding the information that we have received regarding the 1.6 quantity for this supply. Different commenters recommended different quantities of this supply required to furnish the procedure, ranging from 1 to 1.36 to 1.52 to 1.6 to 1.87. We are hesitant to increase the quantity of this supply given the wide-ranging information that we received from commenters. We are also concerned that although commenters referenced studies that found different supply quantities for SL186, commenters did not submit the data associated with these studies for our review. We would be more open to the idea of increasing the supply quantity to 1.6 if this data were supported by clinical data or study. We also note that one commenter stated that one antibody is “generally used” per marker, which supports our contention that the proposed value of 1 antibody for CPT codes 88184 and 88185 would be typical. As a result, we are maintaining a supply quantity of 1 for the flow cytometry antibody supply, which is also the current value for the supply and the RUC-recommended value from the previous set of recommendations.

Comment: Several commenters disagreed with the proposed equipment time for the dye sublimation printer (ED031). Commenters stated that printing is not performed all at one time, with 25-30 pages of information and data printed over a 5 minute time span. Commenters indicated that this time cannot be linked directly to one particular clinical labor task line, and the printer cannot be used for any other task during these 5 minutes even while it is not actively printing.

Response: We appreciated the additional information from the commenters regarding the use of the dye sublimation printer. Due to the presentation of this new information detailing how the equipment time for the printer is disassociated from any clinical labor tasks, we will increase the equipment time to the RUC-recommended 5 minutes for CPT code 88184 and 2 minutes for CPT code 88185.

After consideration of comments received, we are finalizing the proposed work RVUs for CPT code 88187, 88188, and 88189. We are also finalizing the proposed direct PE inputs, with the refinement to the dye sublimation printer detailed above.

(47) Microslide Consultation (CPT Codes 88321, 88323, and 88325)

CPT codes 88321, 88323, and 88325 were reviewed by the RUC in April 2014 for their direct PE inputs only, and the CMS refined values were included in the CY 2016 PFS final rule with comment period. The family of codes was reviewed again at the January 2016 RUC meeting for both work values and direct PE inputs, and new recommendations were submitted to CMS as part of the CY 2017 PFS rulemaking cycle.

In the CY 2016 PFS final rule with comment period, we finalized our proposal to remove many of the inputs for clinical labor, supplies, and equipment for CPT code 88325. The descriptor for this code did not state that slide preparation was taking place, and therefore, we refined the labor, supplies, and equipment inputs to align with the inputs recommended for CPT code 88321, which also does not include the preparation of slides. After further discussion with pathologists and consideration of comments received, we have been persuaded that slide preparation does take place in conjunction with the service described by CPT code 88325. In the RUC-recommended direct PE inputs from the January 2016 meeting, the labor, supplies, and equipment inputs related to slide preparation were added once again to CPT code 88325. We proposed to accept these restorations related to slide preparation without refinement.

Regarding the clinical labor direct PE inputs, we proposed to assign 1 minute of L037B clinical labor for “Complete workload recording logs. Collate slides and paperwork. Deliver to pathologist” for CPT codes 88323 and 88325. We are maintaining this at the current value for CPT code 88323, and adding this 1 minute to CPT code 88325 based on our new understanding that slide preparation is undertaken as part of the service described by this code. We proposed to remove the clinical labor for “Assemble and deliver slides with paperwork to pathologists” from all three codes, as we believe this clinical labor is redundant with the labor assigned for “Complete workload recording logs.” We similarly proposed to remove the clinical labor for “Clean equipment while performing service” from CPT codes 88323 and 88325, as we believe it to be redundant with the clinical labor assigned for “Clean room/equipment following procedure.”

We proposed to maintain the quantity of the “stain, hematoxylin” supply (SL135) at 16 ml for CPT codes 88323 and 88325, as opposed to the RUC-recommended quantity of 32 ml. The RUC recommendation stated that the hematoxylin supply does not include eosin and should not be redundant; the stains are not mixed together, but are instead sequential. The recommendation also made a comparison to the use of the hematoxylin supply quantity in CPT code 88305. However, we note that CPT code 88305 does not include 8 ml of eosin stain (SL201), but instead 8 gm of eosin solution (SL063), and these are not the same supply. Therefore we do not agree that a direct comparison of the supply quantities is the most accurate way to value these procedures. For CPT codes 88323 and 88325, we continue to note that the prior supply inputs for these procedures had quantity 2.4 of the eosin solution (SL063) and quantity 4.8 of the hematoxylin stain (SL135); in other words, a 1:2 ratio between the eosin and hematoxylin. We proposed to maintain that 1:2 ratio with 8 ml of the eosin stain (SL201) and 16 ml of the hematoxylin stain (SL135).

We also proposed to update the use of the eosin solution (sometimes listed as “eosin y”) in our supply database. We believe that the eosin solution supply (SL063), which is measured in grams, reflects an older process of creating eosin stains by hand. This is in contrast to the eosin stain supply (SL201), which is measured in milliliters, and can be ordered in a state that is ready for staining immediately. We do not believe that the use of eosin solution would reflect typical lab practice today, with the readily availability for purchase of inexpensive eosin staining materials. We also note that in the CY 2016 PFS final rule with comment period, we removed 8 gm of the eosin solution and replaced it with 8 ml of the eosin stain, and this substitution was accepted without further change in the most recent set of RUC recommendations. As a result, we proposed to update the price of the eosin stain supply from $0.044 per ml to $0.068 per ml to reflect the current cost of the supply. We also proposed to use CPT codes 88323 and 88325 as a model, and replace the use of eosin solution with an equal quantity of eosin stain for the rest of the codes that make use of this supply. This applies to 15 other CPT codes: 88302 (Level II—Surgical pathology, gross and microscopic examination), 88304 (Level III—Surgical pathology, gross and microscopic examination), 88305 (Level IV—Surgical pathology, gross and microscopic examination), 88307 (Level V—Surgical pathology, gross and microscopic examination), 88309 (Level VI—Surgical pathology, gross and microscopic examination), 88364 (In situ hybridization (e.g., FISH), per specimen; each additional single probe stain procedure), 88365 (In situ hybridization (e.g., FISH), per specimen; initial single probe stain procedure), 88366 (In situ hybridization (e.g., FISH), per specimen; each multiplex probe stain procedure), 88367 (Morphometric analysis, in situ hybridization (quantitative or semi-quantitative), using computer-assisted technology, per specimen; initial single probe stain procedure), 88368 (Morphometric analysis, in situ hybridization (quantitative or semi-quantitative), manual, per specimen; initial single probe stain procedure), 88369 (Morphometric analysis, in situ hybridization (quantitative or semi-quantitative), manual, per specimen; each additional single probe stain procedure), 88373 (Morphometric analysis, in situ hybridization (quantitative or semi-quantitative), using computer-assisted technology, per specimen; each additional single probe stain procedure), 88374 (Morphometric analysis, in situ hybridization (quantitative or semi-quantitative), using computer-assisted technology, per specimen; each multiplex probe stain procedure), 88377 (Morphometric analysis, in situ hybridization (quantitative or semi-quantitative), manual, per specimen; each multiplex probe stain procedure), and G0416 (Surgical pathology, gross and microscopic examinations, for prostate needle biopsy, any method).

The following is a summary of the comments we received regarding our proposed valuation of the Microslide Consultation codes.

Comment: One commenter supported the restoration of the direct PE inputs related to slide preparation in CPT code 88325 and requested that CMS update the PE data files for CY 2016 to reflect these changes.

Response: We appreciate the support from the commenter. The proposed rates for CY 2017 reflected these changes to the direct PE inputs. However, the RVUs for CY 2016 were unaffected by this proposal, as has been our longstanding practice for interim final codes.

Comment: Several commenters requested that CMS add an additional 1 minute for the clinical labor activity “Complete workload recording logs. Collate slides and paperwork. Deliver to pathologist” in CPT code 88321. Commenters stated that this clinical labor task was accidently left off of the April 2014 RUC recommendation for CPT code 88321, and that it was a necessary task that was not redundant with other clinical labor activities.

Response: We agree with the commenters that 1 minute of clinical labor time for this task is an appropriate addition for CPT 88321 to be consistent with the identical clinical labor task taking place in other codes in the family.

After consideration of comments received, we are finalizing our proposed work RVUs for CPT code 88321, 88323, and 88325. We are also finalizing the proposed direct PE inputs, with the addition of 1 minute of clinical labor time as detailed above for CPT code 88321. We note as well that we are finalizing the replacement of eosin solution with eosin stain, as detailed in the PE section of this final rule (see section II.A. of this final rule).

(48) Immunohistochemistry (CPT Codes 88341, 88342, 88344, and 88350)

In the CY 2014 PFS final rule with comment period (78 FR 74341), we assigned a status indicator of I (Not valid for Medicare purposes) to CPT codes 88342 and 88343 and instead created two G-codes, G0461 and G0462, to report immunohistochemistry services. We did this, in part, to avoid creating incentives for overutilization.

For CY 2015, the CPT coding was revised with the creation of two new CPT codes, 88341 and 88344, the revision of CPT code 88342 and the deletion of CPT code 88343. In the past for similar procedures in this family, the RUC recommended a work RVU for the add-on code (CPT code 88364) that was 60 percent of the work RVU for the base code (CPT code 88365). In the CY 2015 PFS final rule with comment period, we stated that the relative resources involved in furnishing an add-on service in this family would be reflected appropriately using the same 60 percent metric and subsequently established an interim final work RVU of 0.42 for CPT code 88341, which was 60 percent of the work RVU of 0.70 for the base CPT code 88342. In the CY 2016 PFS proposed rule, we revised the add-on codes from 60 percent to 76 percent of the base code and subsequently proposed a work RVU of 0.53 for CPT code 88341. However, we inadvertently published work RVUs for CPT code 88341 in Addendum B on the CMS Web site without explicitly discussing it in the preamble text. In the CY 2016 PFS final rule with comment period, we maintained the CY 2015 work RVU of 0.53 for CPT code 88341 as interim final for CY 2016 and requested public comment. Also, in the CY 2016 PFS final rule with comment period, we established an interim final work RVU of 0.70 for CPT codes 88342 and 88344.

Comment on the CY 2016 PFS final rule with comment period: Several commenters objected to a standard discount for the physician work involved in pathology add-on services and urged us to accept the RUC-recommend work RVU of 0.65 for CPT code 88341.

