Medicare Program; Prospective Payment System and Consolidated Billing for Skilled Nursing Facilities for FY 2018, SNF Value-Based Purchasing Program, SNF Quality Reporting Program, Survey Team Composition, and Correction of the Performance Period for the NHSN HCP Influenza Vaccination Immunization Reporting Measure in the ESRD QIP for PY 2020

AGENCY:

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

ACTION:

Final rule.

SUMMARY:

This final rule updates the payment rates used under the prospective payment system (PPS) for skilled nursing facilities (SNFs) for fiscal year (FY) 2018. It also revises and rebases the market basket index by updating the base year from 2010 to 2014, and by adding a new cost category for Installation, Maintenance, and Repair Services. The rule also finalizes revisions to the SNF Quality Reporting Program (QRP), including measure and standardized resident assessment data policies and policies related to public display. In addition, it finalizes policies for the Skilled Nursing Facility Value-Based Purchasing Program that will affect Medicare payment to SNFs beginning in FY 2019. The final rule also clarifies the regulatory requirements for team composition for surveys conducted for investigating a complaint and aligns regulatory provisions for investigation of complaints with the statutory requirements. The final rule also finalizes the performance period for the National Healthcare Safety Network (NHSN) Healthcare Personnel (HCP) Influenza Vaccination Reporting Measure included in the End-Stage Renal Disease (ESRD) Quality Incentive Program (QIP) for Payment Year 2020.

DATES:

These regulations are effective on October 1, 2017.

FOR FURTHER INFORMATION CONTACT:

Penny Gershman, (410) 786-6643, for information related to SNF PPS clinical issues.

John Kane, (410) 786-0557, for information related to the development of the payment rates and case-mix indexes.

Kia Sidbury, (410) 786-7816, for information related to the wage index.

Bill Ullman, (410) 786-5667, for information related to level of care determinations, consolidated billing, and general information.

Michelle King, (410) 786-3667, for information related to skilled nursing facility quality reporting program.

James Poyer, (410) 786-2261, for information related to the skilled nursing facility value-based purchasing program.

Delia Houseal, (410) 786-2724, for information related to the end-stage renal disease quality incentive program.

Rebecca Ward, (410) 786-1732 and Caecilia Blondiaux, (410) 786-2190, for survey type definitions.

SUPPLEMENTARY INFORMATION:

Availability of Certain Tables Exclusively Through the Internet on the CMS Web site

As discussed in the FY 2014 SNF PPS final rule (78 FR 47936), tables setting forth the Wage Index for Urban Areas Based on CBSA Labor Market Areas and the Wage Index Based on CBSA Labor Market Areas for Rural Areas are no longer published in the Federal Register.

Instead, these tables are available exclusively through the Internet on the CMS Web site. The wage index tables for this final rule can be accessed on the SNF PPS Wage Index home page, at http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/SNFPPS/WageIndex.html.

Readers who experience any problems accessing any of these online SNF PPS wage index tables should contact Kia Sidbury at (410) 786-7816.

To assist readers in referencing sections contained in this document, we are providing the following Table of Contents.

Table of Contents

I. Executive Summary

II. Background on SNF PPS

A. Statutory Basis and Scope

B. Initial Transition for the SNF PPS

C. Required Annual Rate Updates

III. Analysis and Responses to Public Comments on the FY 2018 SNF PPS Proposed Rule

A. General Comments on the FY 2018 SNF PPS Proposed Rule

B. SNF PPS Rate Setting Methodology and FY 2018 Update

1. Federal Base Rates

2. SNF Market Basket Update

3. Case-Mix Adjustment

4. Wage Index Adjustment

5. Adjusted Rate Computation Example

C. Additional Aspects of the SNF PPS

1. SNF Level of Care—Administrative Presumption

2. Consolidated Billing

3. Payment for SNF-Level Swing-Bed Services

D. Other Issues

1. Revising and Rebasing the SNF Market Basket Index

2. Skilled Nursing Facility (SNF) Quality Reporting Program (QRP)

3. Skilled Nursing Facility Value-Based Purchasing Program (SNF VBP)

4. Survey Team Composition

5. Correction of the Performance Period for the National Healthcare Safety Network (NHSN) Healthcare Personnel (HCP) Influenza Vaccination Immunization Reporting Measure in the End-Stage Renal Disease (ESRD) Quality Incentive Program (QIP) for Payment Year (PY) 2020

IV. Collection of Information Requirements

V. Economic Analyses

Regulation Text

Acronyms

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

AIDS Acquired Immune Deficiency Syndrome

ALJ Administrative Law Judge

ARD Assessment reference date

BBA Balanced Budget Act of 1997, Public Law 105-33

BBRA Medicare, Medicaid, and SCHIP Balanced Budget Refinement Act of 1999, Public Law 106-113

BIPA Medicare, Medicaid, and SCHIP Benefits Improvement and Protection Act of 2000, Public Law 106-554

CAH Critical access hospital

CARE Continuity Assessment Record and Evaluation

CASPER Certification and Survey Provider Enhanced Reporting

CBSA Core-based statistical area

CCN CMS Certification Number

CFR Code of Federal Regulations

CMI Case-mix index

CMS Centers for Medicare & Medicaid Services

DTI Deep tissue injuries

FFS Fee-for-service

FR Federal Register

FY Fiscal year

HCPCS Healthcare Common Procedure Coding System

HIQR  Hospital Inpatient Quality Reporting

HOQR Hospital Outpatient Quality Reporting

HRRP Hospital Readmissions Reduction Program

HVBP  Hospital Value-Based Purchasing

ICD-10-CM International Classification of Diseases, 10th Revision, Clinical Modification

IGI IHS Global Inc.

IMPACT Improving Medicare Post-Acute Care Transformation Act of 2014, Public Law 113-185

IPPS Inpatient prospective payment system

IRF Inpatient Rehabilitation Facility

IRF-PAI Inpatient Rehabilitation Facility Patient Assessment Instrument

LTC Long-term care

LTCH Long-term care hospital

MACRA Medicare Access and CHIP Reauthorization Act of 2015, Public Law 114-10

MAP Measures Application Partnership

MDS Minimum data set

MFP Multifactor productivity

MMA Medicare Prescription Drug, Improvement, and Modernization Act of 2003, Public Law 108-173

MSA Metropolitan statistical area

NF Nursing facility

NQF National Quality Forum

OASIS Outcome and Assessment Information Set

OBRA 87 Omnibus Budget Reconciliation Act of 1987, Public Law 100-203

OMB Office of Management and Budget

PAC Post-acute care

PAMA Protecting Access to Medicare Act of 2014, Public Law 113-93

PPS Prospective Payment System

PQRS Physician Quality Reporting System

QIES Quality Improvement and Evaluation System

QIES ASAP Quality Improvement and Evaluation System Assessment Submission and Processing

QRP Quality Reporting Program

RAI Resident assessment instrument

RAVEN Resident assessment validation entry

RFA Regulatory Flexibility Act, Public Law 96-354

RIA Regulatory impact analysis

RUG-III Resource Utilization Groups, Version 3

RUG-IV Resource Utilization Groups, Version 4

RUG-53 Refined 53-Group RUG-III Case-Mix Classification System

SCHIP State Children's Health Insurance Program

SNF Skilled nursing facility

SNF PMR Skilled Nursing Facility Payment Models Research

SNF QRP Skilled Nursing Facility Quality Reporting Program

SNF VBP Skilled Nursing Facility Value-Based Purchasing Program

SNFPPR Skilled Nursing Facility Potentially Preventable Readmission Measure

SNFRM Skilled Nursing Facility 30-Day All-Cause Readmission Measure

STM Staff time measurement

STRIVE Staff time and resource intensity verification

TEP Technical expert panel

UMRA Unfunded Mandates Reform Act, Public Law 104-4

VBP Value-based purchasing

I. Executive Summary

A. Purpose

This final rule updates the SNF prospective payment rates for FY 2018 as required under section 1888(e)(4)(E) of the Social Security Act (the Act). It also responds to section 1888(e)(4)(H) of the Act, which requires the Secretary to provide for publication in the Federal Register, before the August 1 that precedes the start of each fiscal year (FY), certain specified information relating to the payment update (see section II.C. of this final rule). This final rule also finalizes updates to the requirements for the Skilled Nursing Facility Quality Reporting Program (SNF QRP), additional policies for the Skilled Nursing Facility Value-Based Purchasing Program (SNF VBP), and clarification of requirements related to survey team composition and investigation of complaints under §§ 488.30, 488.301, 488.308, and 488.314. The final rule also finalizes one proposal related to the performance period for the National Healthcare Safety Network (NHSN) Healthcare Personnel (HCP) Influenza Vaccination Reporting Measure included in the End-Stage Renal Disease (ESRD) Quality Incentive Program (QIP).

B. Summary of Major Provisions

In accordance with sections 1888(e)(4)(E)(ii)(IV) and 1888(e)(5) of the Act, the federal rates in this final rule reflect an update to the rates that we published in the SNF PPS final rule for FY 2017 (81 FR 51970), which reflects the SNF market basket update, as required by section 1888(e)(5)(B)(iii) of the Act for FY 2018. Additionally, in section III.B.1. of this final rule, we are finalizing our proposal to revise and rebase the market basket index for FY 2018 and subsequent FYs by updating the base year from 2010 to 2014, and by adding a new cost category for Installation, Maintenance, and Repair Services. We are also finalizing additional polices, measures and data reporting requirements for the Skilled Nursing Facility Quality Reporting Program (SNF QRP) and requirements for the SNF VBP Program, including an exchange function to translate SNF performance scores calculated using the program's scoring methodology into value-based incentive payments.

We are also clarifying the regulatory requirements for team composition for surveys conducted for the purposes of investigating a complaint and on-site monitoring of compliance, and to align the regulatory provisions for special surveys and investigation of complaints with the statute. The changes clarify that the requirement for an interdisciplinary team that must include a registered nurse is applicable to surveys conducted under sections 1819(g)(2) and 1919(g)(2) of the Act, and not to those surveys conducted to investigate complaints or to monitor compliance on-site under sections 1819(g)(4) and 1919(g)(4) of the Act. Revising the regulatory language under §§ 488.30, 488.301, 488.308, and 488.314 to correspond to the statutory requirements found in sections 1819(g) and 1919(g) of the Act will add clarity to these requirements by making them more explicit. We are also revising the performance period for the National Healthcare Safety Network (NHSN) Healthcare Personnel (HCP) Influenza Vaccination Reporting Measure included in the End-Stage Renal Disease (ESRD) Quality Incentive Program (QIP) for PY 2020.

C. Summary of Cost and Benefits

II. Background on SNF PPS

A. Statutory Basis and Scope

As amended by section 4432 of the Balanced Budget Act of 1997 (BBA, Pub. L. 105-33, enacted on August 5, 1997), section 1888(e) of the Act provides for the implementation of a PPS for SNFs. This methodology uses prospective, case-mix adjusted per diem payment rates applicable to all covered SNF services defined in section 1888(e)(2)(A) of the Act. The SNF PPS is effective for cost reporting periods beginning on or after July 1, 1998, and covers all costs of furnishing covered SNF services (routine, ancillary, and capital-related costs) other than costs associated with approved educational activities and bad debts. Under section 1888(e)(2)(A)(i) of the Act, covered SNF services include post-hospital extended care services for which benefits are provided under Part A, as well as those items and services (other than a small number of excluded services, such as physicians' services) for which payment may otherwise be made under Part B and which are furnished to Medicare beneficiaries who are residents in a SNF during a covered Part A stay. A comprehensive discussion of these provisions appears in the May 12, 1998 interim final rule (63 FR 26252). In addition, a detailed discussion of the legislative history of the SNF PPS is available online at http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/SNFPPS/Downloads/Legislative_History_04152015.pdf.

Section 215(a) of the Protecting Access to Medicare Act of 2014 (PAMA, Pub. L. 113-93, enacted on April 1, 2014) added a new section 1888(g) to the Act, which requires the Secretary to specify an all-cause all-condition hospital readmission measure and an all-condition risk-adjusted potentially preventable hospital readmission measure for the SNF setting. Additionally, section 215(b) of PAMA added a new section 1888(h) to the Act, which requires the Secretary to implement a VBP program for SNFs. Finally, section 2(a) of the Improving Medicare Post-Acute Care Transformation Act of 2014 (IMPACT Act, Pub. L. 113-185, enacted on October 6, 2014) added a new section 1899B to the Act that, among other things, requires SNFs to report standardized resident assessment data, data on quality measures, and data on resource use and other measures. In addition, section 2(c)(4) of the IMPACT Act added a new section 1888(e)(6) to the Act, which requires the Secretary to implement a quality reporting program for SNFs.

B. Initial Transition for the SNF PPS

Under sections 1888(e)(1)(A) and 1888(e)(11) of the Act, the SNF PPS included an initial, three-phase transition that blended a facility-specific rate (reflecting the individual facility's historical cost experience) with the federal case-mix adjusted rate. The transition extended through the facility's first 3 cost reporting periods under the PPS, up to and including the one that began in FY 2001. Thus, the SNF PPS is no longer operating under the transition, as all facilities have been paid at the full federal rate effective with cost reporting periods beginning in FY 2002. As we now base payments for SNFs entirely on the adjusted federal per diem rates, we no longer include adjustment factors under the transition related to facility-specific rates for the upcoming FY.

C. Required Annual Rate Updates

Section 1888(e)(4)(E) of the Act requires the SNF PPS payment rates to be updated annually. The most recent annual update occurred in a final rule that set forth updates to the SNF PPS payment rates for FY 2017 (81 FR 51970, August 5, 2016). Section 1888(e)(4)(H) of the Act specifies that we provide for publication annually in the Federal Register of the following:

• The unadjusted federal per diem rates to be applied to days of covered SNF services furnished during the upcoming FY.
• The case-mix classification system to be applied for these services during the upcoming FY.
• The factors to be applied in making the area wage adjustment for these services.

Along with other revisions discussed later in this preamble, this final rule provides the required annual updates to the per diem payment rates for SNFs for FY 2018.

III. Analysis and Responses to Public Comments on the FY 2018 SNF PPS Proposed Rule

In response to the publication of the FY 2018 SNF PPS proposed rule, we received 247 public comments from individuals, providers, corporations, government agencies, private citizens, trade associations, and major organizations. The following are brief summaries of each proposed provision, a summary of the public comments that we received related to that proposal, and our responses to the comments.

A. General Comments on the FY 2018 SNF PPS Proposed Rule

In addition to the comments we received on specific proposals contained within the proposed rule (which we address later in this final rule), commenters also submitted the following, more general, observations on the SNF PPS and SNF care generally. A discussion of these comments, along with our responses, appears below.

Comment: One commenter requested that we instruct the Medicare Administrative Contractors to refrain from denying coverage and payment for SNF Part B claims for physiatrists visiting residents in SNFs. The commenter goes on to state their concerns regarding the potential for variability in coverage across contractors.

Response: With regard to our instructing the contractors to refrain from denying coverage or payment for SNF claims related to physiatrists visits under Part B, this comment is outside the scope of this final rule. However, we will forward these comments to the appropriate division within CMS for consideration. With regard to the potential for variability among contractors, we will continue to educate the contractors to ensure compliance with all federal guidance and regulations.

Comment: One commenter requested that we consider including recreational therapy time provided to SNF residents by recreational therapists as part of the calculation of the resident's RUG-IV therapy classification or as part of determining the number of restorative nursing services provided to the resident.

Response: We appreciate the commenter raising this issue, but we do not believe there is sufficient evidence at this time regarding the efficacy of recreational therapy interventions or, more notably, data which would substantiate a determination of the effect on payment of such interventions, as such services were not considered separately, as were physical, occupational and speech-language pathology services, when RUG-IV was being developed. That being said, we would note that Medicare Part A originally paid for institutional care in various provider settings, including SNF, on a reasonable cost basis, but now makes payment using PPS methodologies, such as the SNF PPS. To the extent that one of these SNFs furnished recreational therapy to its inpatients under the previous, reasonable cost methodology, the cost of the services would have been included in the base payments when SNF PPS payment rates were derived. Under the PPS methodology, Part A makes a comprehensive payment for the bundled package of items and services that the facility furnishes during the course of a Medicare-covered stay. This package encompasses nearly all services that the beneficiary receives during the course of the stay—including any medically necessary recreational therapy—and payment for such services is included within the facility's comprehensive SNF PPS payment for the covered Part A stay itself.

B. SNF PPS Rate Setting Methodology and FY 2018 Update

1. Federal Base Rates

Under section 1888(e)(4) of the Act, the SNF PPS uses per diem federal payment rates based on mean SNF costs in a base year (FY 1995) updated for inflation to the first effective period of the PPS. We developed the federal payment rates using allowable costs from hospital-based and freestanding SNF cost reports for reporting periods beginning in FY 1995. The data used in developing the federal rates also incorporated a Part B add-on, which is an estimate of the amounts that, prior to the SNF PPS, would have been payable under Part B for covered SNF services furnished to individuals during the course of a covered Part A stay in a SNF.

In developing the rates for the initial period, we updated costs to the first effective year of the PPS (the 15-month period beginning July 1, 1998) using a SNF market basket index, and then standardized for geographic variations in wages and for the costs of facility differences in case mix. In compiling the database used to compute the federal payment rates, we excluded those providers that received new provider exemptions from the routine cost limits, as well as costs related to payments for exceptions to the routine cost limits. Using the formula that the BBA prescribed, we set the federal rates at a level equal to the weighted mean of freestanding costs plus 50 percent of the difference between the freestanding mean and weighted mean of all SNF costs (hospital-based and freestanding) combined. We computed and applied separately the payment rates for facilities located in urban and rural areas, and adjusted the portion of the federal rate attributable to wage-related costs by a wage index to reflect geographic variations in wages.

2. SNF Market Basket Update

a. SNF Market Basket Index

Section 1888(e)(5)(A) of the Act requires us to establish a SNF market basket index that reflects changes over time in the prices of an appropriate mix of goods and services included in covered SNF services. Accordingly, we have developed a SNF market basket index that encompasses the most commonly used cost categories for SNF routine services, ancillary services, and capital-related expenses. In the SNF PPS final rule for FY 2014 (78 FR 47939 through 47946), we revised and rebased the market basket index, which included updating the base year from FY 2004 to FY 2010. For FY 2018, as discussed in section III.D.1. of this final rule, we are rebasing and revising the SNF market basket, updating the base year from FY 2010 to 2014.

The SNF market basket index is used to compute the market basket percentage change that is used to update the SNF federal rates on an annual basis, as required by section 1888(e)(4)(E)(ii)(IV) of the Act. This market basket percentage update is adjusted by a forecast error correction, if applicable, and then further adjusted by the application of a productivity adjustment as required by section 1888(e)(5)(B)(ii) of the Act and described in section III.B.2.d. of this final rule. For FY 2018, the growth rate of the 2014-based SNF market basket is estimated to be 2.6 percent, which is based on the IHS Global Inc. (IGI) second quarter 2017 forecast with historical data through first quarter 2017.

However, we note that section 411(a) of the Medicare Access and CHIP Reauthorization Act of 2015 (MACRA, Pub. L. 114-10, enacted on April 16, 2015) amended section 1888(e) of the Act to add section 1888(e)(5)(B)(iii) of the Act. Section 1888(e)(5)(B)(iii) of the Act establishes a special rule for FY 2018 that requires the market basket percentage, after the application of the productivity adjustment, to be 1.0 percent. In accordance with section 1888(e)(5)(B)(iii) of the Act, we will use a market basket percentage of 1.0 percent to update the federal rates set forth in this final rule. In section III.B.2.e. of this final rule, we discuss the specific application of the MACRA-specified market basket adjustment to the forthcoming annual update of the SNF PPS payment rates. In addition, in section III.D.2. of this final rule, we discuss the 2 percent reduction applied to the market basket update for those SNFs that fail to submit measures data as required by section 1888(e)(6)(A) of the Act.

b. Use of the SNF Market Basket Percentage

Section 1888(e)(5)(B) of the Act defines the SNF market basket percentage as the percentage change in the SNF market basket index from the midpoint of the previous FY to the midpoint of the current FY. Absent the addition of section 1888(e)(5)(B)(iii) of the Act, added by section 411(a) of MACRA, we would have used the percentage change in the SNF market basket index to compute the update factor for FY 2018. Based on the revision and rebasing of the SNF market basket discussed in section III.D.1. of this final rule, this factor is based on the IGI second quarter 2017 forecast (with historical data through the first quarter 2017) of the FY 2018 percentage increase in the 2014-based SNF market basket index reflecting routine, ancillary, and capital-related expenses. As discussed in sections III.B.2.c. and III.B.2.d. of this final rule, this market basket percentage change would have been reduced by the applicable forecast error correction (as described in § 413.337(d)(2)) and by the MFP adjustment as required by section 1888(e)(5)(B)(ii) of the Act. As noted previously, section 1888(e)(5)(B)(iii) of the Act, added by section 411(a) of the MACRA, requires us to use a 1.0 percent market basket percentage instead of the estimated 2.6 percent market basket percentage, adjusted as described below, to adjust the SNF PPS federal rates for FY 2018. Additionally, as discussed in section II.B. of this final rule, we no longer compute update factors to adjust a facility-specific portion of the SNF PPS rates, because the initial three-phase transition period from facility-specific to full federal rates that started with cost reporting periods beginning in July 1998 has expired.

c. Forecast Error Adjustment

As discussed in the June 10, 2003 supplemental proposed rule (68 FR 34768) and finalized in the August 4, 2003 final rule (68 FR 46057 through 46059), § 413.337(d)(2) provides for an adjustment to account for market basket forecast error. The initial adjustment for market basket forecast error applied to the update of the FY 2003 rate for FY 2004, and took into account the cumulative forecast error for the period from FY 2000 through FY 2002, resulting in an increase of 3.26 percent to the FY 2004 update. Subsequent adjustments in succeeding FYs take into account the forecast error from the most recently available FY for which there is final data, and apply the difference between the forecasted and actual change in the market basket when the difference exceeds a specified threshold. We originally used a 0.25 percentage point threshold for this purpose; however, for the reasons specified in the FY 2008 SNF PPS final rule (72 FR 43425, August 3, 2007), we adopted a 0.5 percentage point threshold effective for FY 2008 and subsequent FYs. As we stated in the final rule for FY 2004 that first issued the market basket forecast error adjustment (68 FR 46058, August 4, 2003), the adjustment will reflect both upward and downward adjustments, as appropriate.

For FY 2016 (the most recently available FY for which there is final data), the estimated increase in the market basket index was 2.3 percentage points, while the actual increase for FY 2016 was 2.3 percentage points, resulting in the actual increase being the same as the estimated increase. Accordingly, as the difference between the estimated and actual amount of change in the market basket index does not exceed the 0.5 percentage point threshold, the FY 2018 market basket percentage change of 2.6 percent would not have been adjusted to account for the forecast error correction. Table 1 shows the forecasted and actual market basket amounts for FY 2016.

d. Multifactor Productivity Adjustment

Section 1888(e)(5)(B)(ii) of the Act, as added by section 3401(b) of the Patient Protection and Affordable Care Act (Affordable Care Act, Pub. L. 111-148, enacted on March 23, 2010) requires that, in FY 2012 and in subsequent FYs, the market basket percentage under the SNF PPS (as described in section 1888(e)(5)(B)(i) of the Act) is to be reduced annually by the multifactor productivity (MFP) adjustment described in section 1886(b)(3)(B)(xi)(II) of the Act. Section 1886(b)(3)(B)(xi)(II) of the Act, in turn, defines the MFP adjustment to be equal to the 10-year moving average of changes in annual economy-wide private nonfarm business multi-factor productivity (as projected by the Secretary for the 10-year period ending with the applicable FY, year, cost-reporting period, or other annual period). The Bureau of Labor Statistics (BLS) is the agency that publishes the official measure of private nonfarm business MFP. We refer readers to the BLS Web site at http://www.bls.gov/mfp for the BLS historical published MFP data.