Response in the CY 2017 PFS proposed rule: We responded to the comments by stating our appreciation of the commenters' concerns regarding a standard discount; however, we believed that it was reasonable to estimate work RVUs for a base and an add-on code, and to recognize efficiencies between them, by looking at how similar efficiencies are reflected in work RVUs for other PFS services. Also we noted that the intravascular codes for which we initially established our base/add-on code relationship for CPT codes 88346 and 88350 were deleted in CY 2016 and replaced with two new codes; CPT codes 37252 and 37253. The relationship between CPT codes 37252 and 37253 represents a 20 percent discount for the add-on code as the base CPT code 37252 has a work RVU of 1.80 and CPT code 37253 has a work RVU of 1.44. As CPT codes 37252 and 37253 replaced the codes on which our discounts for base and add-on codes were based (please see the CY 2016 PFS final rule with comment period (80 FR 70972) for a detailed discussion), we believed it would be appropriate to maintain the same 20 percent relationship for CPT codes 88346 and 88350. Therefore, for CY 2017, we proposed a work RVU of 0.56 for CPT code 88341, which represents 80 percent of the work RVU of 0.70 for the base code. For CY 2016, we finalized a work RVU of 0.56 for CPT code 88350 which, represented 76 percent of the work RVU of 0.74 for the base code. To maintain consistency within this code family, for CY 2017 we proposed to revalue CPT code 88350 using the 20 percent discount discussed above. To value CPT code 88350, we multiplied the work RVU of 0.74 for CPT code 88346 by 80 percent, and then subtracted the product from 0.74, resulting in a work RVU of 0.59 for CPT code 88350. For CY 2017, we proposed a work RVU of 0.59 for CPT code 88350.

The following is a summary of the comments we received regarding our proposed valuations for the Immunohistochemistry family:

Comments: Several commenters stated concerns regarding the level of reimbursement these pathology codes would receive if CMS reduced the work RVUs as proposed. The commenters stated the reduced reimbursement would force pathologists to decrease the number of technical staff, which would interfere with pathologists' ability to perform these services accurately and timely.

The RUC stated the CY 2017 proposed work RVUs for CPT codes 88341 and 88350 do not represent the work involved in furnishing the procedure and present a rank order anomaly for other services. The RUC also stated that the services furnished by CPT codes 37252 and 37253, which we used to establish the relationship between the base code and the add-on code, are not medically comparable services to CPT codes 88341 and 88350. Additionally, the RUC stated each pathology service has individual intensities and complexities. Specifically, for additional immunohistochemistry services represented by add-on CPT codes 88341 and 88350, each antibody is evaluated separately on different slides and each additional service is separate and distinct.

Lastly, the RUC stated its approach of evaluating the actual work associated with each unique base and each unique add-on service is far more accurate, rational, and responsive to the specific circumstances than holding codes equal to a fixed discount from the base code. Applying ratio comparisons and fixed discounts to arrive at a work relative value will continue to create inter-specialty rank order anomalies of physician work RVUs.

Another commenter noted there were RUC surveys that evaluated physician work differentials between the base codes and the add-on codes for pathology services. The commenter offered CPT codes 88333 (Pathology consultation during surgery; cytologic examination (e.g., touch prep, squash prep), initial site) and 88334 (Pathology consultation during surgery; cytologic examination (e.g., touch prep, squash prep), each additional site (List separately in addition to code for primary procedure)) and CPT codes 88331 (Pathology consultation during surgery; first tissue block, with frozen section(s), single specimen) and 88332 (Pathology consultation during surgery; each additional tissue block with frozen section(s) (List separately in addition to code for primary procedure) as examples for consideration.

Response: We appreciate commenters' concern regarding the level of reimbursement and will continue to consider input from the medical community on this issue through evaluation of CPT coding changes and associated RUC recommendations, as well as feedback received through public comments, as we value these services through future notice and comment rulemaking. We also note that the PFS is a relative value system and, as such, values services across all specialties. We believe it is important that there are accurate comparisons between codes in different families.

As discussed in detail in previous proposed and final rules, we continue to believe the metric we use to value add-on codes relative to their base codes is appropriate and representative of the work involved and note that there is no rank order anomaly within this particular code family. In response to the commenter's statement that there should be no comparison of intravascular ultrasound services to any pathology service, we continue to believe any difference in work RVUs for codes describing different kinds of services should reflect the relative differences in time and intensity involved in furnishing the services. Therefore, we believe that it is imperative that we can compare the assumptions regarding overall work between any two codes, regardless of their characteristics.

We appreciate commenters' concerns regarding a standard discount, and we do not consider the use of a particular increment to establish a new standard. Instead, we reiterate that we believe that it is reasonable to estimate work RVUs for a base and an add-on code, and to recognize efficiencies between them, by looking at how similar efficiencies are reflected in work RVUs for other PFS services. We appreciate the commenters' concerns regarding the time ratio methodologies and have responded to these concerns about our methodology in section II.L of this final rule.

Therefore, for CY 2017 we are finalizing a work RVU of 0.56, 0.70, 0.70, and 0.59 for CPT codes 88341, 88342, 88344 and 88350, respectively.

(49) Morphometric Analysis (CPT Codes 88364, 88365, 88367, 88368, 88369 and 88373)

For CY 2015, the CPT Editorial Panel revised the code descriptors for the in situ hybridization procedures, CPT codes 88365, 88367 and 88368, to specify “each separately identifiable probe per block.” Additionally, three new add-on codes (CPT codes 88364, 88369 and 88373) were created to specify “each additional separately identifiable probe per slide.” Some of the add-on codes in this family had RUC-recommended work RVUs that were 60 percent of the work RVU of the base procedure. We believed this accurately reflected the resources used in furnishing these add-on codes and subsequently established interim final work RVUs of 0.53 for CPT code 88364 (60 percent of the work RVU of CPT code 88365); 0.53 for CPT code 88369 (60 percent of the work RVU of CPT code 88368); and 0.43 for CPT code 88373 (60 percent of the work RVU of CPT code 88367).

For CY 2016, the RUC re-reviewed these services due to the specialty society's initially low survey response rate. In our review of these codes, we noticed that the latest RUC recommendation was identical to the RUC recommendation provided for CY 2015. Therefore, we proposed to retain the CY 2015 work RVUs and work time for CPT codes 88367 and 88368 for CY 2016. For CPT code 88365 we finalized a work RVU of 0.88 for CY 2016. For CPT codes 88364 and 88369, we increased the work RVUs for both of these add-on codes from 0.53 to 0.67, which reflected 76 percent of the work RVUs of the base procedures for these services. However, we inadvertently omitted the rationale for this revision to the work RVUs in the preamble to CY 2016 proposed rule. Consequently, we maintained the CY 2015 interim final values of the work RVU of 0.67 for CPT codes 88464 and 88369 and sought comment on these values for CY 2016. For CPT code 88373 we finalized a work RVU of 0.43.

Comment on the CY 2016 PFS final rule with comment period: A few commenters stated their objection to our use of a standard discount for pathology add-on services and for suggesting that each service is separate and unique. Commenters also stated there should be no comparison of intravascular ultrasound services to morphometric analysis, immunohistochemistry, immunofluorescence, or any pathology service.

Response in the CY 2017 PFS proposed rule: In reviewing the RUC recommended base/add-on relationships between several pathology codes, we continue to believe the base/add-on code time relationships for pathology services are appropriate and have not been presented with any compelling evidence that conflicts with the RUC-recommended relationships. However, as we stated above, the intravascular codes we initially examined in revaluing CPT codes 88364 and 88369 were deleted in CY 2016 and replaced with CPT codes 37252 and 37253. For the reasons stated above we continue to believe this 20 percent discount relationship between the base and add-on code accurately reflects the work involved in furnishing these services. Therefore, for CY 2017, we are proposing a work RVU of 0.70 for CPT codes 88364 and 88369 which represents a 20 percent discount from the base code. As the relationship between the base code and add-on code now represents a 20 percent difference we are proposing to revalue CPT code 88373 at 0.58 work RVUs.

In the CY 2017 proposed rule, we proposed a work RVU of 0.58 for CPT code 88373.

The following is a summary of the comments we received regarding our proposed valuation of the Morphometric Analysis codes:

Comments: The RUC stated appreciation for the proposed increase in work RVUs for CPT codes 88364 and 88369 although it stated the increase still does not represent the proper work RVU for the work involved and presents a rank order anomaly relative to other services. The RUC, along with other commenters, stated the services described by CPT codes 37252 and 37253 are not comparable medical services to those furnished by CPT codes 88364 and 88369, and there should be no comparison of intravascular ultrasound services to any pathology services.

The RUC also stated that although some medical procedures and services may present efficiencies between base and add-on services, this is not the case for CPT codes 88364 and 88369, as each pathology service is individual so that any rational comparison of the physician work of intravascular ultrasound services with pathology services is impossible. The RUC also stated that no pathology add-on service can be presumed to have a discount in physician work from the base service.

Another commenter stated for CPT code 88373, it is irrational to assume that second and subsequent services designated by convention as “add-on” services require a reduction in resources relative to the corresponding initial service.

Another commenter noted that in the CY 2017 proposed rule, CMS incorrectly stated it was utilizing a RUC recommendation specific to these codes. According to the CY 2015 Final Rule (79 FR 67548), the codes on which CMS based its discount were CPT codes 88334, 88335, 88177, and 88172. The commenter states the distinction between the codes cited in the CY 2015 final rule, CPT codes 88334, 88335, 88177, 88172, and the new add-on codes, CPT codes 88364, 88369 and 88373, is that the discount factor is specific to services for which a diagnosis has already been furnished. For the new codes to which CMS applied this discount, no such corresponding interpretative diagnosis has been made.

The same commenter stated for morphometric codes, the pathologist is reviewing a second, unique and distinct probe with an entirely different signal than that of its base code, and the work involved with these add-on services requires the same level of intensity and time as their base codes.

The commenter also stated that pathology consultation and cytopathology evaluation codes were clinically different and are not valid proxies to identify efficiencies for the new add-on codes.

Response: We do not agree that there are rank order anomalies within this code family, and we note that this code family was valued within itself and not in relation to other services within the PFS. In response to the commenter's statement that there should be no comparison of intravascular ultrasound services to any pathology service as discussed above, we continue to believe it is valid to compare services across the PFS when determining appropriate values.

We also continue to believe that it is reasonable to recognize efficiencies between them a base and an add-on code. In reviewing the RUC-recommended base/add-on relationships between several pathology codes, we continue to believe the base/add-on code time relationships for pathology services are appropriate and have not been presented with any persuasive evidence or rationale that conflicts with the RUC-recommended relationships.

We agree with the commenter that the designation “add-on” does not automatically imply a reduction; however, in the case of these similar pathology services, we continue to believe using the same valuation metrics is valid. Therefore, for CY 2017, we are finalizing a work RVU of 0.70, 0.73, 0.88, 0.70 and 0.58 for CPT codes 88364, 88367, 88368, 88369 and 88373, respectively.

(50) Liver Elastography (CPT Code 91200)

For CY 2016, we received a RUC recommendation of 0.27 work RVUs for CPT code 91200. After careful review of the recommendation, we established the RUC-recommended work RVU and direct PE inputs as interim final for CY 2016.

Comment on the CY 2016 PFS final rule with comment period: A few commenters requested that we reconsider the level of payment assigned to this service when furnished in a nonfacility setting, stating that the code met the definition for the potentially misvalued code list as there is a significant difference in payment between sites of service. The commenters also asked us to reconsider the assigned 50 percent utilization rate for the FibroScan equipment in this procedure as the current utilization rate would translate to over 50 procedures per week. Instead, the commenters suggested the typical number of procedures done per week ranges between 15 and 25 and requested we adopt a 25 percent utilization rate which corresponds to that number of procedures.

Response in the CY 2017 PFS proposed rule: We refer commenters to the CY 2016 final rule with comment period (80 FR 71057-71058) where we discussed and addressed the comparison of the PFS payment amount to the OPPS payment amount for CPT 91200. For the commenter's statement about the utilization rate, we have previously addressed the accuracy of these default assumptions as they apply to particular equipment resources and particular services. In the CY 2008 PFS proposed rule (72 FR 38132), we discussed the 50 percent utilization assumption and acknowledged that the default 50 percent usage assumption is unlikely to capture the actual usage rates for all equipment. However, we stated that we did not believe that we had strong empirical evidence to support any alternative approaches. We indicated that we would continue to monitor the appropriateness of the equipment utilization assumption, and evaluate whether changes should be proposed in light of the data available. The commenters did not provide any verifiable data suggesting a lower utilization rate. Therefore, for CY 2017 we proposed a work RVU of 0.27 for CPT code 91200, consistent with the CY 2016 interim final value, and we continued to explore and solicit comments regarding publically available data sources to identify the most accurate equipment utilization rate assumptions possible. We also noted that following the publication of the CY 2016 PFS final rule with comment period (80 FR 70886) there was an inconsistency in the Work Time file published on the CMS Web site. For CPT code 91200 the RUC recommended 16 minutes total service time whereas our file reflected 18 minutes total time for the service. For CY 2017, we proposed to update the Work Time file to reflect the RUC's recommendation, which is 16 minutes for CPT code 91200.

The following is a summary of the comments we received regarding our proposed valuation of CPT code 91200.

Comment: Many commenters disagreed with the proposed valuation of CPT code 91200, suggesting that the $34 payment rate in the nonfacility setting for CY 2016 underestimated the resource cost of the procedure. Commenters stated that this code is a first-line method used to assess fibrosis scores in patients with chronic liver disease, especially those with chronic Hepatitis C, and that the current reimbursement was not sufficient to cover the cost of providing the service. Some commenters compared the use of the Fibroscan device in CPT code 91200 to more expensive and more invasive liver biopsies, or compared the cost of the procedure to the treatment provided in hospital-based payment systems. Commenters urged CMS to increase the valuation of CPT code 91200 to encourage providers to adopt the use of the Fibroscan device.

Response: We remind commenters that we are obligated by statute to set payment rates based on the resources used to furnish the procedure, and that as a result pricing for codes on the PFS does not necessarily mirror pricing for codes under different payment systems. We also note that we proposed the RUC-recommended work RVU and direct PE inputs for CPT code 91200 without alteration.

Comment: Many of the commenters also addressed the pricing of the Fibroscan equipment (ER101). Commenters provided CMS a range of different prices for this equipment item, individually suggesting that the equipment costs $120k, $130k, $140k, and $150k. One commenter supplied an individual invoice for the Fibroscan, including the device itself along with a CAP option, an S probe, a printer, and shipping/maintenance costs.

Response: We appreciate the submission of additional information regarding the proper pricing of the Fibroscan. We encourage more commenters to include invoices with their comment submissions if they believe that existing supplies or equipment items are undervalued, as we have had longstanding reservations about establishing pricing based on single invoices. In the specific case of the Fibroscan equipment, we agree that the price should be increased based on the submitted invoice. We are pricing ER101 at $183,390 based on a combination of the cost of the Fibroscan itself ($131,950), the CAP option ($22,955), the S probe ($27,950), and the printer ($495). We note that we do not typically pay for shipping costs or maintenance costs, as equipment maintenance costs are built into the equipment cost per minute formula. We are also changing the name of ER101 from “Fibroscan” to “Fibroscan with printer” to reflect the fact that this pricing incorporates a printer.

After consideration of comments received, we are finalizing our proposed work RVUs and direct PE inputs, with the price increase to the Fibroscan with printer equipment.

(51) Closure of Paravalvular Leak (CPT Codes 93590, 93591, and 93592)

The CPT Editorial Panel developed three new codes (two base codes and one add-on code) to describe paravalvular leak closure procedures that were previously reported using an unlisted code. The RUC recommended a work RVU of 17.97 for CPT code 93591. We proposed a work RVU of 14.50 for CPT code 93591, a direct crosswalk from CPT code 37227. We stated in the CY 2017 proposed rule that we believe that a direct crosswalk to CPT code 37227 accurately reflected the time and intensity described in CPT code 93591 since CPT code 37227 also described a transcatheter procedure with similar service times.

To maintain relativity among the codes in this family, we proposed refinements to the recommended work RVUs for CPT code 93590. The RUC noted that the additional work associated with CPT code 93590 compared to CPT code 93591 was due to the addition of a transseptal puncture to access the mitral valve. The RUC identified a work RVU of 3.73 for a transseptal puncture. Therefore, for CPT code 93590, we proposed a work RVU of 18.23 by using our proposed work RVU of 14.50 for CPT code 93591 and adding the value of a transseptal puncture (3.73).

CPT code 93592 is an add-on code used to report placement of additional occlusion devices for percutaneous transcatheter paravalvular leak closure, performed in conjunction with either an initial mitral or aortic paravalvular leak closure. The RUC recommended a work RVU of 8.00 for this code. In the proposed rule, we stated that we considered applying the relative increment between CPT codes 93590 and 93591; however, we believed that a direct crosswalk to CPT code 35572, with a work RVU of 6.81, more accurately reflected the time and intensity of furnishing the service. Therefore, for CPT code 93592, we proposed a work RVU of 6.81.

Comment: For CPT code 93591, commenters opposed CMS' assertion that a cardiovascular intervention performed in an immobile leg is comparable in intensity and patient risk to an intervention performed in a beating, moving heart. Commenters suggested that CMS' proposed crosswalk to CPT code 37227 was not appropriate since CPT code 37227 is generally performed in an outpatient setting, while CPT code 93591 is generally performed in a facility setting due to the intensity and risk associated with the procedure. Subsequently, commenters suggested that CMS finalize the RUC-recommended work RVU of 17.97 for CPT code 93591.

For CPT code 93590, commenters, including the RUC, supported CMS' proposed use of the same building block methodology used in the RUC recommendations, by proposing to apply a work RVU of 3.73 to the base code value of 93591. However, commenters suggested that CMS apply the value of a transeptal puncture to the RUC-recommended value for CPT code 93591, and therefore, finalize the RUC-recommended work RVU of 21.70 for CPT code 93590.

For CPT code 93592, commenters, including the RUC, disagreed with CMS' proposed comparison of CPT code 93592 to CPT code 35572 (Harvest of femoropopliteal vein, 1 segment, for vascular reconstruction procedure (e.g., aortic valve services)). Commenters stated that CMS's proposed crosswalk is inappropriate and does not recognize the intensity and skill level needed to place a device to close a paravalvular leak in a moving, beating heart, frequently in patients with heart failure. Commenters stated that CPT code 35572 was only similar to CPT code 93592 in that both procedures are cardiovascular in nature. Commenters also stated that surgical harvest of the lower extremity vein is not clinically similar to the transcatheter percutaneous structural heart therapies.

Response: We thank the commenters for their feedback on our proposal. After consideration of the comments received, we are finalizing the RUC-recommended work RVUs for each of the codes in this family. Therefore, we are finalizing a work RVU of 21.70 for CPT code 93590, a work RVU of 17.97 for CPT code 93591, and a work RVU of 8.00 for CPT code 93592.

(52) Electroencephalogram (EEG) (CPT Codes 95812, 95813, and 95957)

In February 2016, the RUC submitted recommendations for work and direct PE inputs for CPT codes 95812, 95813, and 95957. We proposed to use the RUC-recommended physician work and direct PE inputs for CPT code 95957 and to use the RUC-recommended work RVUs for CPT codes 95812 and 95813.

In the CY 2016 PFS final rule with comment period (80 FR 70886), we finalized direct PE input refinements for several clinical labor times for CPT codes 95812 and 95813. The RUC's February 2016 direct PE summary of recommendations indicated that the specialty society expert panel disagreed with CMS' refinements to clinical labor time for these two codes. The RUC recommended 62 minutes for clinical labor task “perform procedure” for CPT code 95812 and 96 minutes for the same clinical labor task for CPT code 95813, similar to the values recommended by the RUC in April 2014.

We proposed to maintain the CMS-refined CY 2016 PE inputs for clinical labor task “perform procedure” for CPT codes 95812 (50 minutes) and 95813 (80 minutes), since the RUC's PE summary of recommendations stated that CPT code 95812 required 50 minutes of clinical labor time for EEG recording, and CPT code 95813 required 80 minutes of clinical labor time for the same clinical labor task.

We did not receive any comments on our proposals for this family of codes. Therefore, for CY 2017, we are finalizing our proposed direct PE inputs for these codes without modifications. We are also finalizing for CY 2017 work RVUs of 1.08 for CPT code 95812, 1.63 for CPT code 95813, 1.98 for CPT code 95957.

(53) Analysis of Neurostimulator Pulse Generator System (CPT Codes 95971, 95972)

CPT codes 95971 and 95972 were established as interim final following the CY 2016 final rule with comment period. For CY 2017, we proposed to maintain their work RVUs and direct PE inputs.

Comment: A commenter expressed support for the proposal to maintain the current work and PE RVUs, stating that these codes were revalued in 2015 and there was no reason to make any changes.

Response: We appreciate the support from the commenter.

After consideration of comments received, we are finalizing our proposed work RVUs and proposed direct PE inputs for CPT codes 95971 and 95972.

(54) Patient, Caregiver-focused Health Risk Assessment (CPT Codes 96160 and 96161)

In October 2015, the CPT Editorial Panel created two new PE-only CPT codes, 96160 (Administration of patient-focused health risk assessment instrument (e.g., health hazard appraisal) with scoring and documentation, per standardized instrument) and 96161 (Administration of caregiver-focused health risk assessment instrument (e.g., depression inventory) for the benefit of the patient, with scoring and documentation, per standardized instrument). For CPT code 96160, we proposed the RUC-recommended direct PE inputs. For CPT code 96161, the service is furnished to a patient who may not be a Medicare beneficiary, and therefore, we did not believe the code would be eligible for Medicare payment. We proposed to assign a procedure status of I (Not valid for Medicare purposes) for CPT code 96161.