MFP is derived by subtracting the contribution of labor and capital inputs growth from output growth. The projections of the components of MFP are currently produced by IGI, a nationally recognized economic forecasting firm with which CMS contracts to forecast the components of the market baskets and MFP. To generate a forecast of MFP, IGI replicates the MFP measure calculated by the BLS, using a series of proxy variables derived from IGI's U.S. macroeconomic models. For a discussion of the MFP projection methodology, we refer readers to the FY 2012 SNF PPS final rule (76 FR 48527 through 48529) and the FY 2016 SNF PPS final rule (80 FR 46395). A complete description of the MFP projection methodology is available on our Web site at http://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/MedicareProgramRatesStats/MarketBasketResearch.html.

(1) Incorporating the MFP Adjustment Into the Market Basket Update

Per section 1888(e)(5)(A) of the Act, the Secretary shall establish a SNF market basket index that reflects changes over time in the prices of an appropriate mix of goods and services included in covered SNF services. Section 1888(e)(5)(B)(ii) of the Act, added by section 3401(b) of the Affordable Care Act, requires that for FY 2012 and each subsequent FY, after determining the market basket percentage described in section 1888(e)(5)(B)(i) of the Act, the Secretary shall reduce such percentage by the productivity adjustment described in section 1886(b)(3)(B)(xi)(II) of the Act (which we refer to as the MFP adjustment). Section 1888(e)(5)(B)(ii) of the Act further states that the reduction of the market basket percentage by the MFP adjustment may result in the market basket percentage being less than zero for a FY, and may result in payment rates under section 1888(e) of the Act being less than such payment rates for the preceding fiscal year.

If not for the enactment of section 411(a) of the MACRA, the FY 2018 update would include a calculation of the MFP adjustment as the 10-year moving average of changes in MFP for the period ending September 30, 2018, which is estimated to be 0.6 percent. Also, if not for the enactment of section 411(a) of the MACRA, consistent with section 1888(e)(5)(B)(i) of the Act and § 413.337(d)(2), the market basket percentage for FY 2018 for the SNF PPS would be based on IGI's second quarter 2017 forecast of the SNF market basket update, which is estimated to be 2.6 percent. In accordance with section 1888(e)(5)(B)(ii) of the Act (as added by section 3401(b) of the Affordable Care Act) and § 413.337(d)(3), this market basket percentage would then be reduced by the MFP adjustment (the 10-year moving average of changes in MFP for the period ending September 30, 2018) of 0.6 percent, which would be calculated as described above and based on IGI's second quarter 2017 forecast. Absent the enactment of section 411(a) of MACRA, the resulting MFP-adjusted SNF market basket update would have been equal to 2.0 percent, or 2.6 percent less 0.6 percentage point. However, as discussed above, section 1888(e)(5)(B)(iii) of the Act, added by section 411(a) of the MACRA, requires us to apply a 1.0 percent positive market basket adjustment in determining the FY 2018 SNF payment rates set forth in this final rule, without regard to the market basket update as adjusted by the MFP adjustment described above.

e. Market Basket Update Factor for FY 2018

Sections 1888(e)(4)(E)(ii)(IV) and 1888(e)(5)(i) of the Act require that the update factor used to establish the FY 2018 unadjusted federal rates be at a level equal to the market basket index percentage change. Accordingly, we determined the total growth from the average market basket level for the period of October 1, 2016, through September 30, 2017 to the average market basket level for the period of October 1, 2017, through September 30, 2018. This process yields a percentage change in the 2014-based SNF market basket of 2.6 percent.

As further explained in section III.B.2.c. of this final rule, as applicable, we adjust the market basket percentage change by the forecast error from the most recently available FY for which there is final data and apply this adjustment whenever the difference between the forecasted and actual percentage change in the market basket exceeds a 0.5 percentage point threshold. Since the difference between the forecasted FY 2016 SNF market basket percentage change and the actual FY 2016 SNF market basket percentage change (FY 2016 is the most recently available FY for which there is historical data) did not exceed the 0.5 percentage point threshold, the FY 2018 market basket percentage change of 2.6 percent would not have been adjusted by the forecast error correction.

If not for the enactment of section 411(a) of the MACRA, the SNF market basket for FY 2018 would be determined in accordance with section 1888(e)(5)(B)(ii) of the Act, which requires us to reduce the market basket percentage change by the MFP adjustment (the 10-year moving average of changes in MFP for the period ending September 30, 2018) of 0.6 percent, as described in section III.B.2.d. of this final rule. Thus, absent the enactment of MACRA, the resulting net SNF market basket update would equal 2.0 percent, or 2.6 percent less the 0.6 percentage point MFP adjustment. We note that our policy has been that, if more recent data become available (for example, a more recent estimate of the SNF market basket and/or MFP adjustment), we would use such data, if appropriate, to determine the SNF market basket percentage change, labor-related share relative importance, forecast error adjustment, and MFP adjustment in the SNF PPS final rule.

Commenters submitted the following comments related to the proposed rule's discussion of the market basket update factor for FY 2018. A discussion of these comments, along with our responses, appears below.

Comment: We received a number of comments in relation to applying the FY 2018 market basket update factor in the determination of the FY 2018 unadjusted federal per diem rates, with some commenters supporting its application in determining the FY 2018 unadjusted per diem rates, while others opposed its application. In their March 2017 report (available at http://medpac.gov/docs/default-source/reports/mar17_medpac_ch8.pdf ) and in their comment on the FY 2018 SNF PPS proposed rule, MedPAC recommended that we eliminate the market basket update for SNFs altogether for FY 2018 and FY 2019 and implement revisions to the SNF PPS. A few commenters also encouraged us to consider the “gap” between the customary market basket update, as reflected in the MFP-adjusted market basket update factor described above and the MACRA-required 1.0 percentage point market basket update.

Response: We appreciate all of the comments received on the proposed market basket update for FY 2018. In response to those comments opposing the application of the FY 2018 market basket update factor in determining the FY 2018 unadjusted federal per diem rates (specifically, MedPAC's proposal to eliminate the market basket update for SNFs), we note that under sections 1888(e)(4)(E)(ii)(IV) and (e)(5)(B) of the Act, we are required to update the unadjusted federal per diem rates each fiscal year by the SNF market basket percentage change, as reduced by the MFP adjustment, and that, under section 1888(e)(5)(B)(iii) of the Act (as added by section 411(a) of MACRA), for FY 2018, that update must be 1.0 percentage point.

With regard to those comments on the “gap” between the standard market basket update and the MACRA-required update, we appreciate these commenters' concerns, but we are required in section 1888(e)(5)(B)(iii) of the Act, as added by section 411(a) of MACRA, to apply the 1.0 percentage point update factor for FY 2018.

Comment: One commenter requested that we engage in an ongoing dialogue with the commenter's association on their market basket research, which would serve to inform us and support any analogous CMS reform efforts.

Response: We appreciate the commenter's review of the market basket and interest in continued dialogue regarding their research. The commenter is encouraged to submit any research to CMSDNHS@cms.hhs.gov.

Comment: One commenter stated that we have the statutory authority to implement geographically-specific updates associated with state and/or regional minimum wage laws. The commenter requested that such updates be made at the Core-Based Statistical Area (CBSA) levels.

Response: We would note that any increases in wages resulting from state and/or regional minimum wage laws are likely to be reflected in data used to create the SNF PPS wage index. Therefore, we believe such standards are already taken into account in the calculation of the SNF PPS wage index to the extent that these laws have an impact on wages.

Accordingly, after considering the comments received, for the reasons specified in this final rule and in the FY 2018 SNF PPS proposed rule (82 FR 21017 through 21019), we are finalizing the FY 2018 market basket factor of 1.0 percent, as required by section 411(a) of MACRA. Historically, we have used the SNF market basket, adjusted as described above, to adjust each per diem component of the federal rates forward to reflect the change in the average prices from one year to the next. However, section 1888(e)(5)(B)(iii) of the Act, as added by section 411(a) of the MACRA, requires us to use a market basket percentage of 1.0 percent, after application of the MFP adjustment to adjust the federal rates for FY 2018. Under section 1888(e)(5)(B)(iii) of the Act, the market basket percentage increase used to determine the federal rates set forth in this final rule will be 1.0 percent for FY 2018. Tables 2 and 3 reflect the updated components of the unadjusted federal rates for FY 2018, prior to adjustment for case-mix.

In addition, we note that section 1888(e)(6)(A)(i) of the Act provides that, beginning in FY 2018, SNFs that fail to submit data, as applicable, in accordance with sections 1888(e)(6)(B)(i)(II) and (III) of the Act for a fiscal year will receive a 2.0 percentage point reduction to their market basket update for the fiscal year involved, after application of section 1888(e)(5)(B)(ii) of the Act (the MFP adjustment) and section 1888(e)(5)(B)(iii) of the Act (the 1 percent market basket increase for FY 2018) (for additional information on the SNF QRP, including the statutory authority and the selected measures, we refer readers to section III.D.2. of this final rule). In addition, section 1888(e)(6)(A)(ii) of the Act states that application of the 2.0 percentage point reduction (after application of section 1888(e)(5)(B)(ii) and (iii) of the Act) may result in the market basket index percentage change being less than 0.0 for a fiscal year, and may result in payment rates for a fiscal year being less than such payment rates for the preceding fiscal year. Section 1888(e)(6)(A)(iii) of the Act further specifies that the 2.0 percentage point reduction is applied in a noncumulative manner, so that any reduction made under section 1888(e)(6)(A)(i) of the Act shall apply only for the fiscal year involved, and the Secretary shall not take into account such reduction in computing the payment amount for a subsequent fiscal year. We did not receive any comments specifically on the market basket reduction under the SNF QRP and any comments on the SNF QRP more broadly are discussed in section III.D.2 of this final rule.

3. Case-Mix Adjustment

Under section 1888(e)(4)(G)(i) of the Act, the federal rate also incorporates an adjustment to account for facility case-mix, using a classification system that accounts for the relative resource utilization of different patient types. The statute specifies that the adjustment is to reflect both a resident classification system that the Secretary establishes to account for the relative resource use of different patient types, as well as resident assessment data and other data that the Secretary considers appropriate. In the interim final rule with comment period that initially implemented the SNF PPS (63 FR 26252, May 12, 1998), we developed the RUG-III case-mix classification system, which tied the amount of payment to resident resource use in combination with resident characteristic information. Staff time measurement (STM) studies conducted in 1990, 1995, and 1997 provided information on resource use (time spent by staff members on residents) and resident characteristics that enabled us not only to establish RUG-III, but also to create case-mix indexes (CMIs). The original RUG-III grouper logic was based on clinical data collected in 1990, 1995, and 1997. As discussed in the SNF PPS proposed rule for FY 2010 (74 FR 22208), we subsequently conducted a multi-year data collection and analysis under the Staff Time and Resource Intensity Verification (STRIVE) project to update the case-mix classification system for FY 2011. The resulting Resource Utilization Groups, Version 4 (RUG-IV) case-mix classification system reflected the data collected in 2006 through 2007 during the STRIVE project, and was finalized in the FY 2010 SNF PPS final rule (74 FR 40288) to take effect in FY 2011 concurrently with an updated new resident assessment instrument, version 3.0 of the Minimum Data Set (MDS 3.0), which collects the clinical data used for case-mix classification under RUG-IV.

We note that case-mix classification is based, in part, on the beneficiary's need for skilled nursing care and therapy services. The case-mix classification system uses clinical data from the MDS to assign a case-mix group to each patient that is then used to calculate a per diem payment under the SNF PPS. As discussed in section III.C.1. of this final rule, the clinical orientation of the case-mix classification system supports the SNF PPS's use of an administrative presumption that considers a beneficiary's initial case-mix classification to assist in making certain SNF level of care determinations. Further, because the MDS is used as a basis for payment, as well as a clinical assessment, we have provided extensive training on proper coding and the time frames for MDS completion in our Resident Assessment Instrument (RAI) Manual. For an MDS to be considered valid for use in determining payment, the MDS assessment must be completed in compliance with the instructions in the RAI Manual in effect at the time the assessment is completed. For payment and quality monitoring purposes, the RAI Manual consists of both the Manual instructions and the interpretive guidance and policy clarifications posted on the appropriate MDS Web site at http://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/MDS30RAIManual.html.

In addition, we note that section 511 of the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (MMA, Pub. L. 108-173, enacted December 8, 2003) amended section 1888(e)(12) of the Act to provide for a temporary increase of 128 percent in the PPS per diem payment for any SNF residents with Acquired Immune Deficiency Syndrome (AIDS), effective with services furnished on or after October 1, 2004. This special add-on for SNF residents with AIDS was to remain in effect only until the Secretary certifies that there is an appropriate adjustment in the case mix to compensate for the increased costs associated with such residents. The add-on for SNF residents with AIDS is also discussed in Program Transmittal #160 (Change Request #3291), issued on April 30, 2004, which is available online at www.cms.gov/transmittals/downloads/r160cp.pdf. In the SNF PPS final rule for FY 2010 (74 FR 40288), we did not address this certification in that final rule's implementation of the case-mix refinements for RUG-IV, thus allowing the add-on payment required by section 511 of the MMA to remain in effect for the time being.

For the limited number of SNF residents that qualify for this add-on, there is a significant increase in payments. For example, using FY 2015 data (which still used ICD-9-CM coding), we identified fewer than 5085 SNF residents with a diagnosis code of 042 (Human Immunodeficiency Virus (HIV) Infection). As explained in the FY 2016 SNF PPS final rule (80 FR 46397 through 46398), on October 1, 2015 (consistent with section 212 of PAMA), we converted to using ICD-10-CM code B20 to identify those residents for whom it is appropriate to apply the AIDS add-on established by section 511 of the MMA. For FY 2018, an urban facility with a resident with AIDS in RUG-IV group “HC2” would have a case-mix adjusted per diem payment of $443.08 (see Table 4) before the application of the MMA adjustment. After an increase of 128 percent, this urban facility would receive a case-mix adjusted per diem payment of approximately $1,010.22.

Under section 1888(e)(4)(H) of the Act, each update of the payment rates must include the case-mix classification methodology applicable for the upcoming FY. The FY 2018 payment rates set forth in this final rule reflect the use of the RUG-IV case-mix classification system from October 1, 2017, through September 30, 2018. We list the case-mix adjusted RUG-IV payment rates for FY 2018, provided separately for urban and rural SNFs, in Tables 4 and 5 with corresponding case-mix values. We use the revised OMB delineations adopted in the FY 2015 SNF PPS final rule (79 FR 45632, 45634) to identify a facility's urban or rural status for the purpose of determining which set of rate tables applies to the facility. Tables 4 and 5 do not reflect the add-on for SNF residents with AIDS enacted by section 511 of the MMA, which we apply only after making all other adjustments (such as wage index and case-mix).

4. Wage Index Adjustment

Section 1888(e)(4)(G)(ii) of the Act requires that we adjust the federal rates to account for differences in area wage levels, using a wage index that the Secretary determines appropriate. Since the inception of the SNF PPS, we have used hospital inpatient wage data in developing a wage index to be applied to SNFs. We proposed to continue this practice for FY 2018, as we continue to believe that in the absence of SNF-specific wage data, using the hospital inpatient wage index data is appropriate and reasonable for the SNF PPS. As explained in the update notice for FY 2005 (69 FR 45786), the SNF PPS does not use the hospital area wage index's occupational mix adjustment, as this adjustment serves specifically to define the occupational categories more clearly in a hospital setting; moreover, the collection of the occupational wage data also excludes any wage data related to SNFs. Therefore, we believe that using the updated wage data exclusive of the occupational mix adjustment continues to be appropriate for SNF payments. For FY 2018, the updated wage data are for hospital cost reporting periods beginning on or after October 1, 2013 and before October 1, 2014 (FY 2014 cost report data).

We note that section 315 of the Medicare, Medicaid, and SCHIP Benefits Improvement and Protection Act of 2000 (BIPA, Pub. L. 106-554, enacted on December 21, 2000) authorized us to establish a geographic reclassification procedure that is specific to SNFs, but only after collecting the data necessary to establish a SNF wage index that is based on wage data from nursing homes. However, to date, this has proven to be unfeasible due to the volatility of existing SNF wage data and the significant amount of resources that would be required to improve the quality of that data. More specifically, we believe auditing all SNF cost reports, similar to the process used to audit inpatient hospital cost reports for purposes of the Inpatient Prospective Payment System (IPPS) wage index, would place a burden on providers in terms of responding to documented audit requests. We also believe that adopting such an approach would require a significant commitment of resources by CMS and the Medicare Administrative Contractors, potentially far in excess of those required under the IPPS given that there are nearly five times as many SNFs as there are hospitals. Therefore, while we continue to believe that the development of such an audit process could improve SNF cost reports in such a manner as to permit us to establish a SNF-specific wage index, we do not regard an undertaking of this magnitude as being feasible within the current level of programmatic resources.

In addition, we proposed to continue to use the same methodology discussed in the SNF PPS final rule for FY 2008 (72 FR 43423) to address those geographic areas in which there are no hospitals, and thus, no hospital wage index data on which to base the calculation of the FY 2018 SNF PPS wage index. For rural geographic areas that do not have hospitals and, therefore, lack hospital wage data on which to base an area wage adjustment, we stated in the proposed rule we would use the average wage index from all contiguous Core-Based Statistical Areas (CBSAs) as a reasonable proxy. For FY 2018, there are no rural geographic areas that do not have hospitals, and thus, we stated that this methodology would not be applied. For rural Puerto Rico, we stated that we would not apply this methodology due to the distinct economic circumstances that exist there (for example, due to the close proximity to one another of almost all of Puerto Rico's various urban and non-urban areas, this methodology would produce a wage index for rural Puerto Rico that is higher than that in half of its urban areas); instead, we stated we would continue to use the most recent wage index previously available for that area. For urban areas without specific hospital wage index data, we stated we would use the average wage indexes of all of the urban areas within the state to serve as a reasonable proxy for the wage index of that urban CBSA. For FY 2018, the only urban area without wage index data available is CBSA 25980, Hinesville-Fort Stewart, GA. The wage index applicable to FY 2018 is set forth in Tables A and B available on the CMS Web site at http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/SNFPPS/WageIndex.html.

In the SNF PPS final rule for FY 2006 (70 FR 45026, August 4, 2005), we adopted the changes discussed in the OMB Bulletin No. 03-04 (June 6, 2003), available online at https://www.whitehouse.gov/omb/bulletins_b03-04,, which announced revised definitions for MSAs and the creation of micropolitan statistical areas and combined statistical areas.

In adopting the CBSA geographic designations, we provided for a 1-year transition in FY 2006 with a blended wage index for all providers. For FY 2006, the wage index for each provider consisted of a blend of 50 percent of the FY 2006 MSA-based wage index and 50 percent of the FY 2006 CBSA-based wage index (both using FY 2002 hospital data). We referred to the blended wage index as the FY 2006 SNF PPS transition wage index. As discussed in the SNF PPS final rule for FY 2006 (70 FR 45041), since the expiration of this one-year transition on September 30, 2006, we have used the full CBSA-based wage index values.

In the FY 2015 SNF PPS final rule (79 FR 45644 through 45646), we finalized changes to the SNF PPS wage index based on the newest OMB delineations, as described in OMB Bulletin No. 13-01, beginning in FY 2015, including a 1-year transition with a blended wage index for FY 2015. OMB Bulletin No. 13-01 established revised delineations for Metropolitan Statistical Areas, Micropolitan Statistical Areas, and Combined Statistical Areas in the United States and Puerto Rico based on the 2010 Census, and provided guidance on the use of the delineations of these statistical areas using standards published on June 28, 2010 in the Federal Register (75 FR 37246 through 37252). Subsequently, on July 15, 2015, OMB issued OMB Bulletin No. 15-01, which provides minor updates to and supersedes OMB Bulletin No. 13-01 that was issued on February 28, 2013. The attachment to OMB Bulletin No. 15-01 provides detailed information on the update to statistical areas since February 28, 2013. The updates provided in OMB Bulletin No. 15-01 are based on the application of the 2010 Standards for Delineating Metropolitan and Micropolitan Statistical Areas to Census Bureau population estimates for July 1, 2012 and July 1, 2013. As we previously stated in the FY 2008 SNF PPS proposed and final rules (72 FR 25538 through 25539, and 72 FR 43423), we again wish to clarify that this and all subsequent SNF PPS rules and notices are considered to incorporate any updates and revisions set forth in the most recent OMB bulletin that applies to the hospital wage data used to determine the current SNF PPS wage index. As noted above, the wage index applicable to FY 2018 is set forth in Tables A and B available on the CMS Web site at http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/SNFPPS/WageIndex.html.

Once calculated, we stated in the proposed rule we would apply the wage index adjustment to the labor-related portion of the federal rate. Each year, we calculate a revised labor-related share, based on the relative importance of labor-related cost categories (that is, those cost categories that are labor-intensive and vary with the local labor market) in the input price index. In the SNF PPS final rule for FY 2014 (78 FR 47944 through 47946), we finalized a proposal to revise the labor-related share to reflect the relative importance of the FY 2010-based SNF market basket cost weights for the following cost categories: Wages and Salaries; Employee Benefits; Professional fees: Labor-related; Administrative and Facilities Support Services; All other—Labor-Related Services; and a proportion of Capital-Related expenses. Effective beginning FY 2018, as discussed in section III.D.1. of the proposed rule, we proposed to revise the labor-related share to reflect the relative importance of the 2014-based SNF market basket cost weights for the following cost categories: Wages and Salaries; Employee Benefits; Professional fees: Labor-related; Administrative and Facilities Support services; Installation, Maintenance, and Repair services; All Other: Labor-Related Services; and a proportion of Capital-Related expenses.

We calculate the labor-related relative importance from the SNF market basket, and it approximates the labor-related portion of the total costs after taking into account historical and projected price changes between the base year and FY 2018. The price proxies that move the different cost categories in the market basket do not necessarily change at the same rate, and the relative importance captures these changes. Accordingly, the relative importance figure more closely reflects the cost share weights for FY 2018 than the base year weights from the SNF market basket. The methodology for calculating the labor-related portion for FY 2018 is discussed in section III.D.1. of this final rule and the labor-related share is provided in Table 15.

We invited public comments on these proposals. A discussion of the comments we received, along with our responses, appear below.

Comment: One commenter expressed concern with what appears to be a precipitous drop in the New Bern, North Carolina (CBSA 35100) wage index. The commenter noted that in the SNF PPS final rule for 2017, the wage index for this CBSA was 0.8539, but that in the FY 2018 SNF PPS proposed rule, this value had dropped to 0.5988. The commenter requests that the information used to determine the wage indexes be reviewed prior to the release of the final rule.

Response: We appreciate the commenter's concern regarding the decrease in the wage index for CBSA 35100. There is a wage data verification and correction process which is discussed in the Inpatient Prospective Payment System (IPPS) proposed and final rules each year. The most recent discussion appears in the FY 2018 IPPS proposed rule (82 FR 19899 through 19900. 19911 through 19915). Based on the final wage data for FY 2018, the wage index for CBSA 35100 has been updated to 0.8277, which is only a slight decrease compared to the FY 2017 value.

Comment: Several commenters recommend that we continue exploring potential approaches to establish a SNF-specific wage index either by modifying the use of current hospital wage data by eliminating certain job categories specific to hospitals only, or by utilizing collected SNF-specific wage data only. More specifically, these commenters suggest that a SNF-specific wage index could benefit from weighting it by occupational mix data for SNFs, allowing for a rural floor policy, and by implementation of a reclassification system.

Response: We appreciate the commenters raising these concerns regarding the use of the hospital wage index data under the SNF PPS, and the commenter's recommendation to continue exploring potential approaches for collecting SNF-specific wage data to establish a SNF-specific wage index. However, we note that, consistent with the preceding discussion in this final rule as well as our previous responses to these recurring comments (most recently published in the FY 2017 SNF PPS final rule (81 FR 51979 through 51980)), developing such a wage index would require a resource-intensive audit process similar to that used for IPPS hospital data, to improve the quality of the SNF cost report data in order for it to be used as part of this analysis. We would further note that as this audit process is quite extensive in the case of approximately 3,300 hospitals, it would be significantly more so in the case of approximately 15,000 SNFs. As discussed above, we believe auditing all SNF cost reports, similar to the process used to audit inpatient hospital cost reports for purposes of the Inpatient Prospective Payment System (IPPS) wage index, would place a burden on providers in terms of recordkeeping and completion of the cost report worksheet. We also believe that adopting such an approach would require a significant commitment of resources by CMS and the Medicare Administrative Contractors, potentially far in excess of those required under the IPPS given that there are nearly five times as many SNFs as there are hospitals. Therefore, while we continue to review all available data and contemplate the potential methodological approaches for a SNF-specific wage index in the future, we continue to believe that in the absence of the appropriate SNF-specific wage data, using the pre-reclassified hospital inpatient wage data (without the occupational mix adjustment) is appropriate and reasonable for the SNF PPS.

Further, we appreciate these commenters' suggestion that we modify the current hospital wage data used to construct the SNF PPS wage index to reflect the SNF environment more accurately by eliminating certain job categories specific to hospitals only. While we consider whether or not such an approach may constitute an interim step in the process of developing a SNF-specific wage index, we would note that other provider types also use the hospital wage index as the basis for their associated wage index. As such, we believe that such a recommendation should be part of a broader discussion of wage index reform across Medicare payment systems.

We note that section 315 of BIPA authorized us to establish a geographic reclassification procedure that is specific to SNFs, only after collecting the data necessary to establish a SNF-specific wage index that is based on data from nursing homes. However, to date this has been infeasible due to the volatility of existing SNF wage data and the significant amount of resources that would be required to improve the quality of that data. To the extent we are able to develop and implement a SNF-specific wage index in the future, we may consider at that time whether it would be appropriate to implement a reclassification system and an occupational mix adjustment, as suggested by commenters.

As it relates to the suggestion that we adopt a rural floor policy with a SNF-specific wage index, we do not believe it would be prudent to adopt such a policy under the SNF PPS. As we stated in the FY 2016 SNF PPS final rule (80 FR 46401), MedPAC has recommended eliminating the rural floor policy (which actually sets a floor for urban hospitals) from the calculation of the IPPS wage index (see, for example, Chapter 3 of MedPAC's March 2013 Report to Congress on Medicare Payment Policy, available at http://medpac.gov/docs/default-source/reports/mar13_ch03.pdf,, which notes on page 65 that in 2007, MedPAC had “. . . recommended eliminating these special wage index adjustments and adopting a new wage index system to avoid geographic inequities that can occur due to current wage index policies (Medicare Payment Advisory Commission 2007b.”) As we stated in the FY 2016 SNF PPS final rule, if we were to adopt the rural floor under the SNF PPS, we believe that the SNF PPS wage index could become vulnerable to problems similar to those that MedPAC identified in its March 2013 Report to Congress.

Accordingly, after considering the comments received and for the reasons discussed previously in this section and in the FY 2018 SNF PPS proposed rule (82 FR 21022 through 21026), we are finalizing the FY 2018 wage index adjustment and related policies as proposed in the FY 2018 SNF PPS proposed rule. For FY 2018, the updated wage data are for hospital cost reporting periods beginning on or after October 1, 2013 and before October 1, 2014 (FY 2014 cost report data). As noted above, the wage index applicable to FY 2018 is set forth in Tables A and B available on the CMS Web site at http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/SNFPPS/WageIndex.html. Tables 6 and 7 show the RUG-IV case-mix adjusted federal rates for FY 2018 by labor-related and non-labor-related components.

Section 1888(e)(4)(G)(ii) of the Act also requires that we apply this wage index in a manner that does not result in aggregate payments under the SNF PPS that are greater or less than would otherwise be made if the wage adjustment had not been made. For FY 2018 (federal rates effective October 1, 2017), we stated in the proposed rule that we would apply an adjustment to fulfill the budget neutrality requirement. We stated we would meet this requirement by multiplying each of the components of the unadjusted federal rates by a budget neutrality factor equal to the ratio of the weighted average wage adjustment factor for FY 2017 to the weighted average wage adjustment factor for FY 2018. For this calculation, we stated we would use the same FY 2016 claims utilization data for both the numerator and denominator of this ratio. We define the wage adjustment factor used in this calculation as the labor share of the rate component multiplied by the wage index plus the non-labor share of the rate component. We proposed a budget neutrality factor of 1.0003. We did not receive any comments regarding our proposed budget neutrality calculation. Thus, we are finalizing the budget neutrality methodology as proposed. The final budget neutrality factor for FY 2018 is 1.0013. We note that this is different from the budget neutrality factor provided in the FY 2018 SNF PPS proposed rule (82 FR 21026) due to an updated wage index file and updated claims file used to calculate the budget neutrality factor.

5. Adjusted Rate Computation Example

Using the hypothetical SNF XYZ, Table 8 shows the adjustments made to the federal per diem rates to compute the provider's actual per diem PPS payment for FY 2018. We derive the Labor and Non-labor columns from Table 6. The wage index used in this example is based on the FY 2018 SNF PPS wage index, which may be found in Table A available on the CMS Web site at http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/SNFPPS/WageIndex.html. As illustrated in Table 8, SNF XYZ's total PPS payment for FY 2018 would equal $47,596.42.

C. Additional Aspects of the SNF PPS

1. SNF Level of Care—Administrative Presumption

The establishment of the SNF PPS did not change Medicare's fundamental requirements for SNF coverage. However, because the case-mix classification is based, in part, on the beneficiary's need for skilled nursing care and therapy, we have attempted, where possible, to coordinate claims review procedures with the existing resident assessment process and case-mix classification system discussed in section III.B.3. of this final rule. This approach includes an administrative presumption that utilizes a beneficiary's initial classification in one of the upper 52 RUGs of the 66-group RUG-IV case-mix classification system to assist in making certain SNF level of care determinations.

In accordance with § 413.345, we include in each update of the federal payment rates in the Federal Register the designation of those specific RUGs under the classification system that represent the required SNF level of care, as provided in § 409.30. As set forth in the FY 2011 SNF PPS update notice (75 FR 42910), this designation reflects an administrative presumption under the 66-group RUG-IV system that beneficiaries who are correctly assigned to one of the upper 52 RUG-IV groups on the initial 5-day, Medicare-required assessment are automatically classified as meeting the SNF level of care definition up to and including the assessment reference date (ARD) on the 5-day Medicare-required assessment.

A beneficiary assigned to any of the lower 14 RUG-IV groups is not automatically classified as either meeting or not meeting the definition, but instead receives an individual level of care determination using the existing administrative criteria. This presumption recognizes the strong likelihood that beneficiaries assigned to one of the upper 52 RUG-IV groups during the immediate post-hospital period require a covered level of care, which would be less likely for those beneficiaries assigned to one of the lower 14 RUG-IV groups.

In the July 30, 1999 final rule (64 FR 41670), we indicated that we would announce any changes to the guidelines for Medicare level of care determinations related to modifications in the case-mix classification structure. In this final rule, we continue to designate the upper 52 RUG-IV groups for purposes of this administrative presumption, consisting of all groups encompassed by the following RUG-IV categories:

• Rehabilitation plus Extensive Services.
• Ultra High Rehabilitation.
• Very High Rehabilitation.
• High Rehabilitation.
• Medium Rehabilitation.
• Low Rehabilitation.
• Extensive Services.
• Special Care High.
• Special Care Low.
• Clinically Complex.

However, we note that this administrative presumption policy does not supersede the SNF's responsibility to ensure that its decisions relating to level of care are appropriate and timely, including a review to confirm that the services prompting the beneficiary's assignment to one of the upper 52 RUG-IV groups (which, in turn, serves to trigger the administrative presumption) are themselves medically necessary. As we explained in the FY 2000 SNF PPS final rule (64 FR 41667), the administrative presumption:

. . . is itself rebuttable in those individual cases in which the services actually received by the resident do not meet the basic statutory criterion of being reasonable and necessary to diagnose or treat a beneficiary's condition (according to section 1862(a)(1) of the Act). Accordingly, the presumption would not apply, for example, in those situations in which a resident's assignment to one of the upper . . . groups is itself based on the receipt of services that are subsequently determined to be not reasonable and necessary.

Moreover, we want to stress the importance of careful monitoring for changes in each patient's condition to determine the continuing need for Part A SNF benefits after the ARD of the 5-day assessment.

In connection with the administrative level of care presumption, in the FY 2018 SNF PPS proposed rule (82 FR 21027), we proposed to amend the existing regulations text at § 413.345 by removing the parenthetical phrase “(including the designation of those specific Resource Utilization Groups under the resident classification system that represent the required SNF level of care, as provided in § 409.30 of this chapter)” that currently appears in the second sentence of § 413.345. We stated in the proposed rule that the deletion of the current reference to publishing such material annually in the Federal Register, along with the specific reference to “Resource Utilization Groups,” would serve to conform the text of these regulations more closely to that of the corresponding statutory language at section 1888(e)(4)(H)(ii) of the Act, which refers in more general terms to the applicable “case mix classification system.” Moreover, we noted in the proposed rule that the recurring announcements in the Federal Register of the administrative presumption's designated groups as part of each annual update of the SNF PPS rates has in actual practice proven to be largely a formality, resulting in exactly the same designated groups repetitively being promulgated routinely year after year. Accordingly, we proposed instead to disseminate this standard description of the administrative presumption's designated groups exclusively through the SNF PPS Web site, and to announce such designations in rulemaking only in the event that we are actually proposing to make changes in them.

Along with this proposed revision, we also proposed to make appropriate conforming revisions in other portions of the regulations text (82 FR 21027). Specifically, we proposed to remove from the introductory text of § 409.30, the parenthetical phrase “(in the annual publication of Federal prospective payment rates described in § 413.345 of this chapter)” for the same reasons we proposed to remove the parenthetical phrase from § 413.345, as discussed in the proposed rule and in this final rule above. In addition, we proposed to replace the phrase to “one of the Resource Utilization Groups that is designated” in § 409.30's introductory text with the phrase “one of the case-mix classifiers CMS designates” to conform more closely with the statutory language in section 1888(e)(4)(G) and (H) of the Act, which refers in more general terms to the “resident classification system” or “case mix classification system,” and to clarify that “CMS” makes these designations. Additionally, we proposed to revise § 409.30 to reflect more clearly our longstanding policy that the assignment of a designated case-mix classifier would serve to trigger the administrative presumption only when that assignment is itself correct. As we noted in the FY 2000 SNF PPS final rule (64 FR 41667, July 30, 1999), “. . . the presumption would not apply, for example, in those situations in which a resident's assignment to one of the upper . . . groups is itself based on the receipt of services that are subsequently determined to be not reasonable and necessary.” We also proposed to make similar conforming revisions in the “resident classification system” definition that currently appears in § 413.333 to replace “Resource Utilization Groups” with “resident classification system”, as well as in the material in § 424.20(a)(1)(ii) on SNF level of care certifications to replace the phrase “one of the Resource Utilization Groups designated” with “one of the case-mix classifiers that CMS designates,” in both cases to conform more closely with the statutory language in section 1888(e)(4)(G) and (H) of the Act, as discussed in the proposed rule (82 FR 21027) and in this final rule, which refers in more general terms to the “resident classification system” or “case mix classification system,” and to clarify in § 424.20(a)(1)(ii) that “CMS” designates these case-mix classifiers. Finally, regarding § 424.20, we proposed to revise paragraph (e)(2)(ii)(B)(2) by updating its existing cross-reference to the provision at § 483.40(e) on delegating physician tasks in SNFs, which was recently redesignated as new § 483.30(e) under the revised long-term care facility requirements for participation (81 FR 68861, October 4, 2016). Finally, we proposed to remove the word “Optional” from the title of 42 CFR part 413 (82 FR 21098), as this is an obsolete reference to an optional prospective payment methodology for low-volume SNFs that predated the SNF PPS and is no longer in effect.

Commenters submitted the following comments on our proposals described above related to the SNF Level of Care—Administrative Presumption aspects of the SNF PPS. A discussion of these comments, along with our responses, appears below.

Comment: We received a comment about our proposed revisions to §§ 413.333 and 413.345 that would result in removing the term “Resource Utilization Groups,” and in § 413.333, utilizing the term “resident classification system” in its place. The commenter interpreted our use of the term “resident classification system” in this context as referring specifically to the Resident Classification System, Version I (RCS-I), the particular case-mix classification model that is currently under development as discussed in our advance notice of proposed rulemaking with comment (CMS-1686-ANPRM, 82 FR 20980, May 4, 2017). Based on that assumption, the commenter expressed the view that it would be premature and confusing to adopt terminology referencing a particular model that has not been finalized at this point.

Response: We wish to clarify that our use of the term “resident classification system” in this context refers solely to a case-mix classification system in the generic sense, and not to the particular model discussed in the ANPRM, which we will continue to refer to as the Resident Classification System, Version I (or RCS-I). We note that the term “resident classification system” in the more generic sense has long been utilized as such in the existing regulations at § 413.333, and that our proposed changes were not intended to restrict the regulations text to any one particular type of classification system, but rather, to do the opposite by removing the existing, specific references to the RUG model. As we noted in the proposed rule (82 FR 21027), such revisions would actually serve to conform the regulations text “. . . more closely with the statutory language in section 1888(e)(4)(G) and (H) of the Act, . . . which refers in more general terms to the ‘resident classification system’ . . .” (emphasis added). Accordingly, we are revising these portions of the regulations text as proposed, as discussed in this final rule.

Comment: One commenter inquired about our proposed clarification in § 409.30 which, similar to the existing regulations at § 424.20(a)(1)(ii), would specify that a resident qualifies for the level of care presumption only when “correctly” assigned to one of the case-mix classifiers designated for this purpose. In explaining the reason for this clarification in the proposed rule (82 FR 21027), we cited a prior discussion of the presumption in the FY 2000 final rule (64 FR 41667, July 30, 1999), which had noted that “. . . the presumption would not apply, for example, in those situations in which a resident's assignment to one of the upper . . . groups is itself based on the receipt of services that are subsequently determined to be not reasonable and necessary.” The commenter questioned whether, in this scenario, the resident's assignment to a RUG that turns out to be incorrect would result in disqualifying the resident from SNF coverage altogether. The commenter also requested clarification in the wording of a portion of § 30.1 of the Medicare Benefit Policy Manual (MBPM), Chapter 8 that discusses how services furnished during the prior hospital stay are to be coded on the resident assessment.

Response: Regarding the scenario discussed above (in which the services that triggered a given RUG assignment on the initial assessment are found to be not reasonable and necessary), if the resident is then reassigned to a different RUG that is itself designated as meeting the level of care presumption, the resident would, in fact, still qualify for the presumption on that basis, as the end result of the reassignment would be that the resident has been “correctly assigned” to one of the designated RUGs on that assessment. Alternatively, if the reassignment is to one of the less intensive RUGs that is not designated as meeting the presumption, the resident would still receive an individual level of care determination using the existing administrative criteria. Finally, regarding the request to clarify the MBPM instructions on coding procedures, we believe this comment is beyond the scope of this rule. As we noted in the FY 2002 SNF PPS final rule, “. . . specific operational instructions (such as those describing the details of particular billing procedures) are beyond the scope of the SNF PPS final rule” (66 FR 39588, July 31, 2001). However, we will forward this comment to the appropriate component within CMS for consideration.

After consideration of the comments received, for the reasons discussed above and in the FY 2018 SNF PPS proposed rule (82 FR 21026 through 21027), we are finalizing, without modification, our proposed revisions to §§ 409.30, 413.333, 413.345, 424.20(a)(1)(ii) and (e)(2)(ii)(B)(2), and our revision to the title of 42 CFR part 413 as discussed in this final rule. In addition, as we proposed, we will henceforth disseminate the standard description of the administrative presumption's designated groups exclusively through the SNF PPS Web site, and will announce such designations in rulemaking only in the event that we are actually proposing to make changes in them.

2. Consolidated Billing

Sections 1842(b)(6)(E) and 1862(a)(18) of the Act (as added by section 4432(b) of the BBA) require a SNF to submit consolidated Medicare bills to its Medicare Administrative Contractor (MAC) for almost all of the services that its residents receive during the course of a covered Part A stay. In addition, section 1862(a)(18) of the Act places the responsibility with the SNF for billing Medicare for physical therapy, occupational therapy, and speech-language pathology services that the resident receives during a noncovered stay. Section 1888(e)(2)(A) of the Act excludes a small list of services from the consolidated billing provision (primarily those services furnished by physicians and certain other types of practitioners), which remain separately billable under Part B when furnished to a SNF's Part A resident. These excluded service categories are discussed in greater detail in section V.B.2. of the May 12, 1998 interim final rule (63 FR 26295 through 26297).

A detailed discussion of the legislative history of the consolidated billing provision is available on the SNF PPS Web site at http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/SNFPPS/Downloads/Legislative_History_04152015.pdf. In particular, section 103 of the Medicare, Medicaid, and SCHIP Balanced Budget Refinement Act of 1999 (BBRA) (Pub. L. 106-113, enacted on November 29, 1999) amended section 1888(e)(2)(A) of the Act by further excluding a number of individual high-cost, low probability services, identified by Healthcare Common Procedure Coding System (HCPCS) codes, within several broader categories (chemotherapy items, chemotherapy administration services, radioisotope services, and customized prosthetic devices) that otherwise remained subject to the provision. We discuss this BBRA amendment in greater detail in the SNF PPS proposed and final rules for FY 2001 (65 FR 19231 through 19232, April 10, 2000, and 65 FR 46790 through 46795, July 31, 2000), as well as in Program Memorandum AB-00-18 (Change Request #1070), issued March 2000, which is available online at www.cms.gov/transmittals/downloads/ab001860.pdf.

As explained in the FY 2001 proposed rule (65 FR 19232), the amendments enacted in section 103 of the BBRA not only identified for exclusion from this provision a number of particular service codes within four specified categories (that is, chemotherapy items, chemotherapy administration services, radioisotope services, and customized prosthetic devices), but also gave the Secretary the authority to designate additional, individual services for exclusion within each of the specified service categories. In the proposed rule for FY 2001, we also noted that the BBRA Conference report (H.R. Rep. No. 106-479 at 854 (1999) (Conf. Rep.)) characterizes the individual services that this legislation targets for exclusion as high-cost, low probability events that could have devastating financial impacts because their costs far exceed the payment SNFs receive under the PPS. According to the conferees, section 103(a) of the BBRA is an attempt to exclude from the PPS certain services and costly items that are provided infrequently in SNFs. By contrast, the amendments enacted in section 103 of the BBRA do not designate for exclusion any of the remaining services within those four categories (thus, leaving all of those services subject to SNF consolidated billing), because they are relatively inexpensive and are furnished routinely in SNFs.

As we further explained in the final rule for FY 2001 (65 FR 46790), and as is consistent with our longstanding policy, any additional service codes that we might designate for exclusion under our discretionary authority must meet the same statutory criteria used in identifying the original codes excluded from consolidated billing under section 103(a) of the BBRA: They must fall within one of the four service categories specified in the BBRA; and they also must meet the same standards of high cost and low probability in the SNF setting, as discussed in the BBRA Conference report. Accordingly, we characterized this statutory authority to identify additional service codes for exclusion as essentially affording the flexibility to revise the list of excluded codes in response to changes of major significance that may occur over time (for example, the development of new medical technologies or other advances in the state of medical practice) (65 FR 46791). In the FY 2018 SNF PPS proposed rule (82 FR 21028), we specifically invited public comments identifying HCPCS codes in any of these four service categories (chemotherapy items, chemotherapy administration services, radioisotope services, and customized prosthetic devices) representing recent medical advances that might meet our criteria for exclusion from SNF consolidated billing. We stated that we may consider excluding a particular service if it meets our criteria for exclusion as specified above. We also requested that commenters identify in their comments the specific HCPCS code that is associated with the service in question, as well as their rationale for requesting that the identified HCPCS code(s) be excluded. We note that the original BBRA amendment (as well as the implementing regulations) identified a set of excluded services by means of specifying HCPCS codes that were in effect as of a particular date (in that case, as of July 1, 1999). Identifying the excluded services in this manner made it possible for us to utilize program issuances as the vehicle for accomplishing routine updates of the excluded codes, to reflect any minor revisions that might subsequently occur in the coding system itself (for example, the assignment of a different code number to the same service). Accordingly, we stated in the proposed rule that, in the event that we identify through the current rulemaking cycle any new services that would actually represent a substantive change in the scope of the exclusions from SNF consolidated billing, we would identify these additional excluded services by means of the HCPCS codes that are in effect as of a specific date (in this case, as of October 1, 2017). By making any new exclusions in this manner, we could similarly accomplish routine future updates of these additional codes through the issuance of program instructions.

In the proposed rule, we noted that one category of services which consolidated billing excludes under § 411.15(p)(3) consists of certain exceptionally intensive types of outpatient hospital services. As we explained in the FY 2000 SNF PPS final rule, this exclusion applies to “. . . those types of outpatient hospital services that we specifically identify as being beyond the scope of SNF care plans generally” (64 FR 41676, July 30, 1999, emphasis added). As discussed in the FY 2018 SNF PPS proposed rule (82 FR 21028), to further clarify this longstanding policy noted above that the outpatient hospital exclusion applies solely to those services that we specifically designate for this purpose, we proposed to revise § 411.15(p)(3)(iii) to state this more explicitly. In addition, we note that recent revisions in the long-term care facility requirements for participation (81 FR 68858, October 4, 2016) have moved the comprehensive care plan regulations from their previous location at § 483.20(k) to a new, redesignated § 483.21(b); accordingly, we proposed to make a conforming revision in the existing cross-reference to that provision that appears in § 411.15(p)(3)(iii).

We did not receive any public comments on our proposed revisions to § 411.15(p)(3)(iii). Therefore, for the reasons discussed in this final rule and in the FY 2018 SNF PPS proposed rule, we are finalizing our revisions to § 411.15(p)(3)(iii) as proposed, without modification.

Commenters submitted the following comments related to the proposed rule's discussion of the consolidated billing aspects of the SNF PPS. A discussion of these comments, along with our responses, appears below.

Comment: One commenter suggested that, rather than specifying those particular items and services that are excluded from SNF consolidated billing, CMS should comprehensively identify the full range of items and services that are subject to this provision.