We noted that we believed that CPT code 96160 describes a service that is frequently reasonable and necessary in the treatment of illness or injury, such as when there has been a change in health status. However, when the service described by CPT code 96160 is explicitly included in another service being furnished, such as the Annual Wellness Visit (AWV), this code should not be billed separately, much like other codes that describe services included in codes with broader descriptions. We also noted that this service should not be billed separately if furnished as a preventive service as it would describe a non-covered service. However, we also solicited comment on whether this service may be better categorized as an add-on code and welcomed stakeholder input regarding whether or not there are circumstances when this service might be furnished as a stand-alone service.

Comment: Many commenters recommended that CMS should recognize and make separate payment for CPT code 96160, as proposed, as well as 96161 using the RUC-recommended values. Several of these commenters argued that the medical community has recognized that health risk assessment of caregivers is an integral part of ongoing medical care for patients with particular needs. These commenters offered several examples where such an assessment is integral to treating patients, such as:

• Assessment of maternal depression in the active care of infants,
• Assessment of parental mental health as part of evaluating a child's functioning,
• Assessment of caretaker conditions as indicated where atypical parent/child interactions are observed during care,
• Assessment of caregivers as part of care management for adults whose physical or cognitive status renders them incapable of independent living and dependent on another adult caregiver. Some examples might be intellectually disabled adults, seriously disabled military veterans and adults with significant musculoskeletal or central nervous system impairments.

Because commenters noted that these assessments were generally administered during E/M services, they were receptive to making both CPT codes 96160 and 96161 add-on codes to E/M services.

Response: After considering comments, we believe that CPT codes 96160 and 96161 describe services that, in particular cases, can be necessary components of services furnished to Medicare beneficiaries. While we recognize that in many cases we have previously assigned non-payment indicators to codes that describe interactions with caregivers, we also note that we have also recognized that in current medical practice, practitioner interaction with caregivers is an integral part of treatment for some patients. Accordingly, the descriptions for several payable codes under the PFS include direct interactions between practitioners and caregivers.

In developing our proposal regarding the payment disposition of this code, we noted that it singularly described a service administered to a caregiver. However, based on public comments, including the receptivity to our assignment of add-on code status, we understand that in actual practice, this service is integrated with E/M visits under particular circumstances. Consequently, we believe the appropriate payment status for the code should be determined by looking at the overall service as described by the two codes together. We agree with commenters, then, that there are circumstances where this service is an essential part of a service to a Medicare beneficiary. Therefore, we are assigning an active payment status to both codes for CY 2017. We are also establishing use of the RUC recommended values for these codes. We are also assigning an add-on code status to both of these services. As add-on codes, CPT codes 96160 and 96161 describe additional resource components of a broader service furnished to the patient that are not accounted for in the valuation of the base code.

(55) Reflectance Confocal Microscopy (CPT Codes 96931, 96932, 96933, 96934, 96935, and 96936)

For CY 2015, the CPT Editorial Panel established six new Category I codes to describe reflectance confocal microscopy (RCM) for imaging of skin. For CPT codes 96931 and 96933, the specialty society and the RUC agreed that the physician work required for both codes were identical, and therefore, should be valued the same. The RUC recommended a work RVU of 0.80 for CPT codes 96931 and 96933 based on the 25th percentile of the survey. Based on the similarity of the services being performed in CPT codes 96931 and 96933 and the identical intra-service times of 96931, 96933 and 88305, the key reference code from the survey, we believe a direct crosswalk from CPT code 88305 to CPT codes 96931 and 96933 would more accurately reflect the work involved in furnishing the procedure. Therefore, for CY 2017, we proposed a work RVU of 0.75 for CPT codes 96931 and 96933. In addition, we proposed removing 3 minutes of preservice time from CPT codes 96931 and 96933 since it is not included in CPT code 88305 and as a result, we did not believe it was appropriate in CPT codes 96931 and 96933.

For CPT codes 96934 and 96936, the specialty society and the RUC agreed that the physician work required for both codes were identical, and therefore, should be valued the same. In its recommendation, the RUC stated that it believed the survey respondents somewhat overestimated the work for CPT code 96934 with the 25th percentile yielding a work RVU of 0.79. Consequently, the RUC reviewed the survey results from CPT code 96936 and agreed that the 25th percentile work RVU of 0.76 accurately accounted for the work involved for the service. Therefore, the RUC recommended a work RVU of 0.76 for CPT codes 96934 and 96936.

We believe that the incremental difference between the RUC-recommended values for the base and add-on codes accurately captures the difference in work between the code pairs. However, because we valued the base codes differently than the RUC, we proposed values for the add-on codes that maintain the RUC's 0.04 increment instead of the RUC-recommended values. Therefore we proposed a work RVU of 0.71 for CPT codes 96934 and 96936.

We also proposed to reduce the preservice clinical labor for “Patient clinical information and questionnaire reviewed by technologist, order from physician confirmed and exam protocoled by physician” for CPT codes 96934 and 93936 as this work is performed in the two base CPT codes 93931 and 93933. We proposed to reduce the service period clinical labor for “Prepare and position patient/monitor patient/set up IV” from 2 to 1 minute for CPT codes 93934 and 93936 since we believed that less positioning time is needed with subsequent lesions. We proposed to refine the service period clinical labor for “Other Clinical Activity—Review imaging with interpreting physician” to zero minutes for CPT codes 96933 and 96936 as these are interpretation and report only codes and not image acquisition.

Comment: Several commenters, including the RUC, objected to the proposed valuations for CPT codes 96931, 96933, 96934, and 96936. The RUC disagreed with pre-service time being removed from a survey code simply due to a key reference code not also having pre-service time. The RUC stated CPT codes 96931 and 96933 are distinct procedures from CPT codes 88305 and the CMS proposal to remove 3 minutes of pre-time from the base RCM codes was grounded on faulty logic. The RUC stated its agreement with the specialty society that 3 minutes of preservice time was necessary for the physician to review clinical history and referral information. The RUC further stated with the 3 minutes of pre-service time in its recommendation for the RCM base codes were appropriately in line with top key reference CPT code 88305 and urged CMS to accept the survey 25th percentile work RVUs for CPT codes 96931, 96933, 96934, and 96936. Other commenters stated there were very significant differences in the technologies used and the work involved between the procedures of CPT code 88305, the key reference code, and CPT codes 96931 and 96933, with CPT codes 96931 and 96933 being more complex procedures.

One commenter stated CMS incorrectly removed technician time for “Other Clinical Activity—Review imaging with interpreting physician” for CPT codes 96933 and 96936 noting the technician still must review the imaging with the interpreting physician and urged CMS to accept the RUC recommendations.

Response: After consideration of comments received, we agree with the commenters and will finalize the RUC-recommended work RVUs of 0.80, 0.80, 0.76, and 0.76 for CPT codes 96931, 96933, 96934 and 96936; respectively. We will also restore the 3 minutes of preservice time to CPT codes 96931 and 96933.

(56) Evaluative Procedures for Physical Therapy and Occupational Therapy (CPT Codes 97161, 97162, 97163, 97164, 97165, 97166, 97167, 97168)

For CY 2017, the CPT Editorial Panel deleted four CPT codes (97001, 97002, 97003, and 97004) and created eight new CPT codes (97161-97168) to describe the evaluative procedures furnished by physical therapists and occupational therapists. There are three new codes, stratified by complexity, to replace a single CPT code 97001, for physical therapy (PT) evaluation, and three new codes, also stratified by complexity, to replace a single code CPT code 97003, for occupational therapy (OT) evaluation, and one new code each to replace the re-evaluation codes for physical and occupational therapy, CPT codes 97002 and 97004. Table 23 includes the long descriptors and the required components of each of the eight new CPT codes for the PT and OT services.

The CPT Editorial Panel's creation of the new codes for PT and OT evaluative procedures grew out of a CPT workgroup that was originally convened in January 2012 when contemplating major revision of the Physical Medicine and Rehabilitation CPT section of codes in response to our nomination of therapy codes as potentially misvalued codes, including CPT code 97001 (and, as a result, all four codes in the family) in the CY 2012 PFS proposed rule.

In reviewing the eight new CPT codes for evaluative procedures, the HCPAC forwarded recommendations for work RVUs and direct PE inputs for each code. Currently, CPT codes 97001 and 97003 both have a work RVU of 1.20, and CPT codes 97002 and 97004 both have a work RVU of 0.60. These CPT codes have reflected the same work RVUs since CY 1998 when we accepted the HCPAC values during CY 1998 rulemaking.

i. Valuation of Evaluation Codes

In the CY 2017 PFS proposed rule, we noted that the HCPAC submitted work RVU recommendations for each of the six new PT and OT evaluation codes. These recommendations are intended to be work neutral relative to the valuation for the previous single evaluation code for PT and OT, respectively. However, that assessment for each family of codes is dependent on the accuracy of the utilization forecast for the different complexity levels within the PT or OT family. As used in this section, work neutrality is distinct from the budget neutrality that is applied broadly in the PFS. Specifically, work neutrality is intended to reflect that despite changes in coding, the overall amount of work RVUs for a set of services is held constant from one year to the next. For example, if a service is reported using a single code with a work RVU of 2.0 for one year but that same service would be reported using two codes, one for “simple” and another for “complex” in the subsequent year valued at 1.0 and 3.0 respectively, work neutrality could only be attained if exactly half the services were reported using each of the two new codes. If more than half of the services were reported using the “simple” code, then there would be fewer overall work RVUs. If more than half of the services were reported using the “complex” code, then there would be more overall work RVUs. Therefore, work neutrality can only be assessed with an understanding of the relative frequency of how often particular codes will be reported.

The HCPAC recommended a work RVU of 0.75 for CPT code 97161, a work RVU of 1.18 for CPT code 97162, and a work RVU of 1.5 for CPT code 97163. The PT specialty society projected that the moderate complexity evaluation code would be reported 50 percent of the time because it is the typical evaluation, and the CPT codes for the low and high complexity evaluations are each expected to be billed 25 percent of the time. The HCPAC-recommended work RVU of 1.18 for CPT code 97162 represents the survey median with 30 minutes of intraservice time, 10 minutes of preservice time, and 15 minutes postservice time. The HCPAC notes this work value is appropriately ranked between levels 2 and 3 of the E/M office visit codes for new patients.

The HCPAC recommended a work RVU of 0.88 for CPT code 97165, a work RVU of 1.20 for CPT code 97166, and a work RVU of 1.70 for CPT code 97167. For the OT codes, work neutrality would be achieved only with a projected utilization in which the low complexity evaluation is billed 50 percent of the time; the moderate complexity evaluation is billed 40 percent of the time, and the high complexity evaluation only billed 10 percent of the time. For purposes of calculating work neutrality, the HCPAC recommended assuming that the low complexity code will be most frequently reported even though the HCPAC-recommended work RVU of 1.20 and 45 minutes of intraservice time for moderate complexity code is identical to that of the current OT evaluation code. The HCPAC believes that the work RVU of 1.20 is appropriately ranked between 99202 and 99203, levels 2 and 3 for E/M office visits for new outpatients.

ii. Valuation of Evaluation Codes and Discussion of PAMA

In our review of the HCPAC recommendations, we noted the work neutrality and the inherent reliance on the utilization assumptions. We considered the three complexity levels for the PT evaluations and the three complexity levels for the OT evaluations; and we also considered the evaluation services described by the codes as a whole. The varying work RVUs and the dependence on utilization for each complexity level to ensure work neutrality in the PT and OT code families make it difficult for us evaluate the HCPAC's recommended values or to predict with a high degree of certainty whether physical and occupational therapists will actually bill for these services at the same rate forecast by their respective specialty societies.