Response: We note that the online listing by HCPCS code of those services that are excluded from consolidated billing (in the annual updates that are posted at https://www.cms.gov/Medicare/Billing/SNFConsolidatedBilling/index.html ) follows the overall structure of the statutory provision itself. This statutory provision, in turn, specifies in section 1888(e)(2)(A)(ii) through (iv) of the Act those particular services that are excluded from it, so that any services not so specified would remain subject to the provision (this follows the similar structure that was originally established in the hospital bundling provision at section 1862(a)(14) of the Act, which served as the model for SNF consolidated billing). As discussed in the General Explanation of the Major Categories (available online at https://www.cms.gov/Medicare/Billing/SNFConsolidatedBilling/Downloads/2017-General-Explanation.pdf ), one exception to this overall pattern involves the administrative carve-out from SNF consolidated billing under 42 CFR 411.15(p)(3)(iii) for ambulatory surgical services performed in the outpatient hospital setting (Major Category I.F):

Inclusions, rather than exclusions, are given in this one case, because of the great number of surgery procedures that are excluded and can only be safely performed in a hospital operating room setting. It is easier to automate edits around the much shorter list of inclusions under this category, representing minor procedures that can be performed in the SNF itself (emphasis in the original).

Comment: We received a number of comments regarding the statutory exclusion from consolidated billing for certain high-intensity chemotherapy drugs and the administrative exclusion for certain high-intensity outpatient hospital services. One commenter in particular expressed continuing dissatisfaction with what it characterized as CMS's “inadequate regulatory action” in modifying the consolidated billing requirement to reflect the introduction of expensive new drugs, and the expanded provision of outpatient services in nonhospital settings. The commenter cited as examples some previous comments that it had submitted during the FY 2004 rulemaking cycle, in which it had recommended the exclusion of certain additional chemotherapy drugs, and the expansion of the existing administrative exclusion for certain high-intensity outpatient hospital services to encompass freestanding (nonhospital) settings as well. Regarding the latter recommendation, the commenter indicated that to date, CMS has not revisited this “site of service” rule.

Response: Regarding the commenter's previous recommendation during the FY 2004 rulemaking cycle for additional chemotherapy exclusions, our response in the FY 2004 final rule (68 FR 46060, August 4, 2003) explained that “. . . most of the chemotherapy drugs . . . mentioned by commenters were considered for exclusion under the BBRA, but were not adopted by the Congress in the BBRA list of excluded items and services.” As further explained in several subsequent rulemaking cycles (most recently, in the FY 2016 final rule (80 FR 46407, August 4, 2015)),

. . . our position has always been that the BBRA's discretionary authority to exclude codes within certain designated service categories applies solely to codes that were created subsequent to the BBRA's enactment, and not to those codes that were already in existence as of July 1, 1999 (the date that the legislation itself uses as the reference point for identifying the codes that it designates for exclusion). As we explained in the FY 2010 final rule (74 FR 40354), this position reflects the assumption that if a particular code was already in existence as of that date but not designated for exclusion, this meant that it was intended to remain within the SNF PPS bundle, subject to the BBRA Conference Report's provision for a GAO review of the code set that was conducted the following year (H.R. Rep. 106-479 at 854 (1999) (Conf. Rep.)).

Further, we note that we have indeed continued to solicit recommendations periodically for additional exclusions within those specified service categories (such as chemotherapy services) for which the law authorizes us to do so, and we have, in fact, adopted those recommendations to the extent that the recommended services meet the applicable criteria for exclusion.

With regard to the administrative exclusion for high-intensity outpatient hospital services, we note that we not only addressed this issue in the FY 2004 final rule itself (68 FR 46061, August 4, 2003) but, as discussed below, we have revisited it repeatedly in subsequent rulemaking in response to the recurring public comments that we have received on the issue since that time. For example, the FY 2014 final rule (78 FR 47957 through 47958, August 6, 2013) cited the explanation in numerous previous rules (along with Medicare Learning Network (MLN) Matters article SE0432, available online at https://www.cms.gov/Outreach-and-Education/Medicare-Learning-Network-MLN/MLNMattersArticles/Downloads/SE0432.pdf ) that “. . . the rationale for establishing this exclusion was to address those types of services that are so far beyond the normal scope of SNF care that they require the intensity of the hospital setting in order to be furnished safely and effectively” (emphasis in the original), and also noted that when the Congress enacted the consolidated billing exclusion for certain RHC and FQHC services in section 410 of the MMA, the accompanying legislative history's description of present law directly acknowledged the hospital-specific nature of this exclusion. In addition, the FY 2012 final rule (76 FR 48532, August 8, 2011) indicated that ever since its inception, this exclusion was intended to be hospital-specific: It cited the applicable discussion in the May 12, 1998 interim final rule (63 FR 26298), which explained that this exclusion was created within the context of the concurrent development of a new PPS specifically for outpatient hospital services, reflecting the need “. . . to delineate the respective areas of responsibility for the SNF under the Consolidated Billing provision, and for the hospital under the outpatient bundling provision, with regard to these services.” This point was further reinforced in the subsequent final rule for FY 2000 (64 FR 41676, July 30, 1999), which noted that

. . . a key concern underlying the development of the consolidated billing exclusion of certain outpatient hospital services specifically involves the need to distinguish those services that comprise the SNF bundle from those that will become part of the outpatient hospital bundle that is currently being developed in connection with the outpatient hospital PPS. Accordingly, we are not extending the outpatient hospital exclusion from consolidated billing to encompass any other, freestanding settings.

Finally, the FY 2010 final rule (74 FR 40355, August 11, 2009), while acknowledging that advances in medical technology over time may make it feasible to perform such high-intensity outpatient services more widely in nonhospital settings, then went on to cite the FY 2006 final rule in noting that such a development “. . . would not argue in favor of excluding the nonhospital performance of the service from consolidated billing, . . . but rather, would call into question whether the service should continue to be excluded from consolidated billing at all, even when performed in the hospital setting” (70 FR 45049, August 4, 2005).

Comment: One commenter reiterated a recommendation made in previous rulemaking cycles to exclude the oral chemotherapy drug Revlimid® (lenalidomide).

Response: We note that a discussion of our decision not to adopt the exclusion recommendations regarding this drug appears in the final rule for FY 2015 (79 FR 45641 through 45642, August 5, 2014), which was also referenced in the FY 2017 final rule (81 FR 51985, August 5, 2016) as well.

Comment: Several commenters reiterated the same set of comments that they had submitted previously during last year's rulemaking cycle, which had noted the importance of continuing to exclude certain customized prosthetic devices from consolidated billing, and urged expanding that exclusion to encompass orthotics as well. These commenters had also recommended the following four HCPCS codes for exclusion: L5010—Partial foot, molded socket, ankle height, with toe filler; L5020—Partial foot, molded socket, tibial tubercle height, with toe filler; L5969—Addition, endoskeletal ankle-foot or ankle system, power assist, includes any type motor(s); and L5987—All lower extremity prosthesis, shank foot system with vertical loading pylon. One of the commenters now noted in addition that although our previous response in the FY 2017 final rule (81 FR 51986, August 5, 2016) had indicated that code L5969 “. . . actually appears already on the exclusion list under Major Category III.D. (‘Customized Prosthetic Devices’), where this particular L code has, in fact, been listed ever since its initial assignment in January 2014,” the commenter has been unable to locate this code on the list of exclusions in the 2017 Annual Part B MAC Update.

Response: We refer to the previous discussion in the FY 2017 final rule (81 FR 51986, August 5, 2016) regarding our decision not to adopt the recommendations for excluding orthotics and HCPCS codes L5010, L5020, and L5987. In addition, while that final rule was correct in noting that ever since its initial assignment, code L5969 has appeared as an exclusion under Major Category III.D. (“Customized Prosthetic Devices”) in the Annual Part A MAC Update, this particular code was inadvertently omitted from the corresponding exclusion list in File 1 of the Annual Part B MAC Update. We appreciate being apprised of the omission, and will take the necessary steps to rectify this oversight.

Comment: One commenter made reference to high-cost medications that are currently not excluded from consolidated billing, and requested guidance in this context regarding the applicable policy on residents being requested to supply their own medications to minimize the cost to the nursing home.

Response: In terms of Medicare payment, with limited exceptions (such as certain specified, high-intensity chemotherapy drugs), medications that are required during the course of a Medicare-covered SNF stay are included within the SNF's bundled per diem payment for the covered stay itself, which the SNF is required under the terms of its provider agreement to accept as payment in full (see section 1866(a)(1)(A)(i) of the Act and the implementing regulations at § 489.21(a)). Further, § 489.20(s) requires the SNF to furnish these bundled services either directly with its own resources, or under an “arrangement” in which the SNF itself accepts the professional and financial responsibility for the arranged-for services (see the discussion of arrangements that appears in § 409.3 and in § 10.3 of the Medicare General Information, Eligibility, and Entitlement Manual, Chapter 5). Section 489.21(h) further indicates that even if an SNF fails to furnish directly or make arrangements for such a service, the beneficiary is not to bear the financial liability for the service.

3. Payment for SNF-Level Swing-Bed Services

Section 1883 of the Act permits certain small, rural hospitals to enter into a Medicare swing-bed agreement, under which the hospital can use its beds to provide either acute- or SNF- level care, as needed. For critical access hospitals (CAHs), Part A pays on a reasonable cost basis for SNF-level services furnished under a swing-bed agreement. However, in accordance with section 1888(e)(7) of the Act, SNF-level services furnished by non-CAH rural hospitals are paid under the SNF PPS, effective with cost reporting periods beginning on or after July 1, 2002. As explained in the FY 2002 final rule (66 FR 39562), this effective date is consistent with the statutory provision to integrate swing-bed rural hospitals into the SNF PPS by the end of the transition period, June 30, 2002.

Accordingly, all non-CAH swing-bed rural hospitals have now come under the SNF PPS. Therefore, all rates and wage indexes outlined in earlier sections of this final rule for the SNF PPS also apply to all non-CAH swing-bed rural hospitals. A complete discussion of assessment schedules, the MDS, and the transmission software (RAVEN-SB for Swing Beds) appears in the FY 2002 final rule (66 FR 39562) and in the FY 2010 final rule (74 FR 40288). As finalized in the FY 2010 SNF PPS final rule (74 FR 40356 through 40357), effective October 1, 2010, non-CAH swing-bed rural hospitals are required to complete an MDS 3.0 swing-bed assessment which is limited to the required demographic, payment, and quality items. The latest changes in the MDS for swing-bed rural hospitals appear on the SNF PPS Web site at http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/SNFPPS/index.html. We received no comments on this aspect of the proposed rule.

D. Other Issues

1. Revising and Rebasing the SNF Market Basket Index

Section 1888(e)(5)(A) of the Act requires the Secretary to establish a market basket index that reflects the changes over time in the prices of an appropriate mix of goods and services included in covered SNF services. Accordingly, we have developed a SNF market basket index that encompasses the most commonly used cost categories for SNF routine services, ancillary services, and capital-related expenses. We use the SNF market basket index, adjusted in the manner described in section III.B. of this rule, to update the SNF PPS per diem rates and to determine the labor-related share on an annual basis.

The SNF market basket is a fixed-weight, Laspeyres-type price index. A Laspeyres price index measures the change in price, over time, of the same mix of goods and services purchased in the base period. Any changes in the quantity or mix of goods and services (that is, intensity) purchased over time relative to a base period are not measured.

The index itself is constructed in three steps. First, a base period is selected (in the FY 2018 SNF PPS proposed rule (82 FR 21029), the proposed base period was 2014) and total base period expenditures are estimated for a set of mutually exclusive and exhaustive spending categories with the proportion of total costs that each category represents being calculated. These proportions are called cost or expenditure weights. Second, each expenditure category is matched to an appropriate price or wage variable, referred to as a price proxy. In nearly every instance, these price proxies are derived from publicly available statistical series that are published on a consistent schedule (preferably at least on a quarterly basis). Finally, the expenditure weight for each cost category is multiplied by the level of its respective price proxy. The sum of these products (that is, the expenditure weights multiplied by their price levels) for all cost categories yields the composite index level of the market basket in a given period. Repeating this step for other periods produces a series of market basket levels over time. Dividing an index level for a given period by an index level for an earlier period produces a rate of growth in the input price index over that timeframe.

Effective for cost reporting periods beginning on or after July 1, 1998, we revised and rebased our 1977 routine costs input price index and adopted a total expenses SNF input price index using FY 1992 as the base year. In the FY 2002 SNF PPS final rule (66 FR 39582), we rebased and revised the market basket to a base year of FY 1997. In the FY 2008 SNF PPS final rule (72 FR 43425), we rebased and revised the market basket to a base year of FY 2004. In the FY 2014 SNF PPS final rule (78 FR 47939), we last revised and rebased the SNF market basket, which included updating the base year from FY 2004 to FY 2010. For FY 2018, we proposed (82 FR 21029) to rebase the market basket to reflect 2014 Medicare-allowable total cost data (routine, ancillary, and capital-related) from freestanding SNFs and to revise applicable cost categories and price proxies used to determine the market basket. We proposed to maintain our policy of using data from freestanding SNFs, which represent 93 percent of the total SNFs shown in Table 26. We believe using freestanding MCR data, as opposed to the hospital-based SNF MCR data, for the proposed cost weight calculation is most appropriate because of the complexity of hospital-based data and the representativeness of the freestanding data. Hospital-based SNF expenses, are embedded in the hospital cost report. Any attempt to incorporate data from hospital-based facilities requires more complex calculations and assumptions regarding the ancillary costs related to the hospital-based SNF unit. We believe the use of freestanding SNF cost report data is technically appropriate for reflecting the cost structures of SNFs serving Medicare beneficiaries.

We proposed to use 2014 as the base year as we believe that the 2014 Medicare cost reports represented the most recent, complete set of Medicare cost report (MCR) data available to develop cost weights for SNFs at the time of rulemaking. The 2014 Medicare cost reports are for cost reporting periods beginning on and after October 1, 2013 and before October 1, 2014. While these dates appear to reflect fiscal year data, we note that a Medicare cost report that begins in this timeframe is generally classified as a “2014 cost report.” For example, we found that of the available 2014 Medicare cost reports for SNFs, approximately 7 percent had an October 1, 2013 begin date, approximately 70 percent of the reports had a January 1, 2014 begin date, and approximately 12 percent had a July 1, 2014 begin date. For this reason, and for the reasons explained below, we proposed to define the base year of the market basket as “2014-based” instead of “FY 2014-based”.

Specifically, we proposed to develop cost category weights for the 2014-based SNF market basket in two stages. First, we proposed to derive eight major expenditures or cost weights from the 2014 MCR data (CMS Form 2540-10) for freestanding SNFs: Wages and Salaries; Employee Benefits; Contract Labor; Pharmaceuticals; Professional Liability Insurance; Home Office Contract Labor; Capital-related; and a residual “All Other”. With the exception of the Home Office Contract Labor cost weight, these are the same cost categories calculated using the 2010 MCR data for the FY 2010-based SNF market basket. We provided a detailed discussion of our proposal to use the 2014 MCR data to determine the Home Office Contract Labor cost weight in section IV.A.1.a of the proposed rule and in section III.D.1.a of this final rule. The residual “All Other” category would reflect all remaining costs that are not captured in the other seven cost categories. Second, we proposed to divide the residual “All Other” cost category into subcategories using U.S. Department of Commerce Bureau of Economic Analysis' (BEA) 2007 Benchmark Input-Output (I-O) “use table before redefinitions, purchaser's value” for the Nursing and Community Care Facilities industry (NAICS 623A00) aged forward to 2014 using price changes. Furthermore, we proposed to continue to use the same overall methodology as was used for the FY 2010-based SNF market basket to develop the capital related cost weights of the 2014-based SNF market basket. We note that we are no longer referring to the market basket as a “FY 2014-based” market basket and instead refer to the market basket as simply “2014-based.” We proposed this change in naming convention for the market basket because the base year cost weight data for the proposed market basket do not reflect strictly fiscal year data. For example, the 2014-based SNF market basket uses Medicare cost report data and other government data that reflects fiscal year 2014, calendar year 2014, and state fiscal year 2014 expenses to determine the base year cost weights. Given that it is based on a mix of classifications of 2014 data, we proposed to refer to the market basket simply as “2014-based” as opposed to a “FY 2014-based” or “CY 2014-based”.

We refer readers to the FY 2018 SNF PPS proposed rule (82 FR 21029 through 21041) for a complete discussion of our proposals and associated rationale related to revising and rebasing the SNF market basket. We received a number of comments on the proposed revising and rebasing of the SNF market basket. A discussion of these comments, with our responses, appears throughout this section.

Comment: Several commenters supported the rebasing and revising of the SNF market basket from base year 2010 to base year 2014, stating that the weights for calculating the market basket update should reflect the most up-to-date cost data available. Other commenters requested that we meet with certain health care association representatives before we move forward with the proposed rebasing of the SNF market basket for FY 2018.

Response: We appreciate the commenters' support to rebase the market basket to 2014. We believe that it is reasonable and appropriate to rebase the market basket to 2014 as we believe this reflects the most complete and up-to-date cost data available. We note that we are available to meet with interested parties upon request to discuss their research and ideas for future rebasings.

Comment: Several commenters requested that we align the rebasing schedule of the SNF market basket with the acute inpatient hospital market basket rebasing schedule. They claimed that updating the SNF market basket schedule will improve the accuracy of the SNF market basket updates, particularly since the SNF wage index is directly linked to the hospital wage index. One commenter requested we provide information on ways to work collaboratively with the industry to develop an alternative approach to the SNF market basket methodology and to more appropriately update weights using more current data on a rolling basis. The commenter requested an explanation of why a chained index, which updates cost weights on a continual basis is not employed instead of a fixed-weight index approach.

Response: We appreciate the commenters' suggestion to align the rebasing schedule of the SNF market basket with the acute inpatient hospital market basket rebasing schedule. As discussed in the FY 2006 IPPS final rule (70 FR 47407), in accordance with section 404 of Public Law 108-173, we established a rebasing frequency of every four years for the IPPS hospital market basket. We last rebased the SNF market basket four years ago, reflecting a FY 2010 base year, in the FY 2014 SNF PPS final rule (78 FR 47939). We will continue to monitor the major cost share weights derived from the Medicare cost reports to evaluate whether a rebasing of the SNF market basket is necessary and may consider rebasing the SNF market basket consistent with the IPPS rebasing schedule.

In regards to the use of a fixed-weight index approach, we have found that healthcare provider cost share weights do not change substantially on an annual basis and, therefore, the use of a Laspeyres index formula, with base year weights updated on a regular basis (such as every few years), is technically appropriate for the CMS market baskets. In a 2008 paper, the CMS Office of the Actuary (OACT) investigated the impact of using an alternative price index formula on the inpatient hospital market basket and concluded that market basket rebasings more frequent than every 5 years would not result in any significant changes in update factors. This study also found that the use of an alternative index formula, such as a Paasche, Fisher, or Tornqvist, would not lead to an appreciable change to the results.

a. Development of Cost Categories and Weights

i. Use of Medicare Cost Report Data To Develop Major Cost Weights

To create a market basket that is representative of freestanding SNF providers serving Medicare patients and to help ensure accurate major cost weights (which is the percent of total Medicare allowable costs, as defined below), we proposed to apply edits to remove reporting errors and outliers. Specifically, the SNF Medicare cost reports used to calculate the market basket cost weights excluded any providers that reported costs less than or equal to zero for the following categories: Total facility costs; total operating costs; Medicare general inpatient routine service costs; and Medicare PPS payments. The final sample used included roughly 96 percent of those providers who submitted a Medicare cost report for 2014.

Additionally, for each of the major cost weights, except the Home Office Contract Labor cost weight (Wages and Salaries, Employee Benefits, Contract Labor, Pharmaceuticals, Professional Liability Insurance, and Capital-related Expenses) the data were trimmed to remove outliers (a standard statistical process) by: (1) Requiring that major expenses (such as Wages and Salaries costs) and total Medicare-allowable costs are greater than zero; and (2) excluding the top and bottom five percent of the major cost weight (for example, Wages and Salaries costs as a percent of total Medicare-allowable costs).

We note that in the FY 2018 SNF PPS proposed rule, we mistakenly referenced that we used the same trimming methodology for the Home Office Contract Labor cost weight that we used for the other major cost weights (a top and bottom five percent trimming methodology).

For the Home Office Contract Labor cost weight, we applied a one percent top-only trimming methodology. This allowed all providers' Medicare-allowable costs to be included, even if their home office contract labor costs were zero. We believe, as the Medicare cost report data (Worksheet S2 line 45) indicate, that not all SNF providers have a Home Office. Providers without a Home Office can incur these expenses directly by having their own staff, for which the costs would be included in the Wages and Salaries and Benefits cost weights. Alternatively, providers without a Home Office could also purchase related services from external contractors for which these expenses would be captured in the residual “All-Other” cost weight. We believe this one percent top-only trimming methodology is appropriate as it addresses outliers while allowing providers with zero Home Office Contract Labor costs to be included in the Home Office Contract Labor cost weight calculation. If we applied both top and bottom five percent trimming methodology we would exclude providers who have zero Home Office Contract Labor costs.

The major cost weight trimming process is done for each cost weight individually and, therefore, providers excluded from one cost weight calculation are not automatically excluded from other cost weight calculations. These were the same types of edits utilized for the FY 2010-based SNF market basket (with the exception of the Home Office Contract Labor cost weight which was not broken out using Medicare Cost Reports for the FY 2010 based SNF market basket), as well as other PPS market baskets (including but not limited to IPPS market basket and HHA market basket). We believe this trimming process improves the accuracy of the data used to compute the major cost weights by removing possible data misreporting.

Finally, the final weights of the proposed 2014-based SNF market basket were based on weighted means. For example, the final Wages and Salaries cost weight after trimming is equal to the sum of total Medicare-allowable wages and salaries divided by the sum of total Medicare-allowable costs. This methodology is consistent with the methodology used to calculate the FY 2010-based SNF market basket cost weights and other PPS market basket cost weights.

As stated above, the major cost weights of the 2014-based SNF market basket were derived from 2014 MCR data that is reported on CMS Form 2540-10, effective for freestanding SNFs with a cost reporting period beginning on or after December 1, 2010. The major cost weights for the FY 2010-based SNF market basket were derived from the 2010 MCR data that is reported on CMS Form 2540-96. CMS Form 2540-96 was effective for freestanding SNFs with cost reporting periods beginning on and after October 1, 1997. The OMB control number for both Form 2549-10 and Form 2540-96 is 0938-0463.

For all of the cost weights, we proposed to use Medicare allowable-total costs as the denominator (that is, Wages and Salaries cost weight = Wages and Salaries costs divided by Medicare-allowable total costs). Medicare-allowable total costs were proposed to be equal to total costs (after overhead allocation) from Worksheet B part 1, column 18, for lines 30, 40 through 49, 51, 52, and 71 plus Medicaid drug costs as defined below. We also proposed to include estimated Medicaid drug costs in the pharmacy cost weight, as well as the denominator for total Medicare-allowable costs. This is the same methodology used for the FY 2010-based SNF market basket and the FY 2004-based SNF market basket. The inclusion of Medicaid drug costs was finalized in the FY 2008 SNF PPS final rule (72 FR 43425 through 43430), and for the same reasons set forth in that final rule, we proposed to continue to use this methodology in the 2014-based SNF market basket.

We proposed that for the 2014-based SNF market basket we obtain costs for one new major cost category from the Medicare cost reports that was not used in the FY 2010-based SNF market basket—Home Office Contract Labor Costs.

We described the detailed methodology for obtaining costs for each of the eight major cost categories in section V.A.1.a. of the FY 2018 SNF PPS proposed rule (82 FR 21030) and below in section III.D.1.a. of this rule. The methodology used is similar to the methodology used in the FY 2010-based SNF market basket, as described in the FY 2014 SNF PPS final rule (78 FR 47940 through 47942).

(1) Wages and Salaries: To derive Wages and Salaries costs for the Medicare-allowable cost centers, we proposed first to calculate total unadjusted wages and salaries costs as reported on Worksheet S-3, part II, column 3, line 1. We then proposed to remove the wages and salaries attributable to non-Medicare-allowable cost centers (that is, excluded areas), as well as a portion of overhead wages and salaries attributable to these excluded areas. Excluded area wages and salaries were equal to wages and salaries as reported on Worksheet S-3, part II, column 3, lines 3, 4, and 7 through 11 plus nursing facility and non-reimbursable salaries from Worksheet A, column 1, lines 31, 32, 50, and 60 through 63.

Overhead wages and salaries are attributable to the entire SNF facility; therefore, we proposed to include only the proportion attributable to the Medicare-allowable cost centers. We proposed to estimate the proportion of overhead wages and salaries that is attributable to the non-Medicare-allowable costs centers (that is, excluded areas) by multiplying the ratio of excluded area wages and salaries (as defined above) to total wages and salaries as reported on Worksheet S-3, part II, column 3, line 1 by total overhead wages and salaries as reported on Worksheet S3, Part III, column 3, line 14. We used a similar methodology to derive wages and salaries costs in the FY 2010-based SNF market basket.

(2) Employee Benefits: We proposed Medicare-allowable employee benefits to be equal to total benefits as reported on Worksheet S-3, part II, column 3, lines 17 through 19 minus non-Medicare-allowable (that is, excluded area) employee benefits and minus a portion of overhead benefits attributable to these excluded areas. Non-Medicare-allowable employee benefits were derived by multiplying total excluded wages and salaries (as defined above in the ‘Wages and Salaries’ section) times the ratio of total benefit costs as reported on Worksheet S-3, part II, column 3, lines 17 through 19 to total wages and salary costs as reported on Worksheet S3, part II, column 3, line 1. Likewise, the portion of overhead benefits attributable to the excluded areas was derived by multiplying overhead wages and salaries attributable to the excluded areas (as defined in the ‘Wages and Salaries’ section) times the ratio of total benefit costs to total wages and salary costs (as defined above). We used a similar methodology in the FY 2010-based SNF market basket.

(3) Contract Labor: We proposed to derive Medicare-allowable contract labor costs from Worksheet S-3, part II, column 3, line 17. We note that in the FY 2018 SNF PPS proposed rule (82 FR 21030), we mistakenly referenced line 17. These costs are actually reported in Worksheet S-3, part II, column 3, line 14 as per the CMS Form 2540-10 instructions (which reflects costs for contracted direct patient care services, that is, nursing, therapeutic, rehabilitative, or diagnostic services furnished under contract rather than by employees and management contract services). We note that the processing of the data was correct. We used Worksheet S-3, part II, column 3, line 14 in our analysis. Our written description in the proposed rule of the line we used was, however, incorrect.

(4) Pharmaceuticals: We proposed to calculate pharmaceuticals costs using the non-salary costs from the Pharmacy cost center (Worksheet B, part I, column 0, line 11 less Worksheet A, column 1, line 11) and the Drugs Charged to Patients' cost center (Worksheet B, part I, column 0, line 49 less Worksheet A, column 1, line 49). Since these drug costs were attributable to the entire SNF and not limited to Medicare-allowable services, we proposed to adjust the drug costs by the ratio of Medicare-allowable pharmacy total costs (Worksheet B, part I, column 11, for lines 30, 40 through 49, 51, 52, and 71) to total pharmacy costs from Worksheet B, part I, column 11, line 11. Worksheet B, part I allocates the general service cost centers, which are often referred to as “overhead costs” (in which pharmacy costs are included) to the Medicare-allowable and non-Medicare-allowable cost centers. This adjustment was made for those providers who reported Pharmacy cost center expenses. Otherwise, we assumed the non-salary Drugs Charged to Patients costs were Medicare-allowable.

Second, similar to the FY 2010-based SNF market basket, we proposed to continue to adjust the drug expenses reported on the MCR to include an estimate of total Medicaid drug costs, which are not represented in the Medicare-allowable drug cost weight. Similar to the FY 2010-based SNF market basket, we estimated Medicaid drug costs based on data representing dual-eligible Medicaid beneficiaries. Medicaid drug costs were estimated by multiplying Medicaid dual-eligible drug costs per day times the number of Medicaid days as reported in the Medicare-allowable skilled nursing cost center (Worksheet S3, part I, column 5, line 1) in the SNF MCR. Medicaid dual-eligible drug costs per day (where the day represents an unduplicated drug supply day) were estimated using a sample of 2014 Part D claims for those dual-eligible beneficiaries who had a Medicare SNF stay during the year. Medicaid dual-eligible beneficiaries would receive their drugs through the Medicare Part D benefit, which would work directly with the pharmacy and, therefore, these costs would not be represented in the Medicare SNF MCRs. A random twenty percent sample of Medicare Part D claims data yielded a Medicaid drug cost per day of $19.62. We note that the FY 2010-based SNF market basket also relied on data from the Part D claims, which yielded a dual-eligible Medicaid drug cost per day of $17.39 for 2010.

Provided below are summaries of the comments we received related to the Pharmaceuticals cost category, as well as our responses.

Comment: One commenter was concerned with the lower Pharmaceuticals cost weight in the 2014-based SNF market basket compared to the 2010-based SNF market basket. They were unable to explain the decrease given their experience with annual pharmaceutical price increases and the introduction of new pharmaceuticals.

Several commenters also had specific concerns regarding the methodology utilized to determine the Pharmaceuticals cost weight. The commenters stated that the vast majority of SNFs did not report costs on the cost report line for the “Pharmacy” department. They stated that only a small number of SNFs have in-house Pharmacies and that those SNFs were used as a proxy for the pharmaceutical costs for all SNFs; one commenter requested an alternative method.

Several commenters were also concerned by the addition of estimated Part D medication costs to the “Drugs Charged to Patients” data reported on Row 49 of the cost report. The commenter questioned why this type of “gross up” was not, as far as they could tell, applied to any of the other ancillary cost centers.

Response: The methodology used to determine the cost weights in the 2014-based SNF market basket and 2010-based SNF market basket is the same. The change in the Pharmaceuticals cost weight in the 2014-based SNF market basket (7.3 percent) from the FY 2010-based SNF market basket (7.9 percent) is a function of the growth rate of pharmaceutical expenses relative to other components of the market basket over this time period. Our own internal analysis shows increasing drug costs from FY 2010 to FY 2014; however, during this time period, pharmaceutical costs increased at a slower rate than other components of the market basket—such as capital and contract labor expenses. This relative comparison resulted in a decrease in the Pharmaceuticals cost weight of 0.6 percentage point between the FY 2010-based SNF market basket and 2014-based SNF market basket (7.9 percent to 7.3 percent) while the capital cost weight increased 0.5 percentage point (7.4 percent to 7.9 percent) and contract labor grew 1.3 percentage points (5.5 percent to 6.8 percent). It is also important to consider that the increase in pharmaceutical costs over this period reflects changes in both the price of prescription drugs, proxied by the Producer Price Index for Prescription Drugs, as well the quantity and intensity of prescriptions. Our analysis of the data shows that the decrease in the Pharmaceuticals cost weight was consistent, in aggregate, across urban and rural status SNFs as well as across for-profit, government, and nonprofit ownership type SNFs.

As stated above and in the FY 2018 SNF PPS proposed rule (82 FR 21030 through 21031), we proposed to calculate pharmaceutical costs using the non-salary costs reported in the Pharmacy cost center (Worksheet B, part I, column 0, line 11 less Worksheet A, column 1, line 11) and the Drugs Charged to Patients' cost center (Worksheet B, part I, column 0, line 49 less Worksheet A, column 1, line 49), hereafter referred to as total MCR drug costs. Since these drug costs were attributable to the entire SNF and not limited to Medicare-allowable services, we proposed to adjust the drug costs by the ratio of Medicare-allowable pharmacy total costs (Worksheet B, part I, column 11, for lines 30, 40 through 49, 51, 52, and 71) to total pharmacy costs (Worksheet B, part I, column 11, line 11).

We understand the commenter's concern regarding the adjustment to the total MCR drug costs using the Pharmacy cost center as only 20 percent of providers reported Pharmacy cost center expenses. We are clarifying that the adjustment was only applied to those 20 percent of providers who reported Pharmacy costs. We assumed that all of the drug costs were Medicare-allowable for the remaining 80 percent of providers. We added a clarifying sentence in the Pharmacy cost weight calculation of this final rule. Applying this adjustment had only a marginal impact on the drug cost weight (lowering it by only 0.1 percentage point). As a sensitivity, we also derived an alternative by using the ratio of Skilled Nursing Facility days (as reported on Worksheet S3, part 1, column 7 line 1) to Total Facility days. This would result in a Pharmaceuticals cost weight of 7.1 percent compared to the 2014-based cost weight of 7.3 percent.

As stated in the proposed rule (82 FR 21031), the 2014-based SNF market basket included an adjustment to the drug expenses reported on the MCR to include an estimate of total Medicaid drug costs, which are not represented in the Medicare-allowable drug cost weight. As stated above, the 2014-based SNF market basket reflects total Medicare allowable costs (that is, total costs for all payers for those services reimbursable under the SNF PPS). For the FY 2006-based SNF market basket (72 FR 43426), commenters noted that the total pharmaceutical costs reported on the MCR did not include pharmaceutical costs for dual-eligible Medicaid patients as these were directly reimbursed by Medicaid. Since all of the other cost category weights reflect Medicaid patients (including the compensation costs for dispersing these drugs), we made an adjustment to include these drug expenses. The pharmaceutical cost weight using only 2014 MCR data without any adjustments is 3.0 percent, compared to the proposed Pharmaceuticals cost weight (including the adjustment for Medicaid dual-eligible drug costs) of 7.3 percent.

Comment: One commenter requested further explanation on how Part D drug costs were incorporated into the Pharmaceuticals cost weight. They questioned how the 20 percent sample was selected and the rationale for selecting this population to estimate non-SNF Medicaid drug costs. They questioned if there were analytics to support these decisions and also requested clarification for why the drug costs for patients with a SNF stay would be comparable to patients in a nursing facility that had not had a hospitalization during the year. They also questioned whether the Part D claims were matched to the SNF stay and if Part D claims for the SNF stay were excluded. They further questioned which cost variables in Part D claims were used, how the costs per day were calculated and the rationale for producing this estimate.

Response: As stated previously in this section, the 2014-based SNF market basket reflects total Medicare allowable costs (that is, total costs for all payers for those services reimbursable under the SNF PPS). For the FY 2006-based SNF market basket (72 FR 43426), commenters noted that the total pharmaceutical costs reported on the MCR did not include pharmaceutical costs for dual-eligible Medicaid patients as these were directly reimbursed by Medicaid. Since all of the other cost category weights reflect Medicaid patients (including the compensation costs for dispensing these drugs), we made an adjustment to include these Medicaid drug expenses so the market basket cost weights would be calculated consistently.

For the 2014-based SNF market basket, as stated in the FY 2018 SNF PPS proposed rule (82 FR 21031), we estimated Medicaid drug costs by multiplying Medicaid dual-eligible drug costs per day times the number of Medicaid days as reported in the Medicare-allowable skilled nursing facility cost center (Worksheet S3, part I, column 5, line 1) on the SNF MCR. The Medicaid dual-eligible drug costs per day (where the day represents an unduplicated drug supply day) were estimated using a random 20 percent sample of 2014 Part D claims for those dual-eligible beneficiaries who had a Medicare SNF stay during the year. We believe this sample is a reasonable proxy for total drug costs per day for Medicaid patients residing in a skilled nursing unit under a Medicaid stay. Our analysis of the Part D claims data shows that dual-eligible beneficiaries have higher drug costs per day than “non-duals” and that dual-eligible beneficiaries who have had a SNF Part A stay during the year have higher drug costs per day ($19.62) compared to those dual-eligible beneficiaries with no SNF Part A stay during the year ($14.82).

The total drug costs per unduplicated day represented all drug costs incurred during the 2014 calendar year for those dual-eligible beneficiaries with a SNF Medicare stay during that 2014 calendar year. Therefore, they include drug costs incurred during the Medicaid SNF stay occurring in the 2014 calendar year. The total drug costs from the Part D claims includes the drug ingredient cost, the dispensing fee, vaccine administration fee and sales tax. We used a 20 percent sample of Part D claims (approximately 287 million claims) where claims were randomly selected based on the beneficiary ID number.

Comment: One commenter stated that they see an increase in the number of Veterans being served by SNFs. They further stated that Medicare patients, if they were admitted to a non-VA nursing home, would use their Medicare benefit. However, in a VA home, the commenter claimed that the patient would use their VA benefit which covers the drug costs—and not the nursing home. The commenter concluded that there would be many drug costs that are not represented on the cost report that traditionally would have been. The commenter requested clarification on how we will address this challenge.

Response: We appreciate the commenter raising this concern. We believe the current methodology and resulting Pharmaceutical cost weight is reasonable, in part because VA costs would not have a significant impact on the market basket cost weights (according to the CMS National Health Expenditure Accounts, VA spending accounted for roughly 3 percent of total Nursing Care Facilities and Continuing Care Retirement Communities expenditures in 2014). However, in the future we plan to monitor this issue in more depth to ensure the market basket is adequately capturing the appropriate costs.

(5) Professional Liability Insurance: We proposed to calculate the professional liability insurance costs from Worksheet S-2 of the MCRs as the sum of premiums; paid losses; and self-insurance (Worksheet S-2, column 1 through 3, line 41).

Provided below are summaries of the comments we received related to the Professional Liability Insurance cost category, as well as our responses.

Comment: One commenter stated that we should calculate a weight for professional liability insurance considering other data sources. As an example, the commenter provided a link to AHCA's Aon Professional Liability Study stating that the 2016 report documents a significant and continual increase in professional liability costs.

Response: We thank the commenter for providing the link to this study. As stated in the FY 2018 SNF proposed rule (82 FR 21031), the professional liability insurance cost weight is derived using data from Worksheet S-2 of the Medicare Cost Reports. These data represent the sum of premiums, paid losses, and self-insurance (Worksheet S-2, column 1 through 3, line 41). We continue to believe that using these data submitted by SNFs on the Medicare cost report represent the best data source to derive the professional liability insurance cost weight. We will continue to evaluate other data sources, including the study provided by the commenter, to obtain additional information regarding professional liability insurance costs for SNFs.

(6) Capital-Related: We proposed to derive the Medicare-allowable capital-related costs from Worksheet B, part II, column 18 for lines 30, 40 through 49, 51, 52, and 71.

(7) Home Office Contract Labor Costs: We proposed to calculate Medicare-allowable home office contract labor costs by multiplying total home office contract labor costs (as reported on Worksheet S3, part 2, column 3, line 16) times the ratio of Medicare-allowable operating costs (Medicare-allowable total costs less Medicare-allowable capital costs) to total operating costs (equal to Worksheet B, part I, column 18, line 100 less Worksheet B, part I, column 0, line 1 and 2).

(8) All Other (residual): We proposed to calculate the “All Other” cost weight as a residual, calculated by subtracting the major cost weights (Wages and Salaries, Employee Benefits, Contract Labor, Pharmaceuticals, Professional Liability Insurance, Home Office Contract Labor, and Capital-Related) from 100.

Provided below are summaries of the general comments we received related to the major cost category weights, as well as our responses.

Comment: One commenter noted that the decrease in cost weights related to wages, benefits, contract labor, and pharmaceuticals from FY 2010 to the proposed base year of 2014 did not reflect, in any way, their experience. For geographic locations that have a large proportion of staff whose wages and benefits are driven by collective bargaining agreements, such as the NY metropolitan area where providers have seen regular cost increases over the 4 years, the commenter claimed that the decrease in cost weight does not make sense.

Response: The purpose of the SNF market basket is to measure the price inflation facing average SNFs serving Medicare beneficiaries at the national level. A change in the Wages and Salaries cost weight is a function of the growth rate of Wages and Salaries expenses relative to other components of the market basket, based on data directly supplied to CMS by SNFs. We would further note that differences in wage and wage-related costs among geographic regions are accounted for by the application of the wage index.

Comment: One commenter requested we show the numerators and denominators for the calculation of each weight so that it is possible to comment on any bias that may be introduced by exclusions.

Response: We disagree with the commenter's suggestion that we should provide the numerators and denominators for the calculation as we do not believe this would allow commenters to determine whether any bias may be introduced by exclusions. Rather, we believe that the detailed description of the data (specifically the Medicare cost report worksheet fields) and trimming methodologies allow the commenter to evaluate the bias. Specifically, commenters are able to evaluate the accuracy and reasonableness of the Medicare cost report worksheet fields. They are also able to replicate the results and then compare the trimmed cost share weight samples to the national average distribution of total costs. We reiterate that in deriving the proposed SNF cost weights, we used a similar trimming methodology for each of the major cost weights, with the exception of the Home Office Contract Labor cost weight as discussed earlier in this final rule (as we explained, for the Home Office Contract Labor cost weight, we used an alternative methodology). Our review of the trimmed samples for each of the major cost weights (Wages and Salaries, Employee Benefits, Contract Labor, Professional Liability, Home Office Contract Labor, Pharmaceuticals and Capital) resulted in a total cost distribution that was similar to the cost distribution of the untrimmed sample when compared by urban/rural status, ownership-type (for-profit, nonprofit, or government) and then by census region. We would further note that, as stated above, the trimming of the individual cost weights was done independently of each other, in an effort to produce the most representative data for each of the major cost weights. Finally, we would note that the 5 percent trim is the same methodology used to derive cost share weights (with the exception of the Home Office Contract Labor cost weight) for other CMS market baskets.

Comment: One commenter questioned whether it is time to possibly make some revisions to Worksheet A of the Medicare SNF cost report. They provided suggested additional cost categories that they believe would help construct a more accurate market basket and to account for regional fluctuations (for example, utility costs, property insurance rates, etc).

Response: The commenter's specific detailed recommendations for changes to the Medicare cost report are outside the scope of the FY 2018 SNF PPS proposed rule. However, we appreciate and will consider the commenter's suggestion to capture additional information on the SNF Medicare cost report for possible future use in the SNF market basket.

Comment: One commenter had several questions on the methodology used to develop the major cost weights of the 2014-based SNF market basket. The commenter specifically questioned our trimming methods and whether we excluded partial-year cost reports (that is, providers with cost report data of less than 12 months). They also stated there was no information provided regarding the treatment of missing data in the cost report fields and that zero and missing data do not have the same meaning. They further stated that missing data was high for certain weights with over 40 percent of cost reports having missing values for professional liability insurance, over 70 percent of cost reports having missing values in home office contract labor costs, and over 80 percent having missing values in the Pharmacy cost center used to determine the Pharmaceuticals cost weight.

Response: We appreciate the commenter's review of the methodology used to develop the 2014-based SNF market basket. We made no edits to remove providers with partial cost reporting periods and, therefore, they were included in the initial set of cost reports. In response to this comment, we examined the impact of excluding those providers that reported costs for a period of fewer than 270 days (representing about 3/4 of the cost reporting year) and, similar to the commenter's finding, found that its impact on the major cost weights was minimal with less than 0.1 percentage point in absolute terms. Given its small impact, we do not believe it is necessary to revise the 2014-based SNF market basket to reflect the exclusion of reports with a partial cost reporting period; however, we will consider the merits of this edit for future rebasings.

In regards to the commenter's request for information on the treatment of missing data in the cost report fields, CMS receives Medicare cost report data via the Electronic Cost Reporting file from the Medicare Administrative Contractor. These files do not have missing values for numeric fields; therefore, fields are zero or greater. The public-use files provided on the CMS Web site, however, convert the zero values to missing or null.

We recognize the commenter's concern of providers' reporting zero Professional Liability and Pharmaceutical costs. As stated, in the FY 2018 SNF PPS proposed rule (82 FR 21030), for each of the major cost weights, except for Home Office Contract Labor as discussed above, (that is (Wages and Salaries, Employee Benefits, Contract Labor, Pharmaceuticals, Professional Liability Insurance, Home Office Contract Labor, and Capital-related Expenses) the data were trimmed to remove outliers (a standard statistical process) by first requiring that major expenses and total Medicare-allowable costs are greater than zero. For these major cost weights (Wages and Salaries, Employee Benefits, Contract Labor, Professional Liability, Capital and Pharmaceuticals), we believe that providers should incur these expenses to provide SNF services to beneficiaries. Therefore, cost reports with zero costs for major expenses (except Home Office Contract Labor costs) were excluded from the market basket cost weight calculation before trimming the top and bottom five percent. We note, as stated in the proposed rule, the trimming method is done for each cost weight individually and, therefore, providers excluded from one cost weight calculation are not automatically excluded from other cost weight calculations. This methodology allows us to use the largest possible sample of providers that report expenses for any given category.

However, as discussed earlier, we do not believe, as the Medicare cost report data (Worksheet S2, line 45) indicates, that all SNF providers will have a Home Office and then will also “purchase” services from their home office. Rather, providers can incur these expenses directly by having their own staff, for which the costs would be included in the Wages and Salaries and Benefits cost weights, or be purchased from contractors that are not directly affiliated with SNF, for which these expenses would be captured in the residual “All-Other” cost weight. Therefore, as discussed above, for the Home Office Contract Labor cost weight, we instead applied a one percent top trimming methodology but allowed all providers' Medicare-allowable costs to be included, even if their home office contract labor costs were zero.

Also, we included all data for subcategories of the major cost weights, except Home Office Contract Labor costs, (such as excluded area salaries component of the Wages and Salaries costs) even if they are zero as we believe it is reasonable for some of these specific costs to not be applicable to some providers. We must rely on the data that are submitted by providers and always encourage providers to fill out the cost report forms using the most accurate and complete data available to them.

Comment: One commenter made note of their inability to replicate all of the proposed cost weights using the methodology provided in the proposed rule. Specifically, the commenter was unable to replicate the Contract Labor cost weight and Home Office Contract Labor cost weight.

Response: We appreciate the commenter's review of our methodology and their replication efforts. We note that in the FY 2018 SNF PPS proposed rule, we made an error in the description of which Medicare cost report line is used to determine the Medicare allowable contract labor costs. The proposed rule stated that Medicare allowable contract labor costs would be equal to Worksheet S-3, part II, column 3, line 17, which reflects costs for contracted direct patient care services, that is, nursing, therapeutic, rehabilitative, or diagnostic services furnished under contract, rather than by employees and management contract services. These Medicare allowable contract labor costs are actually reported in Worksheet S-3, part II, column 3, line 14 as per the CMS Form 2540-10 instructions. We note that the processing of the data was correct, and we appropriately used Worksheet S-3, part II, column 3, line 14, but our written description of the line used was not. We apologize for any confusion and have corrected this typographical error in this final rule.

As stated above, in the FY 2018 SNF PPS proposed rule, we mistakenly indicated that we used the same trimming methodology for the Home Office Contract Labor cost weight that we used for the other major cost weights (a top and bottom five percent trimming method). For the Home Office Contract Labor cost weight we applied a one percent top-only trimming methodology. This trimming methodology allowed all providers' Medicare-allowable costs to be included, even if their home office contract labor costs were zero. We believe this one percent trimming methodology is appropriate for the Home Office Contract Labor cost weight as it addresses outliers while allowing providers with zero Home Office Contract Labor costs to be included in the Home Office Contract Labor cost weight calculation. Applying a five percent top and bottom trimming methodology would exclude providers who have zero Home Office Contract Labor costs.

After consideration of the public comments we received, for the reasons discussed above and in the FY 2018 SNF PPS proposed rule, we are finalizing the major cost weights as proposed, without modification. Table 9 below shows the major cost categories and their respective cost weights as derived from the Medicare cost reports for this final rule.