We were concerned that the coding stratification in the PT and OT evaluation codes may result in upcoding incentives, especially while physical and occupational therapists gain familiarity and expertise in the differential coding of the new PT and OT evaluation codes that now include the typical face-to-face times and new required components that are not enumerated in the current codes. We were also concerned that stratified payment rates may provide, in some cases, a payment incentive to therapists to upcode to a higher complexity level than was actually furnished to receive a higher payment.

We understood that there may be multiple reasons for the CPT Editorial Panel to stratify coding for OT and PT evaluation codes based on complexity. We also noted that the codes will be used by payers in addition to Medicare, and other payers may have direct interest in making such differential payment based on complexity of OT and PT evaluation. Given our concerns regarding appropriate valuation, work neutrality, and potential upcoding, however, we did not believe that making different payment based on the reported complexity for these services is, at current, advantageous for Medicare or Medicare beneficiaries.

Given the advantages inherent and public interest in using CPT codes once they become part of the code set, we proposed to adopt the new CPT codes for use in Medicare for CY 2017. However, given our concerns about appropriate pricing and payment for the stratified services, we proposed to price the services described by these stratified codes as a group instead of individually. To do that, we proposed to utilize the authority in section 220(f) of the Protecting Access to Medicare Act (PAMA), which revised section 1848(c)(2)(C) of the Act to authorize the Secretary to determine RVUs for groups of services, rather than determining RVUs at the individual service level. We believed that using this authority instead of proposing to make payment based on Medicare G-codes will preserve consistency in the code set across payers, thus lessening burden on providers, while retaining flexibilities that are beneficial to Medicare.

We proposed a work RVU of 1.20 for both the PT and the OT evaluation groups of services. We proposed this work RVU because we believed it best represents the typical PT and OT evaluation. This is the value recommended by the HCPAC for the OT moderate complexity evaluation and nearly the same work RVU for corresponding PT evaluation (1.18). Additionally, a work RVU of 1.20 is the long-standing value for the current evaluation codes, CPT codes 97001 and 97003, and thus, assures work neutrality without reliance on particular assumptions about utilization, which we believed was the intent of the HCPAC recommendation.

Because we proposed to use the same work RVU for the six evaluation codes, we are not addressing any additional concerns about the utilization assumptions recommended to us. Because we proposed the same work values for each code in the family, there will be no ratesetting impact to work neutrality. As such, we are not revising the utilization crosswalks as projected by the respective therapy specialties to achieve work neutrality. However, were we to value each code in the PT or OT evaluation families individually, we would seek objective data from stakeholders to support the utilization crosswalks, particularly those for the OT family in which the low-level complexity evaluation is depicted as typical and the high complexity is projected to be billed infrequently at 10 percent of the overall number of OT evaluations.

We proposed to use the direct PE inputs forwarded by the HCPAC (with the refinements described below) for the moderate complexity PT and OT evaluations in the development of PE RVUs for the PT and OT codes as a group of services. For the PT codes, we proposed to use the recommended inputs for the moderate complexity code for the direct PE inputs of all three codes based on its assumption as the typical service. Our proposed direct PE inputs reflect the recommended values minus 2 minutes of physical therapist assistant (PTA) time in the service period because we believe that PTA tasks to administer certain assessment tools are appropriately included as part of the physical therapist's work and the time of the PTA to explain and score self-reported outcome measures is not separately included in the clinical labor of other codes. We proposed to include the recommended four sheets of laser paper without an association to a specific equipment item, but we solicited comment regarding the paper's use.

For the OT evaluation codes, we considered proposing to use the direct PE inputs for the low complexity evaluation because the OT specialty organization believes it represents the typical OT evaluation service with a projected 50 percent utilization rate. However, we proposed to use the moderate-level direct inputs instead, because the direct PE for this level is based on a vignette that is valued with the same intraservice time, 45 minutes, as the current code, CPT code 97003. Consequently, we proposed to use the recommended direct PE inputs for the moderate complexity code for use in developing PE RVUs for this group of services.

Our proposed direct PE inputs reflect the recommended values minus 2 minutes of occupational therapist assistant (OTA) time in the service period because we believe that OTA tasks to administer certain assessment tools are appropriately included as part of the occupational therapist's work and the time of the OTA to explain and score self-reported outcome measures is not separately included in the clinical labor of other codes. We also rounded up the recommended 6.8 minutes to 7 minutes to represent the time the OTA assists the occupational therapist during the intraservice time period. For the Vision Kit equipment item, our proposed price reflects the submitted invoice that clearly defined a kit.

iii. Valuation of Re-evaluation Codes

The recommendations the HCPAC sent to us for the PT and OT re-evaluation codes are not work neutral. For the new PT re-evaluation code, CPT code 97164, the HCPAC recommended a work RVU of 0.75 compared to the work RVU of 0.60 for CPT code 97002. This recommended work RVU falls between the 25th percentile of the survey and the survey's median value and was based on a direct crosswalk to CPT code 95992 for canalith repositioning with 20 minutes intraservice time and 10 minutes immediate postservice time. The HCPAC supported this 0.15 work RVU increase based on an anomalous relationship between PT services and E/M office visit codes for established patients, noting that physician E/M codes have historically been used as a relative comparison. The HCPAC stated its recommendation of a work RVU of 0.75 for CPT code 97164 appropriately ranks it between the key reference codes for this service, CPT codes 99212 and 99213, levels 2 and 3 E/M office-visit codes for established patients.

The HCPAC provided a work RVU of 0.80 for the OT re-evaluation code, CPT code 97168, based on the 25th percentile of the survey, which represents an increase over the current work RVU of 0.60 for CPT code 97004. This work value includes 30 minutes of intraservice time, 5 minutes preservice time, and 10 minutes immediate postservice time. The HCPAC noted that the increase in work compared to the PT re-evaluation code (0.75) is because the occupational therapist spends more time observing and assessing the patient and, in general, the OT patient typically has more functional and cognitive disabilities. The HCPAC recommendation notes that the 0.80 work RVU recommendation appropriately ranks it between the level 1 and 2 E/M office-visit codes for new patients.

The HCPAC's recommended increases to work RVUs for the PT and OT re-evaluation codes are not work neutral. We are unclear why the HCPAC did not maintain work neutrality for the OT and PT re-evaluation codes since maintaining work neutrality was important to the establishment of the six new evaluation codes. We proposed to maintain the overall work RVUs for these services by proposing a work RVU of 0.60 for CPT codes 97164 and 97168, consistent with the work RVUs for the deleted re-evaluation codes. We solicited comments from stakeholders on whether there are reasons that the re-evaluation codes should be revalued without regard to work neutrality.

We proposed the HCPAC-recommended direct PE inputs for CPT code 97164 with a reduction in time for the PTA by 1 minute (from 5 to 4) in the service period—the line for “Other Clinical Activity”—because the time to explain and score the self-reported outcome measure (for example, Oswestry) is not separately included in the clinical labor of other codes.

We proposed the HCPAC-recommended direct PE inputs for CPT code 97168 with a reduction in time for the OTA by 1 minute (from 3 to 2) in the service period—the line for “Other Clinical Activity”—for the same reason we proposed to reduce the corresponding line for PTAs—because the time to explain and score any patient-self-administered functional and other standardized outcome measure is not separately included in the clinical labor of other codes.

Because the new CPT code descriptors contain new coding requirements for each complexity level, we solicited comment from the PT and OT specialty organizations, as well as other stakeholders to clarify how therapists will be educated to distinguish the required complexity level components and the selection of the number of elements that impact the plan of care. For example, for the OT codes, we invited comment on how to define performance deficits, what process the occupational therapist uses to identify the number of these performance deficits that result in activity limitations, and performance factors needed for each complexity level. For the PT codes, we sought more information about how the physical therapist differentiates the number of personal factors that actually affect the plan of care. We were also interested in understanding more about how the physical therapist selects the number of elements from any of the body structures and functions, activity limitations, and participation restrictions to make sure there is no duplication during the physical therapist's examination of body systems.

The following is summary of the comments we received:

Comment: Several commenters disagreed with our proposal to accept the new CPT codes for PT and OT evaluations and re-evaluations and urged us to keep the current four-code set. A few of these commenters noted our proposal to accept the stratified code sets for PT and OT evaluations would increase the administrative burden associated with documentation and education training of therapists, billers and coders. Other commenters believed that CMS should first implement the new complexity-defined CPT code set on a demonstration or pilot project basis before we apply it nationally. One commenter proposed that, rather than accepting the new CPT eight-code set with varying descriptors for each PT and OT complexity level, we adopt just two codes that both the PT and OT disciplines could use: a code for PT/OT evaluation and another for PT/OT re-evaluation. Another commenter told us that “implementation of the complex scheme for determining the evaluation level will excessively complicate patient evaluations where clinicians will require more mental effort to meet the demands of the documentation with less time and attention directed at treating the patient.” One commenter suggested that instead of implementing the stratified code sets, CMS should develop an alternative coding and payment model for therapy services and recommended that we create a value-based payment program, consistent with the Triple Aim of health care, which includes reliable and valid outcome and quality measures to demonstrate the outcome and value of therapy.

Response: We thank the commenters for voicing their concerns about our adoption of the new CPT codes for PT and OT evaluative procedures and their alternative coding suggestions. However, we note that we do not have the authority to change CPT code descriptors or use deleted codes without creating G-codes to do so. We also note that adopting a demonstration or pilot program is not a typical CMS payment policy response to the creation of new CPT codes or code sets. After considering these comments, we continue to believe that our proposal to adopt the eight new CPT codes for use in Medicare for CY 2017, rather than retain the current coding structure by creating G-codes, is the best option given the advantages inherent and public interest in using the CPT codes once they become part of the code set. As such, we are finalizing our proposal to adopt new CPT codes 97161-97168 for PT and OT evaluations and re-evaluations.

Comment: Many commenters objected to our proposal to use the PAMA authority to price the services described by the stratified sets of PT and OT evaluation codes as a group instead of individually and asked us to accept or consider the HCPAC work RVU values for each of these six evaluation services. Some commenters expressed concern that we ignored the HCPAC recommendations and proposed to maintain the work RVU of 1.20, since the codes have not been reviewed for this purposes in nearly 20 years. Other commenters stated that CMS, by valuing the PT and OT evaluation complexity levels with the same work RVUs, was failing to appropriately align cost and quality as mandated in the ACA and MACRA.