The Wages and Salaries and Employee Benefits cost weights as calculated directly from the Medicare cost reports decreased by 1.8 and 1.2 percentage points, respectively, while the Contract Labor cost weight increased 1.3 percentage points between the FY 2010-based SNF market basket and 2014-based SNF market basket. The decrease in the Wages and Salaries occurred among most cost centers and in aggregate for the General Service (overhead) and Inpatient Routine Service cost centers, which together account for about 80 percent of total facility costs.

As we did for the FY 2010-based SNF market basket (78 FR 26452), we proposed to allocate contract labor costs to the Wages and Salaries and Employee Benefits cost weights based on their relative proportions under the assumption that contract labor costs are comprised of both wages and salaries and employee benefits. The contract labor allocation proportion for wages and salaries is equal to the Wages and Salaries cost weight as a percent of the sum of the Wages and Salaries cost weight and the Employee Benefits cost weight. Using the 2014 Medicare cost report data, this percentage is 83 percent; therefore, we proposed to allocate approximately 83 percent of the Contract Labor cost weight to the Wages and Salaries cost weight and 17 percent to the Employee Benefits cost weight. For the FY 2010-based SNF market basket, the wages and salaries to employee benefit ratio was 81/19 percent.

We did not receive public comments on our proposed allocation of contract labor costs to Wages and Salaries and Employee Benefits. For the reasons discussed above and in the FY 2018 SNF PPS proposed rule, we are finalizing the allocation methodology and percentages as proposed, without modification. Table 10 below shows the Wages and Salaries and Employee Benefits cost weights after contract labor allocation for the FY 2010-based SNF market basket and the 2014-based SNF market basket.

ii. Derivation of the Detailed Operating Cost Weights

To further divide the “All Other” residual cost weight estimated from the 2014 Medicare cost report data into more detailed cost categories, we proposed to use the 2007 BenchmarkI-O “Use Tables/Before Redefinitions/Purchaser Value” for Nursing and Community Care Facilities industry (NAICS 623A00), published by the Census Bureau's Bureau of Economic Analysis (BEA). These data are publicly available at the following Web site: http://www.bea.gov/industry/io_annual.htm. The BEA Benchmark I-O data are generally scheduled for publication every 5 years with the most recent data available for 2007. The 2007 Benchmark I-O data are derived from the 2007 Economic Census and are the building blocks for BEA's economic accounts. Therefore, they represent the most comprehensive and complete set of data on the economic processes or mechanisms by which output is produced and distributed. BEA also produces Annual I-O estimates. However, while based on a similar methodology, these estimates reflect less comprehensive and less detailed data sources and are subject to revision when benchmark data become available. Instead of using the less detailed Annual I-O data, we proposed to inflate the 2007 Benchmark I-O data aged forward to 2014 by applying the annual price changes from the respective price proxies to the appropriate market basket cost categories that are obtained from the 2007 Benchmark I-O data. We repeated this practice for each year. We then calculated the cost shares that each cost category represents of the 2007 data inflated to 2014. These resulting 2014 cost shares were applied to the “All Other” residual cost weight to obtain the detailed cost weights for the proposed 2014-based SNF market basket. For example, the cost for Food: Direct Purchases represents 13.7 percent of the sum of the “All Other” 2007 Benchmark I-O Expenditures inflated to 2014. Therefore, the Food: Direct Purchases cost weight represents 3.1 percent of the proposed 2014-based SNF market basket's “All Other” cost category (0.137 × 22.6 percent = 3.1 percent). For the FY 2010-based SNF market basket (78 FR 26456), we used the same methodology utilizing the 2002 Benchmark I-O data (aged to FY 2010).

Using this methodology, we proposed to derive 21 detailed SNF market basket operating cost category weights from the proposed 2014-based SNF market basket “All Other” residual cost weight (22.6 percent). These categories are: (1) Fuel: Oil and Gas; (2) Electricity; (3) Water and Sewerage; (4) Food: Direct Purchases; (5) Food: Contract Services; (6) Chemicals; (7) Medical Instruments and Supplies; (8) Rubber and Plastics; (9) Paper and Printing Products; (10) Apparel; (11) Machinery and Equipment; (12) Miscellaneous Products; (13) Professional Fees: Labor-Related; (14) Administrative and Facilities Support Services; (15) Installation, Maintenance, and Repair Services; (16) All Other: Labor-Related Services; (17) Professional Fees: Nonlabor-Related; (18) Financial Services; (19) Telephone Services; (20) Postage; and (21) All Other: Nonlabor-Related Services.

We note that the machinery and equipment expenses are for equipment that is paid for in a given year and not depreciated over the asset's useful life. Depreciation expenses for movable equipment are reflected in the capital component of the proposed 2014-based SNF market basket (described in section V.A.1.c. of the proposed rule (82 FR 21032) and section III.D.1.c. of this final rule).

We would also note that for ease of reference we proposed to rename the Nonmedical Professional Fees: Labor-Related and Nonmedical Professional Fees: Nonlabor-related cost categories (as labeled in the FY 2010-based SNF market basket) to be Professional Fees: Labor-Related and Professional Fees: Nonlabor-Related in the 2014-based SNF market basket. These cost categories still represent the same nonmedical professional fees that were included in the FY 2010-based SNF market basket, which we describe in section V.A.4. of the proposed rule (82 FR 21039) and section III.D.1.d. of this final rule.

For the 2014-based SNF market basket, we proposed to include a separate cost category for Installation, Maintenance, and Repair Services to proxy these costs by a price index that better reflects the price changes of labor associated with maintenance-related services. Previously these costs were included in the All Other: Labor-Related Services category of the FY 2010-based SNF market basket.

Provided below are summaries of the comments we received regarding the derivation of the detailed operating cost weights, as well as our responses.

Comment: Several commenters believe a SNF cost distribution study from 2007 is out-of-date and not likely to represent the distribution of cost in 2014 or going forward. For example, according to the commenter, operational changes driven by the Requirements of Participation will have substantial impacts. The commenter stated that the function of a market basket is to update SNF payment based on real changes in cost over time. The commenter claimed that the use of a static 2007 study is inconsistent with the fundamental intent of the market basket. The commenter requested information regarding how CMS could gather more current data on SNF costs.

Response: To further divide the “All Other” residual cost weight of 22.6 percent into more detailed cost categories, we proposed to use the 2007 Benchmark I-O for Nursing and Community Care Facilities industry (NAICS 623A00). For each of the detailed expenses (such as food: Direct purchase), we inflate the 2007 expense to 2014 using the relevant price proxies. The resulting 2014 cost shares based on these inflated expenses were applied to the “All Other” residual cost weight to obtain the detailed cost weights for the 2014-based SNF market basket.

Thus, our methodology does in fact reflect changes in expenses from 2007 to 2014, but is based on the assumption that the change in quantities over this period is equal to the change in prices. We believe this is a reasonable assumption as it is consistent with historical data which shows the cost shares changing over time. We believe this is a better methodology for developing the market basket rather than keeping the shares fixed between 2007 and 2014 or proxying the “All Other” residual by an aggregate index such as the CPI All-Items, which would not reflect the unique cost structures of SNFs.

It is not until late 2018, when BEA is expected to release 2012 BenchmarkI-O data, that we will be able to determine whether the growth in quantities for these specific costs grew similarly to prices over this period, as we currently assume in the market basket. We will evaluate these data and consider its inclusion for the development of the SNF market basket in the future.

After consideration of the public comments we received, for the reasons discussed above and in the FY 2018 SNF PPS proposed rule, we are finalizing the detailed operating cost weights and methodology for deriving such weights as proposed, without modification.

iii. Derivation of the Detailed Capital Cost Weights

Similar to the FY 2010-based SNF market basket, we proposed to further divide the Capital-related cost weight into: Depreciation, Interest, Lease and Other Capital-related cost weights.

We proposed to calculate the depreciation cost weight (that is, depreciation costs excluding leasing costs) using depreciation costs from Worksheet S-2, column 1, lines 20 and 21. Since the depreciation costs reflect the entire SNF facility (Medicare and non-Medicare-allowable units), we proposed to use total facility capital costs as the denominator. This methodology assumes that the depreciation of an asset is the same regardless of whether the asset was used for Medicare or non-Medicare patients. This methodology yielded depreciation as a percent of capital costs of 27.3 percent for 2014. We then applied this percentage to the proposed 2014-based SNF market basket Medicare-allowable Capital-related cost weight of 7.9 percent, yielding a Medicare-allowable depreciation cost weight (excluding leasing expenses, which is described in more detail below) of 2.2 percent. To further disaggregate the Medicare-allowable depreciation cost weight into fixed and moveable depreciation, we proposed to use the 2014 SNF MCR data for end-of-the-year capital asset balances as reported on Worksheet A7. The 2014 SNF MCR data showed a fixed/moveable split of 83/17. The FY 2010-based SNF market basket, which utilized the same data from the FY 2010 MCRs, had a fixed/moveable split of85/15.

We also proposed to derive the interest expense share of capital-related expenses from 2014 SNF MCR data, specifically from Worksheet A, column 2, line 81. Similar to the depreciation cost weight, we proposed to calculate the interest cost weight using total facility capital costs. This methodology yielded interest as a percent of capital costs of 27.4 percent for 2014. We then applied this percentage to the proposed 2014-based SNF market basket Medicare-allowable Capital-related cost weight of 7.9 percent, yielding a Medicare-allowable interest cost weight (excluding leasing expenses) of 2.2 percent. As done with the last SNF market basket rebasing (78 FR 26454), we proposed to determine the split of interest expense between for-profit and not-for-profit facilities based on the distribution of long-term debt outstanding by type of SNF (for-profit or not-for-profit/government) from the 2014 SNF MCR data. We estimated the split between for-profit and not-for-profit interest expense to be 27/73 percent compared to the FY 2010-based SNF market basket with 41/59 percent.

Because the detailed data were not available in the MCRs, we proposed to use the most recent 2014 Census Bureau Service Annual Survey (SAS) data to derive the capital-related expenses attributable to leasing and other capital-related expenses. The FY 2010-based SNF market basket used the 2010 SAS data. Based on the 2014 SAS data, we determined that leasing expenses are 63 percent of total leasing and capital-related expenses costs. In the FY 2010-based SNF market basket, leasing costs represent 62 percent of total leasing and capital-related expenses costs. We then applied this percentage to the proposed 2014-based SNF market basket residual Medicare-allowable capital costs of 3.6 percent derived from subtracting the Medicare-allowable depreciation cost weight and Medicare-allowable interest cost weight from the 2014-based SNF market basket of total Medicare-allowable capital cost weight (7.9 percent − 2.2 percent − 2.2 percent = 3.6 percent). This produced the proposed 2014-based SNF Medicare-allowable leasing cost weight of 2.3 percent and all-other capital-related cost weight of 1.3 percent.

Lease expenses are not broken out as a separate cost category in the SNF market basket, but are distributed among the cost categories of depreciation, interest, and other capital-related expenses, reflecting the assumption that the underlying cost structure and price movement of leasing expenses is similar to capital costs in general. As was done with past SNF market baskets and other PPS market baskets, we assumed 10 percent of lease expenses are overhead and proposed to assign them to the other capital-related expenses cost category. This is based on the assumption that leasing expenses include not only depreciation, interest, and other capital-related costs but also additional costs paid to the lessor. We distributed the remaining lease expenses to the three cost categories based on the proportion of depreciation, interest, and other capital-related expenses to total capital costs, excluding lease expenses.

We did not receive any public comments on our proposed methodology for deriving the detailed capital cost weights. Therefore, for the reasons discussed above and in the FY 2018 SNF PPS proposed rule, we are finalizing the detailed capital cost weights and methodology as proposed, without modification.

Table 11 shows the capital-related expense distribution (including expenses from leases) in the final 2014-based SNF market basket and the FY 2010-based SNF market basket.

Table 12 presents the final 2014-based SNF market basket and the FY 2010-based SNF market basket.

b. Price Proxies Used To Measure Operating Cost Category Growth

After developing the 30 cost weights for the 2014-based SNF market basket, we selected the most appropriate wage and price proxies currently available to represent the rate of change for each expenditure category. With four exceptions (three for the capital-related expenses cost categories and one for Professional Liability Insurance (PLI)), we base the wage and price proxies on Bureau of Labor Statistics (BLS) data, and group them into one of the following BLS categories:

• Employment Cost Indexes: Employment Cost Indexes (ECIs) measure the rate of change in employment wage rates and employer costs for employee benefits per hour worked. These indexes are fixed-weight indexes and strictly measure the change in wage rates and employee benefits per hour. ECIs are superior to Average Hourly Earnings (AHE) as price proxies for input price indexes because they are not affected by shifts in occupation or industry mix, and because they measure pure price change and are available by both occupational group and by industry. The industry ECIs are based on the 2004 North American Classification System (NAICS).
• Producer Price Indexes: Producer Price Indexes (PPIs) measure price changes for goods sold in other than retail markets. PPIs are used when the purchases of goods or services are made at the wholesale level.
• Consumer Price Indexes: Consumer Price Indexes (CPIs) measure change in the prices of final goods and services bought by consumers. CPIs are only used when the purchases are similar to those of retail consumers rather than purchases at the wholesale level, or if no appropriate PPI were available.

We evaluated the price proxies using the criteria of reliability, timeliness, availability, and relevance. Reliability indicates that the index is based on valid statistical methods and has low sampling variability. Widely accepted statistical methods ensure that the data were collected and aggregated in a way that can be replicated. Low sampling variability is desirable because it indicates that the sample reflects the typical members of the population. (Sampling variability is variation that occurs by chance because only a sample was surveyed rather than the entire population.) Timeliness implies that the proxy is published regularly, preferably at least once a quarter. The market baskets are updated quarterly, and therefore, it is important for the underlying price proxies to be up-to-date, reflecting the most recent data available. We believe that using proxies that are published regularly (at least quarterly, whenever possible) helps to ensure that we are using the most recent data available to update the market basket. We strive to use publications that are disseminated frequently, because we believe that this is an optimal way to stay abreast of the most current data available. Availability means that the proxy is publicly available. We prefer that our proxies are publicly available because this will help ensure that our market basket updates are as transparent to the public as possible. In addition, this enables the public to be able to obtain the price proxy data on a regular basis. Finally, relevance means that the proxy is applicable and representative of the cost category weight to which it is applied. The CPIs, PPIs, and ECIs that we have selected to propose in this regulation meet these criteria. Therefore, we believe that they continue to be the best measure of price changes for the cost categories to which they would be applied.

Table 15 in the proposed rule (82 FR 21039) lists all price proxies for the 2014-based SNF market basket. Below is a detailed explanation of the proposed price proxies used for each operating cost category.

• Wages and Salaries: We proposed to use the ECI for Wages and Salaries for Private Industry Workers in Nursing Care Facilities (NAICS 6231; BLS series code CIU2026231000000I) to measure price growth of this category. NAICS 623 includes facilities that provide a mix of health and social services, with many of the health services being largely some level of nursing services. Within NAICS 623 is NAICS 6231, which includes nursing care facilities primarily engaged in providing inpatient nursing and rehabilitative services. These facilities, which are most comparable to Medicare-certified SNFs, provide skilled nursing and continuous personal care services for an extended period of time, and, therefore, have a permanent core staff of registered or licensed practical nurses. This is the same index used in the FY 2010-based SNF market basket.
• Employee Benefits: We proposed to use the ECI for Benefits for Nursing Care Facilities (NAICS 6231) to measure price growth of this category. The ECI for Benefits for Nursing Care Facilities is calculated using BLS's total compensation (BLS series ID CIU2016231000000I) for nursing care facilities series and the relative importance of wages and salaries within total compensation. We believe this constructed ECI series is technically appropriate for the reason stated above in the Wages and Salaries price proxy section. This is the same index used in the FY 2010-based SNF market basket.
• Electricity: We proposed to use the PPI Commodity for Commercial Electric Power (BLS series code WPU0542) to measure the price growth of this cost category. This is the same index used in the FY 2010-based SNF market basket.
• Fuel: Oil and Gas: We proposed to change the proxy used for the Fuel: Oil and Gas cost category. The FY 2010-based SNF market basket uses the PPI Commodity for Commercial Natural Gas (BLS series code WPU0552) to proxy these expenses. For the 2014-based SNF market basket, we proposed to use a blend of the PPI Industry for Petroleum Refineries (BLS series code PCU32411-32411) and the PPI Commodity for Natural Gas (BLS series code WPU0531). Our analysis of the Bureau of Economic Analysis' 2007 Benchmark I-O data for Nursing and Community Care Facilities shows that petroleum refineries expenses accounts for approximately 65 percent and natural gas accounts for approximately 35 percent of the fuel: Oil and gas expenses. Therefore, we proposed a blended proxy of 65 percent of the PPI Industry for Petroleum Refineries (BLS series code PCU32411-32411) and 35 percent of the PPI Commodity for Natural Gas (BLS series code WPU0531). We believe that these two price proxies are the most technically appropriate indices available to measure the price growth of the Fuel: Oil and Gas category in the 2014-based SNF market basket.
• Water and Sewerage: We proposed to use the CPI All Urban for Water and Sewerage Maintenance (BLS series code CUUR0000SEHG01) to measure the price growth of this cost category. This is the same index used in the FY 2010-based SNF market basket.
• Professional Liability Insurance: We proposed to use the CMS Hospital Professional Liability Insurance Index to measure price growth of this category. We were unable to find a reliable data source that collects SNF-specific PLI data. Therefore, we proposed to use the CMS Hospital Professional Liability Index, which tracks price changes for commercial insurance premiums for a fixed level of coverage, holding non-price factors constant (such as a change in the level of coverage). This is the same index used in the FY 2010-based SNF market basket. We believe this is an appropriate proxy to measure the price growth associated of SNF professional liability insurance as it captures the price inflation associated with other medical institutions that serve Medicare patients.
• Pharmaceuticals: We proposed to use the PPI Commodity for Pharmaceuticals for Human Use, Prescription (BLS series code WPUSI07003) to measure the price growth of this cost category. This is the same index used in the FY 2010-based SNF market basket.
• Food: Wholesale Purchases: We proposed to use the PPI Commodity for Processed Foods and Feeds (BLS series code WPU02) to measure the price growth of this cost category. This is the same index used in the FY 2010-based SNF market basket.
• Food: Retail Purchase: We proposed to use the CPI All Urban for Food Away From Home (All Urban Consumers) (BLS series code CUUR0000SEFV) to measure the price growth of this cost category. This is the same index used in the FY 2010-based SNF market basket.
• Chemicals: For measuring price change in the Chemicals cost category, we proposed to use a blended PPI composed of the Industry PPIs for Other Basic Organic Chemical Manufacturing (NAICS 325190) (BLS series code PCU32519-32519), Soap and Cleaning Compound Manufacturing (NAICS 325610) (BLS series code PCU32561-32561), and Other Miscellaneous Chemical Product Manufacturing (NAICS 3259A0) (BLS series code PCU325998325998).

Using the 2007 Benchmark I-O data, we found that these three NAICS industries accounted for approximately 96 percent of SNF chemical expenses. The remaining four percent of SNF chemical expenses are for three other incidental NAICS chemicals industries such as Paint and Coating Manufacturing. We proposed to create a blended index based on those three NAICS chemical expenses listed above that account for 96 percent of SNF chemical expenses. We proposed to create this blend based on each NAICS' expenses as a share of their sum. These expenses as a share of their sum are listed in Table 34.

The FY 2010-based SNF market basket also used a blended chemical proxy that was based on 2002 Benchmark I-O data. We believe our proposed chemical blended index for the 2014-based SNF market basket is technically appropriate as it reflects more recent data on SNFs purchasing patterns. Table 13 in the proposed rule (82 FR 21035) provided the weights for the 2014-based blended chemical index and the FY 2010-based blended chemical index. The table is also shown below.

As discussed below, we are finalizing the weights for the 2014-based blended chemical index as proposed, without modification.

• Medical Instruments and Supplies: We proposed to use a blend for the Medical Instruments and Supplies cost category. The 2007 Benchmark I-O data shows an approximate 60/40 split between `Medical and Surgical Appliances and Supplies' and `Surgical and Medical Instruments'. Therefore, we proposed a blend composed of 60 percent of the PPI Commodity for Medical and Surgical Appliances and Supplies (BLS series code WPU1563) and 40 percent of the PPI Commodity for Surgical and Medical Instruments (BLS series code WPU1562).

The FY 2010-based SNF market basket used the single, higher level PPI Commodity for Medical, Surgical, and Personal Aid Devices (BLS series code WPU156). We believe that the proposed price proxy better reflects the mix of expenses for this cost category as obtained from the 2007 Benchmark I-O data.

• Rubber and Plastics: We proposed to use the PPI Commodity for Rubber and Plastic Products (BLS series code WPU07) to measure price growth of this cost category. This is the same index used in the FY 2010-based SNF market basket.
• Paper and Printing Products: We proposed to use the PPI Commodity for Converted Paper and Paperboard Products (BLS series code WPU0915) to measure the price growth of this cost category. This is the same index used in the FY 2010-based SNF market basket.
• Apparel: We proposed to use the PPI Commodity for Apparel (BLS series code WPU0381) to measure the price growth of this cost category. This is the same index used in the FY 2010-based SNF market basket.
• Machinery and Equipment: We proposed to use the PPI Commodity for Machinery and Equipment (BLS series code WPU11) to measure the price growth of this cost category. This is the same index used in the FY 2010-based SNF market basket.
• Miscellaneous Products: For measuring price change in the Miscellaneous Products cost category, we proposed to use the PPI Commodity for Finished Goods less Food and Energy (BLS series code WPUFD4131). Both food and energy are already adequately represented in separate cost categories and should not also be reflected in this cost category. This is the same index used in the FY 2010-based SNF market basket.
• Professional Fees: Labor-Related: We proposed to use the ECI for Total Compensation for Private Industry Workers in Professional and Related (BLS series code CIU2010000120000I) to measure the price growth of this category. This is the same index used in the FY 2010-based SNF market basket (which was called the Nonmedical Professional Fees: Labor-Related cost category).
• Administrative and Facilities Support Services: We proposed to use the ECI for Total Compensation for Private Industry Workers in Office and Administrative Support (BLS series code CIU2010000220000I) to measure the price growth of this category. This is the same index used in the FY 2010-based SNF market basket.
• Installation, Maintenance and Repair Services: We proposed to include a separate cost category for Installation, Maintenance, and Repair Services to proxy these costs by a price index that better reflects the price changes of labor associated with maintenance-related services. We proposed to use the ECI for Total Compensation for All Civilian Workers in Installation, Maintenance, and Repair (BLS series code CIU1010000430000I) to measure the price growth of this new cost category. Previously these costs were included in the All Other: Labor-Related Services category and were proxied by the ECI for Total Compensation for Private Industry Workers in Service Occupations (BLS series code CIU2010000300000I).
• All Other: Labor-Related Services: We proposed to use the ECI for Total Compensation for Private Industry Workers in Service Occupations (BLS series code CIU2010000300000I) to measure the price growth of this cost category. This is the same index used in the FY 2010-based SNF market basket.

Professional Fees: NonLabor-Related: We proposed to use the ECI for Total Compensation for Private Industry Workers in Professional and Related (BLS series code CIU2010000120000I) to measure the price growth of this category. This is the same index used in the FY 2010-based SNF market basket (which was called the Nonmedical Professional Fees: Nonlabor-Related cost category).

Financial Services: We proposed to use the ECI for Total Compensation for Private Industry Workers in Financial Activities (BLS series code CIU201520A000000I) to measure the price growth of this cost category. This is the same index used in the FY 2010-based SNF market basket.

Telephone Services: We proposed to use the CPI All Urban for Telephone Services (BLS series code CUUR0000SEED) to measure the price growth of this cost category. This is the same index used in the FY 2010-based SNF market basket.

• Postage: We proposed to use the CPI All Urban for Postage (BLS series code CUUR0000SEEC) to measure the price growth of this cost category. This is the same index used in the FY 2010-based SNF market basket.
• All Other: NonLabor-Related Services: We proposed to use the CPI All Urban for All Items Less Food and Energy (BLS series code CUUR0000SA0L1E) to measure the price growth of this cost category. This is the same index used in the FY 2010-based SNF market basket.

We did not receive any public comments on our proposed price proxies for each of the operating cost categories. For the reasons discussed above and in the FY 2018 SNF PPS proposed rule, we are finalizing the price proxies of the operating cost categories as proposed, without modification. In addition, we did not receive any public comments on our proposed weights for the 2014-based blended chemical index. Thus, for the reasons discussed above and in the FY 2018 SNF PPS proposed rule, we are finalizing the weights for 2014-based blended chemical index as proposed, without modification.

c. Price Proxies Used To Measure Capital Cost Category Growth

We proposed to apply the same price proxies as were used in the FY 2010-based SNF market basket, and below is a detailed explanation of the price proxies used for each capital cost category. We also proposed to continue to vintage weight the capital price proxies for Depreciation and Interest to capture the long-term consumption of capital. This vintage weighting method is the same method that was used for the FY 2010-based SNF market basket and is described below.

• Depreciation—Building and Fixed Equipment: We proposed to use the BEA Chained Price Index for Private Fixed Investment in Structures, Nonresidential, Hospitals and Special Care (BEA Table 5.4.4. Price Indexes for Private Fixed Investment in Structures by Type). This BEA index is intended to capture prices for construction of facilities such as hospitals, nursing homes, hospices, and rehabilitation centers.
• Depreciation—Movable Equipment: We proposed to use the PPI Commodity for Machinery and Equipment (BLS series code WPU11). This price index reflects price inflation associated with a variety of machinery and equipment that would be utilized by SNFs including but not limited to medical equipment, communication equipment, and computers.
• Nonprofit Interest: We proposed to use the average yield on Municipal Bonds (Bond Buyer 20-bond index).
• For-Profit Interest: We proposed to use the average yield on Moody's AAA corporate bonds (Federal Reserve). We proposed different proxies for the interest categories because we believe interest price pressures differ between nonprofit and for-profit facilities.
• Other Capital: Since this category includes fees for insurances, taxes, and other capital-related costs, we proposed to use the CPI All Urban for Owners' Equivalent Rent of Primary Residence (BLS series code CUUR0000SEHC01), which would reflect the price growth of these costs.

We believe that these price proxies continue to be the most appropriate proxies for SNF capital costs that meet our selection criteria of relevance, timeliness, availability, and reliability.

As stated above, we proposed to continue to vintage weight the capital price proxies for Depreciation and Interest to capture the long-term consumption of capital. To capture the long-term nature, the price proxies are vintage-weighted; and the vintage weights are calculated using a two-step process. First, we determined the expected useful life of capital and debt instruments held by SNFs. Second, we identified the proportion of expenditures within a cost category that is attributable to each individual year over the useful life of the relevant capital assets, or the vintage weights.

We proposed to rely on Bureau of Economic Analysis (BEA) fixed asset data to derive the useful lives of both fixed and movable capital, which is the same data source used to derive the useful lives for the FY 2010-based SNF market basket. The specifics of the data sources used are explained below.

i. Calculating Useful Lives for Moveable and Fixed Assets

Estimates of useful lives for movable and fixed assets for the 2014-based SNF market basket are 10 and 23 years, respectively. These estimates are based on three data sources from the BEA: (1) Current-cost average age; (2) historical-cost average age; and (3) industry-specific current cost net stocks of assets.

BEA current-cost and historical-cost average age data by asset type are not available by industry but are published at the aggregate level for all industries. The BEA does publish current-cost net capital stocks at the detailed asset level for specific industries. There are 61 detailed movable assets (including intellectual property) and there are 32 detailed fixed assets in the BEA estimates. Since we seek aggregate useful life estimates applicable to SNFs, we developed a methodology to approximate movable and fixed asset ages for nursing and residential care services (NAICS 623) using the published BEA data. For the proposed FY 2014 SNF market basket, we used the current-cost average age for each asset type from the BEA fixed assets Table 2.9 for all assets and weight them using current-cost net stock levels for each of these asset types in the nursing and residential care services industry, NAICS 6230. (For example, nonelectro medical equipment current-cost net stock (accounting for about 37 percent of total moveable equipment current-cost net stock in 2014) is multiplied by an average age of 4.7 years. Current-cost net stock levels are available for download from the BEA Web site at http://www.bea.gov/national/FA2004/Details/Index.html. We then aggregated the “weighted” current-cost net stock levels (average age multiplied by current-cost net stock) into moveable and fixed assets for NAICS 6230. We then adjusted the average ages for moveable and fixed assets by the ratio of historical-cost average age (Table 2.10) to current-cost average age (Table 2.9).

This produced historical cost average age data for movable (equipment and intellectual property) and fixed (structures) assets specific to NAICS 6230 of 4.8 and 11.6 years, respectively. The average age reflects the average age of an asset at a given point in time, whereas we want to estimate a useful life of the asset, which would reflect the average over all periods an asset is used. To do this, we multiplied each of the average age estimates by two to convert to average useful lives with the assumption that the average age is normally distributed (about half of the assets are below the average at a given point in time, and half above the average at a given point in time). This produced estimates of likely useful lives of 9.6 and 23.2 years for movable and fixed assets, which we rounded to 10 and 23 years, respectively. We proposed an interest vintage weight time span of 21 years, obtained by weighting the fixed and movable vintage weights (23 years and 10 years, respectively) by the fixed and movable split (87 percent and 13 percent, respectively). This is the same methodology used for the FY 2010-based SNF market basket which had useful lives of 22 years and 6 years for fixed and moveable assets, respectively. The impact of revising the useful life for moveable assets from 6 years to 10 years had little to no impact on the growth rate of the 2014-based SNF market basket capital cost weight. Over the 2014 to 2026 time period, the impact on the growth rate of the capital cost weight was no larger than 0.01 percent in absolute terms.

ii. Constructing Vintage Weights

Given the expected useful life of capital (fixed and moveable assets) and debt instruments, we then must determine the proportion of capital expenditures attributable to each year of the expected useful life for each of the three asset types: Building and fixed equipment, moveable equipment, and interest. These proportions represent the vintage weights. We were not able to find a historical time series of capital expenditures by SNFs. Therefore, we proposed to approximate the capital expenditure patterns of SNFs over time, using alternative SNF data sources. For building and fixed equipment, we used the stock of beds in nursing homes from the National Nursing Home Survey (NNHS) conducted by the National Center for Health Statistics (NCHS) for 1962 through 1999. For 2000 through 2010, we extrapolated the 1999 bed data forward using a 5-year moving average of growth in the number of beds from the SNF MCR data. For 2011 to 2014, we proposed to extrapolate the 2010 bed data forward using the average growth in the number of beds over the 2011 to 2014 time period. We then used the change in the stock of beds each year to approximate building and fixed equipment purchases for that year. This procedure assumes that bed growth reflects the growth in capital-related costs in SNFs for building and fixed equipment. We believe that this assumption is reasonable because the number of beds reflects the size of a SNF, and as a SNF adds beds, it also likely adds fixed capital.

As was done for the FY 2010-based SNF market basket (as well as prior market baskets), we proposed to estimate moveable equipment purchases based on the ratio of ancillary costs to routine costs. The time series of the ratio of ancillary costs to routine costs for SNFs measures changes in intensity in SNF services, which are assumed to be associated with movable equipment purchase patterns. The assumption here is that as ancillary costs increase compared to routine costs, the SNF caseload becomes more complex and would require more movable equipment. The lack of movable equipment purchase data for SNFs over time required us to use alternative SNF data sources. A more detailed discussion of this methodology was published in the FY 2008 SNF final rule (72 FR 43428). We believe the resulting two time series, determined from beds and the ratio of ancillary to routine costs, reflect real capital purchases of building and fixed equipment and movable equipment over time.

To obtain nominal purchases, which are used to determine the vintage weights for interest, we converted the two real capital purchase series from 1963 through 2014 determined above to nominal capital purchase series using their respective price proxies (the BEA Chained Price Index for Nonresidential Construction for Hospitals & Special Care Facilities and the PPI for Machinery and Equipment). We then combined the two nominal series into one nominal capital purchase series for 1963 through 2014. Nominal capital purchases are needed for interest vintage weights to capture the value of debt instruments.

Once we created these capital purchase time series for 1963 through 2014, we averaged different periods to obtain an average capital purchase pattern over time: (1) For building and fixed equipment, we averaged 30, 23-year periods; (2) for movable equipment, we averaged 43, 10-year periods; and (3) for interest, we averaged 32, 21-year periods. We calculate the vintage weight for a given year by dividing the capital purchase amount in any given year by the total amount of purchases during the expected useful life of the equipment or debt instrument. To provide greater transparency, we posted on the CMS market basket Web site at http://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/MedicareProgramRatesStats/MarketBasketResearch.html,, an illustrative spreadsheet that contains an example of how the vintage-weighted price indexes are calculated.

We did not receive any public comments on our proposed price proxies used for each of the detailed capital cost categories or on our methodology for deriving the vintage weights. For the reasons discussed above and in the FY 2018 SNF PPS proposed rule, we are finalizing the price proxies of the capital cost categories, the vintage weights, and the methodology for deriving the vintage weights, as proposed without modification.

The vintage weights for the 2014-based SNF market basket and the FY 2010-based SNF market basket are presented in Table 14.

Table 15 shows all the price proxies for the final 2014 based SNF market basket.

c. Labor-Related Share

We define the labor-related share (LRS) as those expenses that are labor-intensive and vary with, or are influenced by, the local labor market. Each year, we calculate a revised labor-related share based on the relative importance of labor-related cost categories in the input price index. Effective beginning with FY 2018, we proposed to revise and update the labor-related share to reflect the relative importance of the 2014-based SNF market basket cost categories that we believe are labor-intensive and vary with, or are influenced by, the local labor market. For the proposed 2014-based SNF market basket, these are: (1) Wages and Salaries (including allocated contract labor costs as described above); (2) Employee Benefits (including allocated contract labor costs as described above); (3) Professional fees: Labor-related; (4) Administrative and Facilities Support Services; (5) Installation, Maintenance, and Repair services; (6) All Other: Labor-Related Services; and (7) a proportion of capital-related expenses. We proposed to continue to include a proportion of capital-related expenses because a portion of these expenses are deemed to be labor-intensive and vary with, or are influenced by, the local labor market. For example, a proportion of construction costs for a medical building would be attributable to local construction workers' compensation expenses.

Consistent with previous SNF market basket revisions and rebasings, the All Other: Labor-related services cost category is mostly comprised of building maintenance and security services (including, but not limited to, landscaping services, janitorial services, waste management services, and investigation and security services). Because these services tend to be labor-intensive and are mostly performed at the SNF facility (and therefore, unlikely to be purchased in the national market), we believe that they meet our definition of labor-related services.

The proposed inclusion of the Installation, Maintenance, and Repair Services cost category into the labor-related share remains consistent with the current labor-related share, since this cost category was previously included in the FY 2010-based SNF market basket All Other: Labor-related Services cost category. We proposed to establish a separate Installation, Maintenance, and Repair Services cost category so that we can use the ECI for Total Compensation for All Civilian Workers in Installation, Maintenance, and Repair to reflect the specific price changes associated with these services. We also use this cost category in the 2012-based IRF market basket (80 FR 47059), 2012-based IPF market basket (80 FR 46667), and 2013-based LTCH market basket (81 FR 57091).

As discussed in the FY 2014 SNF PPS proposed rule (78 FR 26462), in an effort to determine more accurately the share of nonmedical professional fees (included in the 2014-based SNF market basket Professional Fees cost categories) that should be included in the labor-related share, we surveyed SNFs regarding the proportion of those fees that are attributable to local firms and the proportion that are purchased from national firms. Based on these weighted results, we determined that SNFs purchase, on average, the following portions of contracted professional services inside their local labor market:

• 78 percent of legal services.
• 86 percent of accounting and auditing services.
• 89 percent of architectural, engineering services.
• 87 percent of management consulting services.

Together, these four categories represent 3.3 percentage points of the total costs for the 2014-based SNF market basket. We applied the percentages from this special survey to their respective SNF market basket weights to separate them into labor-related and nonlabor-related costs. As a result, we proposed to designate 2.8 percentage points of the 3.3 percentage points to the labor-related share, with the remaining 0.5 percentage point is categorized as nonlabor-related.

For the proposed 2014-based SNF market basket, we conducted a similar analysis of home office data. The Medicare cost report CMS Form 2540-10 requires a SNF to report information regarding their home office provider. Approximately 57 percent of SNFs reported some type of home office information on their Medicare cost report for 2014 (for example, city, state, zip code). Using the data reported on the Medicare cost report, we compared the location of the SNF with the location of the SNF's home office. For the FY 2010-based SNF market basket, we used the Medicare HOMER database to determine the location of the provider's home office as this information was not available on the Medicare cost report CMS Form 2540-96. For the 2014-based SNF market basket, we proposed to determine the proportion of home office contract labor costs that should be allocated to the labor-related share based on the percent of total SNF home office contract labor costs as reported in Worksheet S-3, Part II attributable to those SNFs that had home offices located in their respective local labor markets—defined as being in the same Metropolitan Statistical Area (MSA). We determined a SNF's and home office's MSAs using their zip code information from the Medicare cost reports.

Using this methodology, we determined that 28 percent of SNFs' home office contract labor costs were for home offices located in their respective local labor markets. Therefore, we proposed to allocate 28 percent of home office expenses to the labor-related share. The FY 2010-based SNF market basket allocated 32 percent of home office expenses to the labor-related share.

In the proposed 2014-based SNF market basket, home office expenses that were subject to allocation based on the home office allocation methodology represent 0.7 percent of the 2014-based SNF market basket. Based on the home office results, we proposed to apportion 0.2 percentage point of the 0.7 percentage point figure into the labor-related share (0.7 × 0.28 = 0.193, or 0.2) and designate the remaining 0.5 percentage point as nonlabor-related. Therefore, based on the two allocations mentioned above, we proposed to apportion 3.0 percentage points into the labor-related share. This amount is added to the portion of professional fees that we continue to identify as labor-related using the I-O data such as contracted advertising and marketing costs (0.8 percentage point of total operating costs) resulting in a Professional Fees: Labor-Related cost weight of 3.8 percent.

We did not receive any public comments on our proposed methodology for deriving the labor-related share. For the reasons discussed above and in the FY 2018 SNF PPS proposed rule, we are finalizing our proposals, without modification, as discussed above to update and revise the labor-related share effective October 1, 2017, to reflect the relative importance of the following 2014-based SNF market basket cost weights that we believe are labor-intensive and vary with, or are influenced by, the local labor market: (1) Wages and Salaries (including allocated contract labor costs as described above); (2) Employee Benefits (including allocated contract labor costs as described above); (3) Professional fees: Labor-related; (4) Administrative and Facilities Support Services; (5) Installation, Maintenance, and Repair services; (6) All Other: Labor-Related Services; and (7) a proportion of capital-related expenses.

Table 16 compares the 2014-based labor-related share and the FY 2010-based labor-related share based on the relative importance of IGI's most recent second quarter 2017 forecast with historical data through the first quarter of 2017. The FY 2018 SNF PPS proposed rule (82 FR 21040) reflected IGI's first quarter 2017 forecast with historical data through the fourth quarter of 2016. As stated in the FY 2018 SNF PPS proposed rule (82 FR 21019), our policy has been that, if more recent data becomes available (for example, a more recent estimate of the SNF market basket and/or MFP adjustment), we would use such data, if appropriate, to determine the SNF market basket percentage change, labor-related share relative importance, forecast error adjustment, and MFP adjustment in the SNF PPS final rule.

We note that in Table 16 of the FY 2018 SNF PPS proposed rule (82 FR 21041), we misreported the FY 2017 labor-related share as 69.1 percent (this was the FY 2016 labor-related share (80 FR 46402)). The FY 2017 labor-related share was 68.8 percent as finalized in the FY 2017 SNF PPS final rule (81 FR 51979, 51980). We present the FY 2017 labor-related share in Table 16 below.

The FY 2018 SNF labor-related share (LRS) is 2.0 percentage points higher than the FY 2017 SNF LRS, which is based on the FY 2010-based SNF market basket relative importance. This implies an increase in the quantity of the labor-related services because rebasing the index contributed significantly to the increase. Also contributing to the higher labor-related share is a higher capital-related cost weight in the 2014-based SNF market basket compared to the FY 2010-based SNF market basket. As stated above, we include a proportion of capital-related expenses in the labor-related share as we believe a portion of these expenses (such as construction labor costs) are deemed to be labor-intensive and vary with, or are influenced by, the local labor market.

d. Market Basket Estimate for the FY 2018 SNF PPS Update

As discussed previously in this final rule, beginning with the FY 2018 SNF PPS update, we are adopting the 2014-based SNF market basket as the appropriate market basket of goods and services for the SNF PPS. Based on IHS Global Inc.'s (IGI) second quarter 2017 forecast with historical data through the first quarter of 2017, the most recent estimate of the 2014-based SNF market basket for FY 2018 is 2.6 percent. As stated above, the FY 2018 SNF PPS proposed rule reflected IGI's first quarter 2017 forecast with historical data through the fourth quarter of 2016. IGI is a nationally recognized economic and financial forecasting firm that contracts with CMS to forecast the components of CMS' market baskets.

Table 17 compares the 2014-based SNF market basket and the FY 2010-based SNF market basket percent changes. For the historical period between FY 2013 and FY 2016, the average difference between the two market baskets is −0.3 percentage point. This is primarily the result of the lower pharmaceuticals cost category weight, increased Fuel: Oil and Gas cost category weight, and the change in the Fuels price proxy. For the forecasted period between FY 2017 and FY 2019, there is no difference in the average growth rate.

While we ordinarily would adopt the use of this 2014-based SNF market basket percentage to update the SNF PPS per diem rates for FY 2018, we note that section 411(a) of the MACRA amended section 1888(e) of the Act to add section 1888(e)(5)(B)(iii) of the Act, which establishes a special rule for FY 2018 that requires the market basket percentage, after the application of the productivity adjustment, to be 1.0 percent. In accordance with section 1888(e)(5)(B)(iii) of the Act, we will use a market basket percentage of 1.0 percent to update the federal rates set forth in this final rule. We proposed to use the 2014-based SNF market basket to determine the market basket percentage update for the SNF PPS per diem rates effective FY 2019. For the reasons discussed above and in the FY 2018 SNF PPS proposed rule, we are finalizing our proposal to use the 2014-based SNF market basket to determine the market basket percentage update for the SNF PPS per diem rates, effective FY 2019. In addition, as stated in section III.D.1.d. in this preamble, we are adopting the use of the 2014-based SNF market basket to determine the labor-related share effective October 1, 2017.

2. Skilled Nursing Facility (SNF) Quality Reporting Program (QRP)

a. Background and Statutory Authority

Section 1888(e)(6)(A)(i) of the Act, as added by section 2(c)(4) of the Improving Medicare Post-Acute Care Transformation Act of 2014 (IMPACT Act), requires that for fiscal years beginning with FY 2018, in the case of a SNF that does not submit data as applicable in accordance with sections 1888(e)(6)(B)(i)(II) and (III) of the Act for a fiscal year, the Secretary reduce the market basket percentage described in section 1888(e)(5)(B)(i) of the Act for payment rates during that fiscal year by two percentage points. In section III.B.2. of this final rule, we discuss revisions in the market basket update regulations at § 413.337(d) that will implement this provision. In accordance with this statutory mandate, we have implemented a SNF Quality Reporting Program (QRP), which we believe promotes higher quality and more efficient health care for Medicare beneficiaries. The SNF QRP applies to freestanding SNFs, SNFs affiliated with acute care facilities, and all non-CAH swing-bed rural hospitals. We refer readers to the FY 2016 SNF PPS final rule (80 FR 46427 through 46429) for a full discussion of the statutory background and policy considerations that have shaped the SNF QRP.

When we use the term “FY (year)SNF QRP,” we are referring to the fiscal year for which the SNF QRP requirements applicable to that fiscal year must be met in order for a SNF to receive the full market basket percentage when calculating the payment rates applicable to it for that fiscal year.

The IMPACT Act (Pub. L. 113-185) amended Title XVIII of the Act, in part, by adding a new section 1899B that requires the Secretary to establish new data reporting requirements for certain post-acute care (PAC) providers, including SNFs. Specifically, new sections 1899B(a)(1)(A)(ii) and (iii) of the Act require SNFs, inpatient rehabilitation facilities (IRFs), Long Term Care Hospitals (LTCHs), and home health agencies (HHAs), under the provider-type's respective quality reporting program (which, for SNFs, is found at section 1888(e)(6) of the Act), to report data on quality measures specified under section 1899B(c)(1) of the Act for at least five domains, and data on resource use and other measures specified under section 1899B(d)(1) of the Act for at least three domains. Section 1899B(a)(1)(A)(i) of the Act further requires each of these PAC provider-types to report under its respective quality reporting program standardized resident assessment data in accordance with subsection (b), for at least the quality measures specified under subsection (c)(1), and that is for at least five specific categories: Functional status; cognitive function and mental status; special services, treatments, and interventions; medical conditions and co-morbidities; and impairments. Section 1899B(a)(1)(B) of the Act requires that all of the data that must be reported in accordance with section 1899B(a)(1)(A) of the Act be standardized and interoperable to allow for the exchange of the information among PAC providers and other providers and the use of such data to enable access to longitudinal information and to facilitate coordinated care. We refer readers to the FY 2016 SNF PPS final rule (80 FR 46427 through 46429) for additional information on the IMPACT Act and its applicability to SNFs.

b. General Considerations Used for Selection of Quality Measures for the SNF QRP

We refer readers to the FY 2016 SNF PPS final rule (80 FR 46429 through 46431) for a detailed discussion of the considerations we apply in measure selection for the SNF QRP, such as alignment with the CMS Quality Strategy, which incorporates the three broad aims of the National Quality Strategy.

As part of our consideration for measures for use in the SNF QRP, we review and evaluate measures that have been implemented in other programs and take into account measures that have been endorsed by NQF for provider settings other than the SNF setting. We have previously adopted measures that we referred to as “applications” of those measures. We have received questions pertaining to the term “application” and want to clarify that when we refer to a proposed or implemented measure as an “application of” the measure, we mean that the measure will be used in the SNF setting, rather than the setting for which it was endorsed by the NQF. For example, in the FY 2016 SNF PPS final rule (80 FR 46440 through 46444), we adopted a measure entitled Application of Percent of Residents Experiencing One or More Falls With Major Injury (Long Stay) (NQF #0674), which is currently endorsed for the nursing home setting but not for the SNF setting. For such measures, we intend to seek NQF endorsement for the SNF setting, and if the NQF endorses one or more of them, we will update the title of the measure to remove the reference to “application”.

We received several comments generally related to the proposed measures, the IMPACT Act, NQF endorsement, and training needs. The comments and our responses are discussed below.

Comment: A few commenters expressed concern that CMS has not provided a timeline for seeking NQF endorsement for non-NQF-endorsed quality measures in the SNF QRP. One commenter expressed further concern that non-NQF-endorsed measures may be implemented before undergoing adequate testing, as required for NQF endorsement. Another commenter expressed concern regarding the adequacy of resources allocated to complete necessary testing and obtain consensus endorsement for measures as required by the IMPACT Act. All commenters commenting on this topic requested further information from CMS regarding the process and timeline for seeking NQF endorsement.