Because we proposed the same values, a few commenters were concerned that we failed to discuss the difference in the PT and OT evaluation services. These commenters told us that the HCPAC recommendations included higher work RVUs for the OT services because they reflected greater intraservice times from the surveys, and these times led, in part, to the HCPAC's belief that the typical patient receiving OT services is more complex and intense to treat than the patient receiving PT services. The HCPAC and the OT specialty society urged us to consider the increase in work RVUs for the OT evaluative services, indicating in their comments that while the HCPAC recommendations for the PT evaluations were work-neutral, those for the OT evaluations were not. The HCPAC requested that we consider the difference in PT services versus OT services.

Some commenters presumed that our proposal to value the work the same for each evaluation complexity level was temporary. Another commenter expressed hope that we did not intend to equally value the PT high complexity evaluation the same as the low complexity one in perpetuity. Several commenters requested that CMS describe our future plans to revisit these code sets and asked that the future proposal for these payment amounts be subject to public comment. One of these commenters that favored keeping the current code structure urged us not to adopt the new CPT codes until we are ready to differentiate payments based on the complexity of the provided service.

Some commenters told us that our lack of payment stratification for the three PT and three OT evaluation codes would likely prompt coding and billing behavioral change by some therapists and other providers of therapy services. One of these commenters claimed that assigning the same work RVU to each evaluation complexity level would cause some providers not to adhere to the new coding stratification which could result in inaccurate data on the levels being reported. Another commenter stated that the lack of payment stratification to reflect the therapist's time and expertise at each complexity level could signal to therapists that the accurate coding of evaluations is of diminished interest to CMS. Other commenters stated that the failure to recognize payment stratification between the complexity levels would be detrimental to patient care and the practice of therapy, for example, by reducing incentives for therapists to thoroughly evaluate patients with multiple and complex conditions who fall into the high complexity evaluation.

Response: After a review of the comments, we continue to believe that using the PAMA RVU authority to value the PT and OT evaluation codes as a group of services is appropriate. Given our concerns about appropriate pricing and payment for the PT and OT stratified evaluation services as described in the CY 2017 proposed rule, we are finalizing our proposal to use the PAMA authority to value services as groups rather than individually—valuing each complexity level at 1.2 work RVUs for the PT and OT family of evaluation codes for CY 2017. We believe this policy has advantages for the Medicare program. It limits the incentives for and consequences of upcoding by therapists and providers, especially as therapists become more familiar with the new set of codes. Additionally, the policy assures work neutrality for these PT and OT code families while allowing us to collect and analyze utilization data of the complexity levels for possible future rulemaking.

We understand commenters' concerns about the possibility that the absence of payment stratification in the complexity levels of the PT and OT evaluations could have an effect on some therapists' coding behavior in for these services in CY 2017. However, we are also concerned with the implication that financial incentives are the primary drivers for accurate coding for a significant number of therapists, and if that is the case, we believe that implementing stratified coding would likely encourage upcoding since that is consistent with the financial incentives. We believe that the implementation of these new PT and OT code sets carries with it an inherent change for the therapists furnishing the services since there will be three complexity levels to replace just one and each new code contains newly defined necessary components. We also believe that it is premature to predict how therapists will code and bill the new complexity levels before therapists gain familiarity with the new codes.

Comment: We received several comments on utilization assumptions inherent to the HCPAC recommendations. Several commenters questioned why we did not treat the HCPAC-recommended utilization assumptions for the PT and OT complexity-stratified evaluation code sets as we have historically treated other codes sets that come to us from the HCPAC or RUC; that is, using the utilization assumptions provided in the recommendations. The HCPAC explained that if the assumptions are overestimated, the HCPAC or RUC will examine and determine whether to recommend reductions.

We received several comments from stakeholders in response to our statement in the proposed rule that we would request additional objective data to support the utilization crosswalks, especially for the OT codes, if we were to value the codes individually for the PT and OT evaluation complexity levels. In its comments, the OT specialty society explained that their frequency estimations of the three complexity levels were based on the most recent utilization frequency data from the 2014 Medicare utilization from the five percent sample file. The OT specialty society also stated that it defined the complexity levels using certain groups of diagnoses and patient types. The PT specialty society stated that because its survey process included a broad cross-section of therapists working in the various Medicare settings, it believed its utilization projections for the low, moderate and high complexity evaluation were representative. Many commenters told us because some therapists may not initially code the complexity levels correctly, that we would need to consider an entire year of utilization data to ensure its accuracy.

Response: We appreciate the views expressed and the information that the commenters forwarded to us. However, we continue to have concerns that therapists, particularly occupational therapists, will not bill with the same utilization frequencies forecast by their specialty societies for the low, moderate, and high complexity evaluations as described in the CY 2017 proposed rule. In other words, we are concerned with the possibility that we would establish rates (including for purposes of PFS budget neutrality) that rely on the national organizations' assessment of what ought to be billed, but Medicare spending and subsequent PFS budget neutrality assumptions will reflect actual billing given the financial incentives inherent in stratified payment. Should we propose to value the evaluation codes individually in future rulemaking, we would seek additional objective data at that time. We agree that an entire year of data is likely needed to appropriately analyze the utilization of these evaluation services. We appreciate that our historical practice regarding significant revision of CPT coding scheme has required us to make significant assumptions regarding utilization for new codes. We note that in many cases, we have not accepted the assumptions recommended by specialty societies and the RUC and that we were not pricing groups of services together in the past.

Comment: Several commenters expressed concern about the new PT and OT CPT code descriptors, specifically, that each descriptor includes minimal coding requirements. Several commenters expressed skepticism that therapists will be able to report the new codes accurately—one of these commenters believes the new codes rely on subjective clinical reasoning and decision making that will lead to further significant coding and audit concerns for CMS. Several commenters told us that they believe the true complexity of evaluating patients cannot be solely based on personal factors, comorbidities, performance deficits, or time requirements. One of these commenters noted that some patients with multiple comorbidities and body structures involved are not complicated, while others with few comorbidities and body structures involved are deceptively very complex, difficult to diagnose and treat. Another commenter specifically recommended that each PT and OT evaluation complexity level should have the same timeframes, as well as the same component requirements. A few commenters voiced concern about how CMS and our contractors will note these multiple required components of each CPT code. One commenter noted that an evaluation may have characteristics that fall between two complexity levels and told us that it should be up to the clinician to determine which level is most appropriate. A few commenters noted that the new detailed requirements that dictate the level of each code's definition may cause confusion for physical and occupational therapists, especially as they begin to navigate the new codes.

Response: We appreciate the commenters' concerns about new code descriptors that detail the minimal required components for each of the eight new PT and OT evaluative procedures. We realize that it may take time to train therapists about the various required components of each new PT and OT evaluative procedure code and we have addressed this training in the comment and response below. We also appreciate the commenters' concern that the evaluative process is likely more complex than the component parts comprising each code's new coding requirements; however, as noted in the CY 2017 proposed rule, we proposed to adopt the new CPT codes for PT and OT evaluative procedures rather than propose a different coding structure using G-codes. We would like to clarify for the commenters that were concerned about “time requirements” in the new PT and OT CPT code descriptors for evaluative procedures that these “typical times” are included as a frame of reference and do not represent a minimum coding requirement. Just as the typical times included for each E/M code represent the physician face-to-face time with the patient, the typical times in the new PT and OT CPT codes represent the typical face-to-face time of the physical or occupational therapist with the patient. Regarding the commenter's concern about evaluations that fall between two complexity levels, we would note general coding principles applicable to all codes—that the therapist should select the evaluation complexity level that best represents the furnished service and for which the medical necessity is clearly documented.

Comment: Many commenters requested that we delay documentation requirements for the new PT and OT evaluative procedure codes; several commenters requested a one year reprieve from application of medical review and audit requirements; and a few commenters requested that we delay the implementation of the new CPT codes until CY 2018. Most of these delay requests, commenters told us, were related to the time needed to educate therapists about the new codes. Most of these commenters who asked us not to implement new documentation requirements also supported payment stratification of the complexity levels for the PT and OT evaluation complexity levels. Concerned about the proposed lack of payment stratification, the PT specialty society noted in its comments that it asked CPT to postpone the codes for CPT 2017, but CPT denied the request. In its comments, the PT society, along with a few other stakeholders, also asked CMS to delay implementing the new CPT codes for CY 2017 “if there is any way possible that does not disrupt patient care.”

A few commenters say they will need a delay of six months, at a minimum, to train therapists, since all new descriptors include various required elements and the typical time for each PT and OT complexity level and the re-evaluation codes. The majority of commenters, though, indicated they would need a year for their educational efforts to be successful. In addition to therapists, a few commenters told us they would have to educate coders and billers in the use of the new CPT codes. A few commenters noted the time to implement these new codes into their billing systems was too short.

The PT and OT specialty societies each told us about their plans to educate their therapist members and nonmembers to ensure coding accuracy. Each therapy association has already begun this training, some of which will include webinars, self-paced online courses, frequently asked questions, documentation resources, published articles, etc.

Some commenters asked CMS to work with various stakeholders and to either establish guidelines or assist in educating therapists about the new codes through Open Door Forums, MLN articles, etc. Additionally, they also wanted to work with CMS on LCDs established by contractors. One commenter stated that CMS must provide clear guidance regarding the selection of the appropriate level of evaluation services provided by physical and occupational therapists and the associated documentation requirements to ensure consistency and appropriate reporting of these services.

Several commenters asked us to consider a one-year reprieve from the payment consequences of medical review and audit requirements that address lack of documentation to support the complexity level of the code billed.

Response: We understand that implementing the new code sets for PT and OT evaluative procedures will require time for therapists to be educated in their proper use. We would like to remind those requesting we assist in writing guidelines that the CPT manual PMR subsections for PT and OT Evaluations contain official CPT guidelines. We understand the many requests for delay of new documentation requirements during the initial year of their use. As such, for CY 2017, we will delay changes to our current manual instructions for documentation for evaluations and re-evaluations in the Medicare Benefits Policy Manual (MBPM), chapter 15, section 220.3.

We understand and appreciate that the PT and OT specialty societies are already underway in their educational efforts of therapists, as it has been our past experience with the implementation of other CPT codes and code sets that the leading educational role is assumed by the specialty societies responsible for the code changes.

Comment: We received many comments objecting to our proposal to maintain a work RVU of 0.60 for the re-evaluation codes. Many commenters—including therapy specialty societies and organization representing therapy providers and private practice physical and occupational therapists, among other stakeholders—disagreed with our proposal to maintain the work RVUs for the PT and OT re-evaluation codes and expressed their disappointment that we did not consider or accept the HCPAC recommendations for increased work RVUs of 0.75 for PT (CPT code 97164) and 0.8 for OT (CPT code 97168).

One commenter supported increasing the work RVUs, but suggested that the PT and OT re-evaluation codes should be equally valued for the relative work, PE and MP RVUs. This same commenter contended that because the patients treated by the PT and OT disciplines for hand rehabilitation are the same; that is, have the same functional and cognitive deficits, the same time and expertise of both physical and occupational therapists is required to perform a thorough re-evaluation. The commenter recommended that both re-evaluations reflect the 30-minute typical time that is inherent to the OT re-evaluation code.

Several commenters reminded us that the work RVU recommendations forwarded to us were not considered work neutral because the HCPAC accepted the PT and OT specialty societies' beliefs as compelling evidence that the practice of PT and OT have each significantly changed over the past two decades.

Some commenters reasoned that we should accept the HCPAC-recommended work RVUs for these codes, in part, because the PT and OT specialty societies completed the RUC-HCPAC defined survey process, including time and intensity of the services.

Comments from the HCPAC, the PT and OT specialty societies, and a few other stakeholders provided the rationale that the practice of PT and OT has significantly changed since 1997, including the work of physical therapists and occupational therapists. Some of their rationale included: (a) advances in technology has created opportunities for additional types of treatment approaches; and, (b) the work RVUs for the PT and OT re-evaluation codes have not kept pace with the relativity of increases in work RVUs of comparable E/M codes that have historically been used as comparison: In 1997 the 0.60 work RVUs for CPT codes 97002 and 97004 was 90 percent of that for CPT code 99213; today, it is just 62 percent. Other rationales included ones often cited by commenters requesting increases in RVUs, including increased patient acuity and administrative and reporting burdens.

Response: We appreciate the commenters' remarks and the rationale forwarded in response to our request for comments. After a careful consideration of the comments, we agree that modification of our proposal, to recognize the change in practice since 1997 for the work of physical and occupational therapists, is appropriate. Because we believe that PT and OT have similar work, though, we are finalizing the value of both codes at the same work RVUs by assigning a work RVU of 0.75—the HCPAC-recommended work RVU for the PT re-evaluation and the PT low complexity evaluation.

We would like to take this opportunity to remind physical and occupational therapists about our manual instructions regarding the reporting of a both the evaluation and re-evaluation codes (MBPM, Chapter 15, section 220). Of note, to be separately payable, the re-evaluation requires a significant change in the patient's condition or functional status that was not anticipated in the plan of care. The MBPM full definitions follow in Table 22.

Comment: We received a few comments regarding our PE proposals in the CY 2017 proposed rule for the PT and OT evaluation and re-evaluation codes. In its comments, the PT specialty society, in response to our PE proposal, explained, per our request, the use of the 4 sheets of paper as supply items in the PT evaluation and re-evaluation codes. The OT specialty society noted that they accepted the PE refinements we proposed in the proposed rule.

Response: We appreciate the comments from both the PT and OT specialty societies. We will finalize the PE input changes as proposed and include them in the calculation of the final PE RVUs of the PT and OT evaluation and re-evaluation codes.

After considering the comments, in summary, we are finalizing our proposals to (a) accept the new CPT codes 97161-97168 for PT and OT evaluative procedures and (b) use the PAMA smoothing authority to value the PT and OT complexity level evaluations as groups of services rather than individually by assigning a work RVU of 1.2 to each complexity level. We are modifying our proposal for the valuation of the PT and OT re-evaluation codes and are finalizing a work RVU of 0.75 for each code. Lastly, we are finalizing the PE inputs as proposed.

iv. Always Therapy Codes

It is important to note that CMS defines the codes for these evaluative services as “always therapy.” This means that they always represent therapy services regardless of who performs them and always require a therapy modifier, GP or GO, to signify that the services are furnished under a PT or OT plan of care, respectively. These codes will also be subject to the therapy MPPR and to statutory therapy caps.

v. Potentially Misvalued Therapy Codes

Since 2010, in addition to the codes for evaluative services, CMS has periodically added codes that represent therapy services to the list of potentially misvalued codes. The current list of ten therapy codes was based on the statutory category “codes that account for the majority of spending under the physician fee schedule,” as specified in section 1848(c)(2)(K)(ii)(VII) of the Act. We understand that the therapy specialty organizations have pursued the development of coding changes through the CPT process for these modality and procedure services. While we understand that, in some cases, it may take several years to develop appropriate coding revisions, we are, in the meantime, seeking information regarding appropriate valuation for the existing codes. See Table 24.

Comment: We received multiple comments on our nomination of the ten therapy codes to the potentially misvalued code list. The PT and OT specialty societies each expressed concern that we issued the potentially misvalued code list knowing that they are currently working with the AMA Relativity Assessment Workgroup (RAW) to survey and submit CPT changes to certain intervention codes in the PMR family, including some codes on the misvalued code list. Nonetheless, the PT specialty society told us that it will work with the RUC (as the appropriate venue) this fall to survey and value the codes; but asked to meet with us in early 2017 to discuss their progress. The OT specialty society stated that it has already begun work with AMA to expedite valuation surveys for relevant codes, but also noted its intent to resume work with the RAW to replace some of the codes on the misvalued code list, including CPT code 97535, as soon as the misvalued code survey process is complete. In addition, the OT specialty society noted its belief that CMS staff attendance at the RAW condoned their timeline for proceeding with various PMR code revisions.

A few commenters believe the codes on the potentially misvalued code list are already valued correctly as the PE inputs for many therapy codes, including those defined by 15-minute intervals, have already been adjusted by the PEAC/RUC/HCPAC to account for efficiencies when billed with other therapy codes. Several commenters cautioned that any review must also consider that all of these codes are already subject to a 50 percent MPPR reduction. One commenter believes the work of CPT code 97140 is undervalued compared to other codes since it requires the more skilled therapist using manual techniques to touch the patient.

Response: We will include a valuation discussion during CY 2018 rulemaking of those codes for which we receive RUC recommendations by/at its February 2017 meeting.

(57) Valuation of Services Where Moderate Sedation is an Inherent Part of the Procedure and Valuation of Moderate Sedation Services (CPT Codes 99151, 99152, 99153, 99155, 99156, and 99157; and HCPCS Code G0500)

In the CY 2015 PFS proposed rule (79 FR 40349), we noted that it appeared that practice patterns for endoscopic procedures were changing. Anesthesia services are increasingly being separately reported for endoscopic procedures, meaning that resource costs associated with sedation were no longer incurred by the practitioner reporting the procedure. Subsequently, in the CY 2016 PFS proposed rule (80 FR 41707), we solicited public comment and recommendations on approaches to address the appropriate valuation of moderate sedation related to the approximately 400 diagnostic and therapeutic procedures for which the CPT Editorial Panel has determined that moderate sedation is an inherent part of furnishing the service. The CPT Editorial Panel created separate codes for reporting moderate sedation services (see Table 25).

For the newly created moderate sedation CPT codes, we proposed to use the RUC-recommended work RVUs for CPT codes 99151, 99152, 99155, and 99157. We stated in the CY 2017 proposed rule that CPT codes 99151 and 99152 make a distinction between moderate sedation services furnished to patients younger than 5 years of age and patients 5 years or older, with CPT codes 99155 and 99156 making a similar distinction. The RUC recommendations included a work RVU increment of 0.25 between CPT codes 99151 and 99152. For CPT code 99156, we proposed a work RVU of 1.65 to maintain the 0.25 increment relative to CPT code 99155 (a RUC-recommended work RVU of 1.90) and maintain relativity among the CPT codes in this family. We proposed to use the RUC-recommended direct PE inputs for all six codes.

We stated in the CY 2017 proposed rule that when moderate sedation is reported for Medicare beneficiaries, we expect that it would most frequently be reported using the code that describes moderate sedation furnished by the same person who also performs the primary procedure for patients 5 years of age or older. Under the new coding structure, these moderate sedation services would be reported using CPT code 99152, for which we proposed a work RVU of 0.25, consistent with the RUC recommendations for this code. Stakeholders presented information that illustrated that the specialty group survey data regarding the work involved in furnishing moderate sedation described by CPT code 99152 showed a significant bimodal distribution between procedural services furnished by gastroenterologists (GI) and those services furnished by other specialties. The GI societies' survey data reported a median valuation of 0.10 work RVUs for moderate sedation furnished by the same person furnishing the base procedure. Given the significant volume of moderate sedation furnished by GI practitioners and the significant difference in RVUs reported in the survey data, we proposed to make payment using a GI endoscopy-specific moderate sedation code (HCPCS code G0500) that would be used in lieu of the new CPT moderate sedation coding for use with other services.

• G0500: moderate sedation services provided by the same physician or other qualified health care professional performing a gastrointestinal endoscopic service (excluding biliary procedures) that sedation supports, requiring the presence of an independent trained observer to assist in the monitoring of the patient's level of consciousness and physiological status; initial 15 minutes of intra-service time; patient age 5 years or older.

We proposed to value HCPCS code G0500 at 0.10 work RVUs based on the median survey result for GI respondents in the survey data. We proposed that when moderate sedation services are furnished by the same practitioner reporting the GI endoscopy procedure, practitioners would report the sedation services using HCPCS code G0500 instead of CPT code 99152. In all other cases, we proposed that practitioners would report moderate sedation using one of the new moderate sedation CPT codes consistent with CPT guidance. This would include the full range of codes for those furnishing moderate sedation with the remaining (non-GI endoscopy) base procedures, as well as for the other circumstances during which moderate sedation is furnished along with a GI endoscopy (for example, to a patient under 5 years of age or for a biliary procedure, the endoscopist furnishing moderate sedation should not use HCPCS code G0500, but instead use the appropriate CPT code.

In addition to proposing work RVUs for the new codes used to separately report moderate sedation, we stated in the proposed rule that the RUC provided recommendations that valued the procedural services without moderate sedation. However, the RUC recommended removing fewer RVUs from the procedures than it recommended for valuing the moderate sedation services that were removed from the procedure codes. In other words, the RUC recommended that overall payments for these procedures should be increased now that practitioners would be required to report the sedation services that were previously included as inherent parts of the procedures. We stated in the proposed rule that we believe that if we were to use the RUC recommendations for revaluation of the procedural services without refinement, the RVUs currently attributable to the redundant payment for sedation services when anesthesia is separately reported would be used exclusively to increase overall payment for these services. We refer readers to section II.D.5. of this final rule, which includes a more extensive discussion of our general principle that overall resource costs for procedures that include moderate sedation do not inherently change based solely on changes in coding.

To account for the separate billing of moderate sedation services, we proposed to maintain current values for the procedure codes less the work RVUs associated with the most frequently reported corresponding moderate sedation code so that practitioners furnishing the moderate sedation services previously considered to be inherent in the procedure would have no change in overall work RVUs. Since we proposed 0.10 work RVUs for moderate sedation for the GI endoscopy procedures, we proposed a corresponding 0.10 reduction in work RVUs for these same procedures. For all other Appendix G procedures that currently include moderate sedation as an inherent part of the procedure, we proposed to remove 0.25 work RVUs from the current values.

We received 22 comments from medical professionals, ambulatory surgical centers (ASCs), manufacturers, and professional medical specialty societies representing radiation oncology, brachytherapy, colon and rectal surgeons, certified registered nurse anesthetists (CRNAs), pediatrics, cardiology, thorasic surgery, general surgery, gastroenterology, emergency medicine, interventional radiology, and vascular surgery. Commenters were generally supportive of CMS' proposals related to valuation of the new moderate sedation codes. A few commenters disagreed with our proposed refinements for one of the new moderate sedation CPT codes. While most commenters were supportive of CMS' proposal to use a methodology to revalue the procedural services without moderate sedation, some commenters suggested that we should revalue certain procedures differently (for example, apply a lower work RVU reduction or make no reduction). A few commenters were opposed to separate reporting of moderate sedation and suggested alternatives for CMS to consider. Our responses to commenters' specific issues are included below.

Comment: Many commenters expressed support for CMS' proposal to accept the RUC's recommendations for new moderate sedation CPT codes 99151, 99152, 99153, 99155, and 99157. Several commenters, including the RUC and medical specialty societies, disagreed with CMS' proposal to value CPT code 99156 at 1.65 work RVUs. Commenters requested that CMS finalize the RUC-recommended work RVU of 1.84 (the 25th percentile survey result). Commenters stated that there were clinical differences in the typical patients that receive services that would be reported using CPT code 99156, disagreeing with CMS' proposal to reduce the work RVU for CPT code 99156 to maintain relativity among the code pairs in this family. Commenters suggested that CPT code 99156 would be used to report moderate sedation services that are currently reported using CPT code 99149. Commenters stated that CPT code 99149 was typically performed in the emergency department (approximately 58 percent of the time), indicating that the typical patient is either acutely ill or injured, and that moderate sedation services are typically performed without support staff, which commenters suggested further justified a work RVU of 1.84 for CPT code 99156.

Response: The code descriptors for each of the new moderate sedation CPT codes make distinctions between the ages of the patients and the clinical staff involved in furnishing the moderate sedation services. The typical patient vignettes used in the specialty societies' surveys did not indicate clinical differences between patients receiving moderate sedation services reported using CPT code 99156 compared to services reported with CPT code 99155. Additionally, the typical patient vignettes for CPT codes 99151 and 99152 did not indicate clinical differences in the patients. We continue to believe that the work RVU increment of 0.25 should be maintained between CPT codes 99155 and 99156 since these codes have the same younger than age 5/older than age 5 dynamic as described by CPT codes 99151 and 99152.

Therefore, for CY 2017, we are finalizing work RVUs for the moderate sedation codes as follows:

• Work RVU of 0.50 for CPT code 99151;
• Work RVU of 0.25 for CPT code 99152;
• Work RVU of 1.90 for CPT code 99155;
• Work RVU of 1.65 for CPT code 99156; and
• Work RVU of 1.25 for CPT code 99157.

We note that CPT code 99153 is a PE-only code and we are finalizing the proposed PE inputs for CPT code 99153, as well as finalizing the proposed PE inputs for all other codes in this family without modification.

Comment: While many commenters supported use of a methodological approach to revaluing Appendix G procedural services, some commenters disagreed with CMS' proposed refinements to the RUC's recommended methodology. Commenters suggested that the RUC's approach was consistent with how the services were originally valued and was budget neutral within these services. The RUC, along with several commenters representing specialty medical societies, requested that CMS use the same RUC-approved two-tier methodology for removing work RVUs associated with the work of moderate sedation from Appendix G services based on whether the code was assigned to one of two preservice time packages used by the RUC in developing recommendations. Using the same two-tier methodology, the RUC suggested removal of 0.10 work RVUs for some GI services and 0.19 work RVUs from other GI services, depending on the RUC's assignment of pre-service time.

Response: We understand that some stakeholders would prefer that we use the RUC's recommendations so that the RVUs currently attributable to the redundant payment for sedation services when anesthesia is separately reported would be used exclusively to increase overall payment for these services. We also understand that the RUC assumes that the amount of pre-service time for particular services may reflect a different level of preparation required for sedation services. However, we continue to believe that the overall resource costs for the procedures including moderate sedation do not inherently change based solely on changes in coding, so we do not believe that our assignment of overall work RVUs should increase in cases where the moderate sedation is performed by the proceduralists, as previously assumed. Therefore, we believe that a different amount of work RVUs should be removed from the Appendix G services only in cases where the typical moderate sedation code also has a different amount of assigned work RVUs, such as the case with codes that would be reported with G0500. In other words, we believe that there should be a direct relationship, for each code, between the work RVUs attributable to moderate sedation, regardless of whether it is automatically included in payment for a given procedure (at current) or separately reported (as proposed).

Comment: A few commenters expressed concerns that the proposed revaluation methodology would disturb the relativity of many of the Appendix G codes, along with the increasing administrative burden by requiring separate reporting of the procedural and moderate sedation services. Other commenters suggested that CMS consider alternatives including only addressing revaluation of Appendix G services where moderate sedation is no longer inherent or only those procedural services reported with separate anesthesia services the majority of the time.

Response: We appreciate the concerns of commenters regarding both the issues of relativity within families of codes, as well as concerns regarding administrative burden. However, we believe that it serves relativity to maintain the overall work RVUs for each of the services when reported with moderate sedation, which would be typical for many of these codes. While we understand the value in reducing the number of codes required to be reported for payment under the PFS, we also believe that it is important that the coding be granular enough to allow us to identify which services are furnished to Medicare beneficiaries by which practitioners. It is also clear to us that the accuracy of the assumption of moderate sedation as inherent for particular procedures may change over time, as we have seen reflected in the claims data. We do not believe that a shifting set of services where moderate sedation values are alternatively included or not included in the valuation of particular codes based on annual analysis of claims data would be likely to be administratively easier for practitioners.

Comment: A few commenters requested that CMS not finalize its proposal to reduce the work RVUs for certain procedures. Some commenters indicated that certain codes identified in Appendix G were valued before Appendix G was established, or the work of moderate sedation was not included in the valuation of certain procedures.

Response: We appreciate the commenters' feedback regarding our proposals. We remind stakeholders that the potentially misvalued code process is intended to improve the accuracy of the RVUs assigned to particular codes. We welcome feedback from interested individuals, stakeholders, and specialty societies regarding the valuation of specific codes for consideration in future rulemaking.

Comment: A few commenters stated that CMS did not provide a rationale to support that moderate sedation furnished with GI endoscopy services required less work than moderate sedation furnished with other Appendix G procedures.

Response: Our proposal was based on the GI societies' survey data included in the RUC recommendations that reported a median valuation of 0.10 work RVUs for moderate sedation furnished by the same person furnishing the base procedure.

Comment: A few commenters suggested that creation of HCPCS code G0500 would cause confusion among practitioners since the new CPT codes developed to report moderate sedation do not differentiate between GI and non-GI procedures. One commenter stated that HCPCS code G0500 is time based, and therefore, to report the code, at least 50 percent of the time (7.5 minutes) is required, but the GI subset of data that CMS accepted to create the HCPCS code G0500 indicates an intraservice time of 5 minutes. The commenter went on to state that it would appear that a majority of the GI endoscopists would never be able to report HCPCS code G0500.

Response: We expect that practitioners will report the appropriate CPT or HCPCS code that most accurately describes the services performed during a patient encounter, including those services performed concurrently and in support of a procedural service consistent with CPT guidance. We note that the commenter refers to the time for moderate sedation in the survey data, while the time thresholds for the moderate sedation codes are intended to match the intraservice time of the procedure itself. We reviewed the intraservice time assumptions for the procedure codes, and only one includes an intraservice time as low as 7.5 minutes and none lower. Table 26 identifies the GI endoscopic services for which HCPCS code G0500 will be used to report moderate sedation services (available in the “downloads” section of the PFS Web site at http://www.cms.gov/Medicare/Medicare-Feefor-Service-Payment/PhysicianFeeSched/PFS-Federal-Regulation-Notices.html).

Comment: Several commenters disagreed with CMS' proposal to revalue Appendix G esophageal dilation, biliary endoscopy, and ERCP procedures minus 0.25 work RVUs instead of minus 0.10 work RVUs, similar to other endoscopy services identified in Appendix G. Commenters requested that CMS only reduce these procedural services with a 0.10 work RVU reduction, and allow reporting of moderate sedation using HCPCS code G0500, similar to other endoscopy procedures identified in Appendix G.

Response: While we continue to believe that the moderate sedation work for Appendix G esophageal dilation, biliary endoscopy, and ERCP procedures is more extensive than for other endoscopy procedures identified in Appendix G, for CY 2017, after considering the comments, we are finalizing a revaluation of certain esophageal dilation, biliary endoscopy, and ERCP procedures minus 0.10 work RVUs instead of the 0.25 work RVU reduction as proposed (see Table 26 for additional information). We will continue to monitor claims data related to separately billed anesthesia services performed in conjunction with these procedures to inform future rulemaking related to the valuation of these codes. We are also modifying the code descriptor for HCPCS code G0500 to reflect these changes. Therefore, we are finalizing the descriptor for HCPCS code G0500 as:

• G0500: Moderate sedation services provided by the same physician or other qualified health care professional performing a gastrointestinal endoscopic service that sedation supports, requiring the presence of an independent trained observer to assist in the monitoring of the patient's level of consciousness and physiological status; initial 15 minutes of intra-service time; patient age 5 years or older. (additional time may be reported with 99153, as appropriate).

Comment: Some commenters requested that CMS provide practitioners and providers with instructions on use of the newly created moderate sedation codes, allow for additional time to implement the coding changes, and provide MACs appropriate claims processing instructions specific to these codes.

Response: We plan to issue appropriate claims processing instructions to the local MACs. We do not believe that an implementation delay is necessary since the new CPT and HCPCS codes will be effective January 1, 2017 and available for use by practitioners and providers at that time.

In summary, after consideration of the comments, we are finalizing our proposed modifications to maintain the current values for the procedure codes less the work RVUs associated with the most frequently reported corresponding moderate sedation code. Practitioners furnishing the moderate sedation services previously considered to be inherent in the procedure will have no change in overall work RVUs. Since we are finalizing a work RVU of 0.10 (HCPCS code G0500) for moderate sedation for the GI endoscopy procedures, we are finalizing a corresponding 0.10 reduction in work RVUs for the corresponding procedural services. For all other Appendix G procedures that currently include moderate sedation as an inherent part of the procedure, we are finalizing a 0.25 work RVU reduction from the current values.

Table 26 lists the CY 2016 work RVUs for each applicable service and our proposed and final CY 2017 refined work RVUs using the finalized revaluation methodology described above. Additionally, the table identifies the GI endoscopic services for which HCPCS code G0500 will be used to report moderate sedation services (available in the “downloads” section of the PFS Web site at http://www.cms.gov/Medicare/Medicare-Feefor-Service-Payment/PhysicianFeeSched/PFS-Federal-Regulation-Notices.html).