Response: We recognize that the NQF endorsement process is an important part of measure development and plan to submit non-NQF-endorsed quality measures in the SNF QRP adopted in this rule for NQF endorsement as soon as feasible, with an intended timeframe of 2018. With regard to adequate testing prior to implementation, we wish to note that we engage in multiple testing activities prior to measure implementation. These activities include testing of items and measures in their intended settings, public posting of measure testing data, when possible, seeking public comment on measures in the various stages of their development, and utilization of technical expert input on measure development, including expert evaluation of the validity and importance of measures. We interpret the commenter's comment regarding the adequacy of the resources necessary to obtain consensus endorsement as efforts to engage stakeholders. We believe that we commit an adequate level of resources to the measure development process and the NQF endorsement process. Such resources are outlined above and include engaging in pilot testing with providers, seeking public comment, convening TEPs, and engaging subject matter experts to provide feedback throughout the measure development process.

Comment: One commenter recommended aligning the SNF QRP quality measures with other CMS initiatives such as the Financial Alignment Initiative, the value-based payment program and the Medicaid managed care initiatives under the Section 1115 waiver authorities.

Response: We acknowledge the value of aligning the SNF QRP measures to other CMS initiatives and we will seek to align measures with other initiatives in an effort to reduce provider burden where feasible.

(1) Measuring and Accounting for Social Risk Factors in the SNF QRP

In, the FY 2018 SNF PPS proposed rule (82 FR 21042 through 21043), we discussed accounting for social risk factors in the SNF QRP. We stated that we consider related factors that may affect measures in the SNF QRP. We understand that social risk factors such as income, education, race and ethnicity, employment, disability, community resources, and social support (certain factors of which are also sometimes referred to as socioeconomic status (SES) factors or socio-demographic status (SDS) factors) play a major role in health. One of our core objectives is to improve beneficiary outcomes including reducing health disparities, and we want to ensure that all beneficiaries, including those with social risk factors, receive high quality care. In addition, we seek to ensure that the quality of care furnished by providers and suppliers is assessed as fairly as possible under our programs while ensuring that beneficiaries have adequate access to excellent care.

We have been reviewing reports prepared by the Office of the Assistant Secretary for Planning and Evaluation (ASPE) and the National Academies of Sciences, Engineering, and Medicine on the issue of measuring and accounting for social risk factors in CMS' quality measurement and payment programs, and considering options on how to address the issue in these programs. On December 21, 2016, ASPE submitted a Report to Congress on a study it was required to conduct under section 2(d) of the IMPACT Act. The study analyzed the effects of certain social risk factors of Medicare beneficiaries on quality measures and measures of resource use used in one or more of nine Medicare value-based purchasing programs. The report also included considerations for strategies to account for social risk factors in these programs. In a January 10, 2017 report released by The National Academies of Sciences, Engineering, and Medicine, that body provided various potential methods for measuring and accounting for social risk factors, including stratified public reporting.

In addition, the NQF undertook a 2-year trial period in which new measures, measures undergoing maintenance review, and measures endorsed with the condition that they enter the trial period were assessed to determine whether risk adjustment for selected social risk factors was appropriate for these measures. This trial entailed temporarily allowing inclusion of social risk factors in the risk-adjustment approach for these measures. The trial has concluded and NQF will issue recommendations on the future inclusion of social risk factors in risk adjustment for quality measures.

As we continue to consider the analyses and recommendations from these reports and await the recommendations of the NQF trial on risk adjustment for quality measures, we are continuing to work with stakeholders in this process. As we have previously communicated, we are concerned about holding providers to different standards for the outcomes of their patients with social risk factors because we do not want to mask potential disparities or minimize incentives to improve the outcomes for disadvantaged populations. Keeping this concern in mind, while we sought input on this topic previously, we continue to seek public comment on whether we should account for social risk factors in measures in the SNF QRP, and if so, what method or combination of methods would be most appropriate for accounting for social risk factors. Examples of methods include: Confidential reporting to providers of measure rates stratified by social risk factors, public reporting of stratified measure rates, and potential risk adjustment of a particular measure as appropriate based on data and evidence.

In addition, in the FY 2018 SNF PPS proposed rule (82 FR 21042 through 21043), we sought public comment on which social risk factors might be most appropriate for reporting stratified measure scores and/or potential risk adjustment of a particular measure. Examples of social risk factors include, but are not limited to, dual eligibility/low-income subsidy, race and ethnicity, and geographic area of residence. We also sought comments on which of these factors, including current data sources where this information would be available, could be used alone or in combination, and whether other data should be collected to better capture the effects of social risk. We will take commenters' input into consideration as we continue to assess the appropriateness and feasibility of accounting for social risk factors in the SNF QRP. We note that any such changes would be proposed through future notice and comment rulemaking.

We look forward to working with stakeholders as we consider the issue of accounting for social risk factors and reducing health disparities in CMS programs. Of note, implementing any of the above methods would be taken into consideration in the context of how this and other CMS programs operate (for example, data submission methods, availability of data, statistical considerations relating to reliability of data calculations, among others), so we sought comment on operational considerations. We are committed to ensuring that Medicare beneficiaries have access to and receive excellent care, and that the quality of care furnished by providers and suppliers is assessed fairly in CMS programs. A discussion of the comments we received on this topic, along with our responses, appears below.

Comment: Some commenters were generally supportive of accounting for social risk factors for the SNF QRP quality measures. Many commenters stated that there was evidence demonstrating that these factors can have substantial influence on patient health outcomes. Some commenters noted that social risk factors are beyond the control of the facility and were concerned that without risk adjustment, differences in quality scores may reflect differences in patient populations rather than differences in quality. Commenters also recommended incorporating the results of the NQF SES trial period into consideration of adopting risk-adjustment strategies.

A few commenters, while acknowledging the influence of social risk factors on health outcomes, cautioned against adjusting for them in quality measurement due to the potential for unintended consequences. These commenters expressed concern over the possibility that risk-adjusted measures may remove incentives for quality improvement among facilities that serve higher levels of underserved populations.

Regarding the methodology for risk adjustment, some commenters made specific recommendations regarding the type of risk adjustment that should be used. One commenter suggested that both risk stratification and statistical risk adjustment be used. Commenters stated that any risk stratification should be considered on a measure-by-measure basis, and that measures that are broadly within the control of the facility and reflective of direct care, such as pressure ulcers, should not be stratified. Multiple commenters recommended that we conduct further research and testing of risk-adjustment methods. A few commenters noted the importance of continued monitoring of the effect of social risk factors on health outcomes and on the SNF QRP over time. Other commenters recommended adjusting for social risk factors, specifically for resource use measures assessing potentially preventable readmissions, Medicare Spending Per Beneficiary, and social and environmental risk factors for functional improvement measures. Another commenter noted there are meaningful SES, clinical or other differences between traditional Medicare versus Medicare Advantage (MA) enrollees that could affect comparisons between facilities with different proportion of Medicare Advantage and Part A stays. The commenter further requested that this possibility should be investigated.

In addition to support for our suggested categories of race and ethnicity, dual eligibility status, and geographical location, specific social risk factors suggested by commenters included: Patient-level factors such as lack of personal resources, education level, healthcare literacy, employment, and limited English proficiency. Commenters also suggested community resources and other factors such as access to adequate food, medications, availability of primary care and therapy services, living conditions including living alone, lack of an adequate support system or caregiver availability. Regarding sources for data collection, a commenter suggested the use of confidential patient-reported data to determine social risk and another commenter suggested using confidential electronic health records to collect data relevant to social risk factors.

There were a few comments discussing confidential and public reporting of data adjusted for social risk factors. While a commenter recommended that risk-stratified measures should be publicly reported for purposes of transparency, another commenter noted that the public reporting of stratified rates could create a disincentive to care for disadvantaged populations.

Response: As we have previously stated, we are concerned about holding providers to different standards for the outcomes of their patients with social risk factors, because we do not want to mask potential disparities. We believe that the path forward should incentivize improvements in health outcomes for disadvantaged populations while ensuring that beneficiaries have adequate access to excellent care.

We will consider all suggestions as we continue to assess each measure and the overall program. We intend to explore options including but not limited to measure stratification by social risk factors in a consistent manner across programs, informed by considerations of stratification methods described in section IX.A.13 of the preamble of the FY 2018 IPPS/LTCH PPS final rule. We thank commenters for this important feedback and will continue to consider options to account for social risk factors that would allow us to view disparities and potentially incentivize improvement in care for patients and beneficiaries. We will also consider providing feedback to providers on outcomes for individuals with social risk factors in confidential reports.

c. Collection of Standardized Resident Assessment Data Under the SNF QRP

(1) Definition of Standardized Resident Assessment Data

Section 1888(e)(6)(B)(i)(III) of the Act requires that for fiscal year 2019 (beginning October 1, 2018) and each subsequent year, SNFs report standardized resident assessment data required under section 1899B(b)(1) of the Act. For purposes of meeting this requirement, section 1888(e)(6)(B)(ii) of the Act requires a SNF to submit the standardized resident assessment data required under section 1819(b)(3) of the Act using the standard instrument designated by the state under section 1819(e)(5) of the Act.

For purposes of the SNF QRP, we refer to beneficiaries who receive services from SNFs as “residents,” and we collect certain information about the SNF services they receive using the Resident Assessment Instrument Minimum Data Set (MDS).

Section 1899B(b)(1)(B) of the Act describes standardized resident assessment data as data required for at least the quality measures described in sections 1899B(c)(1) of the Act and that is for the following categories:

• Functional status, such as mobility and self-care at admission to a PAC provider and before discharge from a PAC provider;
• Cognitive function, such as ability to express ideas and to understand and mental status, such as depression and dementia;
• Special services, treatments and interventions such as the need for ventilator use, dialysis, chemotherapy, central line placement and total parenteral nutrition;
• Medical conditions and comorbidities such as diabetes, congestive heart failure and pressure ulcers;
• Impairments, such as incontinence and an impaired ability to hear, see or swallow; and
• Other categories deemed necessary and appropriate.

As required under section 1899B(b)(1)(A) of the Act, the standardized resident assessment data must be reported at least for SNF admissions and discharges, but the Secretary may require the data to be reported more frequently.

In the FY 2018 SNF PPS proposed rule (82 FR 21043 through 21044), we proposed to define the standardized resident assessment data that SNFs must report to comply with section 1888(e)(6) of the Act, as well as the requirements for the reporting of these data. The collection of standardized resident assessment data is critical to our efforts to drive improvement in health care quality across the four post-acute care (PAC) settings to which the IMPACT Act applies. We intend to use these data for a number of purposes, including facilitating their exchange and longitudinal use among health care providers to enable high quality care and outcomes through care coordination, as well as for quality measure calculation, and identifying comorbidities that might increase the medical complexity of a particular admission.

SNFs are currently required to report resident assessment data through the MDS by responding to an identical set of assessment questions using an identical set of response options (we refer to each solitary question/response option as a data element and we refer to a group of questions/responses as data elements), both of which incorporate an identical set of definitions and standards. The primary purpose of the identical questions and response options is to ensure that we collect a set of standardized resident assessment data elements across SNFs which we can then use for a number of purposes, including SNF payment and measure calculation for the SNF QRP.

LTCHs, IRFs, and HHAs are also required to report patient assessment data through their applicable PAC assessment instruments, and they do so by responding to identical assessment questions developed for their respective settings using an identical set of response options (which incorporate an identical set of definitions and standards). Like the MDS, the questions and response options for each of these other PAC assessment instruments are standardized across the PAC provider type to which the PAC assessment instrument applies. However, the assessment questions and response options in the four PAC assessment instruments are not currently standardized with each other. As a result, questions and response options that appear on the MDS cannot be readily compared with questions and response options that appear, for example, on the Inpatient Rehabilitation Facility-Patient Assessment Instrument (IRF-PAI) the PAC assessment instrument used by IRFs. This is true even when the questions and response options are similar. This lack of standardization across the four PAC provider types has limited our ability to compare one PAC provider type with another for purposes such as care coordination and quality improvement.

To achieve a level of standardization across SNFs, LTCHs, IRFs, and HHAs that enables us to make comparisons between them, we proposed to define “standardized resident assessment data” as patient or resident assessment questions and response options that are identical in all four PAC assessment instruments, and to which identical standards and definitions apply. Standardizing the questions and response options across the four PAC assessment instruments will also enable the data to be interoperable allowing it to be shared electronically, or otherwise, between PAC provider types. It will enable the data to be comparable for various purposes, including the development of cross-setting quality measures, which may enhance provider and resident choice when selecting a post-acute care setting that will deliver the best outcome possible, and to inform payment models that take into account patient characteristics rather than setting, as described in the IMPACT Act.

We sought comment on this definition. A discussion of these comments, along with our responses, appears below.

Comment: Most commenters expressed general support for the definition of standardized patient/resident assessment data. One commenter further expressed support for CMS efforts to standardize assessment data to promote care coordination and quality improvements as required under the IMPACT Act.

Response: We thank the commenters for their support.

Final Decision: We are finalizing our definition of standardized resident assessment data as proposed.

(2) General Considerations Used for the Selection of Standardized Resident Assessment Data

As part of our effort to identify appropriate standardized resident assessment data for purposes of collecting under the SNF QRP, we sought input from the general public, stakeholder community, and subject matter experts on items that would enable person-centered, high quality health care, as well as access to longitudinal information to facilitate coordinated care and improved beneficiary outcomes.

To identify optimal data elements for standardization, our data element contractor organized teams of researchers for each category, and each team worked with a group of advisors made up of clinicians and academic researchers with expertise in PAC. Information-gathering activities were used to identify data elements, as well as key themes related to the categories described in section 1899B(b)(1)(B) of the Act. In January and February 2016, our data element contractor also conducted provider focus groups for each of the four PAC provider types, and a focus group for consumers that included current or former PAC patients and residents, caregivers, ombudsmen, and patient advocacy group representatives. The Development and Maintenance of Post-Acute Care Cross-Setting Standardized Patient Assessment Data Focus Group Summary Report is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

We also assembled a 16-member TEP that met on April 7 and 8, 2016, and January 5 and 6, 2017, in Baltimore, Maryland, to provide expert input on data elements that are currently in each PAC assessment instrument, as well as data elements that could be standardized. The Development and Maintenance of Post-Acute Care Cross-Setting Standardized Patient Assessment Data TEP Summary Reports are available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

As part of the environmental scan, data elements currently in the four existing PAC assessment instruments were examined to see if any could be considered for proposal as standardized resident assessment data. Specifically, this evaluation included consideration of data elements in OASIS-C2 (effective January 2017); IRF-PAI, v1.4 (effective October 2016); LCDS, v3.00 (effective April 2016); and MDS 3.0, v1.14 (effective October 2016). Data elements in the standardized assessment instrument that we tested in the Post-Acute Care Payment Reform Demonstration (PAC PRD)—the Continuity Assessment Record and Evaluation (CARE) were also considered. A literature search was also conducted to determine whether additional data elements to propose as standardized resident assessment data could be identified.

We additionally held four Special Open Door Forums (SODFs) on October 27, 2015; May 12, 2016; September 15, 2016; and December 8, 2016, to present data elements we were considering and to solicit input. At each SODF, some stakeholders provided immediate input, and all were invited to submit additional comments via the CMS IMPACT Mailbox at PACQualityInitiative@cms.hhs.gov.

We also convened a meeting with federal agency subject matter experts (SMEs) on May 13, 2016. In addition, a public comment period was open from August 12, to September 12, 2016, to solicit comments on detailed candidate data element descriptions, data collection methods, and coding methods. The IMPACT Act Public Comment Summary Report containing the public comments (summarized and verbatim) and our responses, is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

We specifically sought to identify standardized resident assessment data that we could feasibly incorporate into the LTCH, IRF, SNF, and HHA assessment instruments and that have the following attributes: (1) Being supported by current science; (2) testing well in terms of their reliability and validity, consistent with findings from the Post-Acute Care Payment Reform Demonstration (PAC PRD); (3) the potential to be shared (for example, through interoperable means) among PAC and other provider types to facilitate efficient care coordination and improved beneficiary outcomes; (4) the potential to inform the development of quality, resource use and other measures, as well as future payment methodologies that could more directly take into account individual beneficiary health characteristics; and (5) the ability to be used by practitioners to inform their clinical decision and care planning activities. We also applied the same considerations that we apply with quality measures, including the CMS Quality Strategy which is framed using the three broad aims of the National Quality Strategy.

d. Policy for Retaining SNF QRP Measures and Application of That Policy to Standardized Resident Assessment Data

In the FY 2016 SNF PPS final rule (80 FR 46431 through 46432), we adopted our policy for measure removal and also finalized that when we initially adopt a measure for the SNF QRP, this measure will be automatically retained in the SNF QRP for all subsequent payment determinations unless we propose to remove, suspend, or replace the measure. In the FY 2018 SNF PPS proposed rule (82 FR 21044) we proposed to apply this policy to the standardized resident assessment data that we adopt for the SNF QRP.

We sought public comment on our proposal. A discussion of these comments, along with our responses, appears below.

Comment: Several commenters supported applying the existing policy for retaining SNF QRP measures to standardized resident assessment data.

Response: We thank the commenters for their support.

Final Decision: After consideration of the public comments we received, we are finalizing our proposal to apply the policy for retaining SNF QRP measures to the standardized resident assessment data as proposed.

e. Policy for Adopting Changes to SNF QRP Measures and Application of That Policy to Standardized Resident Assessment Data

In the FY 2016 SNF PPS final rule (80 FR 46432), we finalized our policy pertaining to the process for adoption of non-substantive and substantive changes to SNF QRP measures. We did not propose to make any changes to this policy in the FY 2018 SNF PPS proposed rule (82 FR 21044 through 20145). We did propose to apply this policy to the standardized resident assessment data that we adopt for the SNF QRP.

We sought public comment on our proposal. A discussion of these comments, along with our responses, appears below.

Comment: All commenters who commented on this topic expressed support for our subregulatory process for adopting non-substantive changes to SNF QRP measures, recognizing that the measures will require adjustments over time to reflect changes in practice or populations. All of these commenters also specifically expressed support for our proposal to apply this approach to the standardized resident assessment data proposed for the SNF QRP. Many of these commenters further supported our policy to make substantive changes to quality measures using the rulemaking process. The commenters also recognized that corrections and adjustments to measures may become necessary over time and that we will provide a clear rationale for such changes, as well as a mechanism for public comment on these changes.

Response: We appreciate the commenters' support.

Final Decision: After consideration of the public comments we received, we are finalizing our proposal to apply our policy for adopting changes to the SNF QRP measures to the standardized resident assessment data as proposed.

f. Quality Measures Currently Adopted for the SNF QRP

The SNF QRP currently has seven adopted measures as outlined in Table 18.

We received several comments about quality measures currently adopted for the SNF QRP which are summarized and discussed below.

Comment: A few commenters expressed views regarding the Potentially Preventable 30-Day Post-Discharge Readmission Measure for SNF QRP, a measure previously finalized in the FY 2017 SNF PPS final rule (81 FR 52030 through 52034). Comments included recommendations for additional testing and evaluation of the PPR definition and measure exclusions. One commenter supported the public reporting thresholds. Another commenter requested that patient-level data be made available to SNFs to facilitate quality improvement and review and corrections. We also received some comments related to accounting for social risk factors.

Response: While we received comments regarding this previously finalized measure, the changes we proposed pertain only to the years of data used to calculate this measure and therefore we consider these comments to be out of scope of this current rule. We did address these issues in the FY 2017 SNF PPS final rule (81 FR 52030 through 52034), and we refer the reader to that detailed discussion. We continue to believe that the measure specifications are appropriate for this measure. We also refer readers to section III.D.2.b.1 of this rule for responses to comments received related to social risk factors for this measure.

Comment: We received a comment regarding the Drug Regimen Review Conducted with Follow-Up for Identified Issues-PAC SNF QRP measure, a measure previously finalized in the FY 2017 SNF PPS final rule. The commenter expressed support for MedPAC comments regarding the measure, including the MedPAC recommendation that we develop a measure to evaluate PAC provider support for medication reconciliation throughout the care continuum, including provider transfer of the patient medication list to the follow-up provider at patient discharge. The commenter stated the importance of provider access to patient medication lists and suggested that requiring providers to transmit the patient medication list to the follow-up provider at discharge may improve patient safety and prevent avoidable readmissions.

Response: We appreciate the comments received for this finalized measure. We refer readers to the FY 2017 SNF PPS final rule (81 FR 52034 through 52039) for detailed responses related to the previously finalized Drug Regimen Review Conducted with Follow-Up for Identified Issues-PAC SNF QRP measure.

Comment: A few commenters expressed views regarding the Medicare Spending per Beneficiary-PAC SNF QRP, a measure finalized in the FY 2017 SNF PPS final rule (81 FR 52014 through 52021). Commenters addressed the risk-adjustment approach, clinically unrelated services, confidential feedback reporting, accounting for social risk factors, MSPB-PAC measure alignment, and unintended consequences related to implementation of the measure. One commenter felt that the measure was confusing, and that patients and providers might incorrectly interpret it as a measure of quality rather than efficiency. Another commenter encouraged CMS to utilize claims and patient assessment data to incorporate functional status into the risk-adjustment. Another commenter expressed concern that PAC providers' performance on this measure would focus on costs per patient, without fully accounting for patient outcomes, and that efficiency should not be based solely on the MSPB-PAC measures. This commenter also noted that this measure may result in limiting access to certain patients. One commenter stated that the MSPB-PAC measures should be more uniformly defined so as to facilitate a meaningful comparison of spending for beneficiaries across PAC settings. Another commenter felt that the measure was flawed with regard to putting SNFs at risk for post-discharge services beyond their control. The commenter encouraged CMS to provide additional details regarding the types of services that would be considered “included and associated services.” Another commenter urged CMS to provide the opportunity for confidential feedback between CMS and providers before publicly displaying the MSPB-PAC measures.

Response: While we received comments regarding the previously finalized measure, Medicare Spending per Beneficiary-PAC SNF QRP, since no changes were proposed to this measure, we consider comments received to be outside the scope of the current rule. We addressed these issues in the FY 2017 SNF PPS final rule (81 FR 52014 through 52021), and we refer readers to that detailed discussion. We continue to believe that the measure specifications, including the risk-adjustment, are appropriate for this measure. With regard to comments related to accounting for social risk factors, we refer readers to section III.D.2.b.1. of this rule.

Comment: We received comments related to the Discharge to Community-PAC SNF QRP measure, a measure previously finalized in the FY 2017 SNF PPS final rule. Comments included suggestions to adjust for sociodemographic and socioeconomic risk factors and caregiver support, to adjust for factors unique to providers offering dedicated services to specialty residents (for example, those with HIV/AIDS) who may encounter greater challenges with community transitions, to exclude patients who died in the observation window following return to a community setting, to distinguish between a patient's return to home in the community versus home in a custodial nursing facility, to assess reliability and validity of the claims discharge status code used to calculate the measure, and to submit the measure for NQF endorsement. Commenters also shared concerns about risk adjustment for social factors as this could mask disparities in care, potential unintended consequences for patients expected to have difficult transitions to the community such as decreased PAC access and increased healthcare costs due to more costly acute care stays, lack of adjustment for regional differences in community-based needs and supports, and lack of adjustment for patients' goals in the community, such as those seeking end-of-life care outside of formal hospice services.

Response: While we received comments regarding the previously finalized Discharge to Community-PAC SNF QRP measure, since no changes were proposed to this measure, we consider comments received to be outside the scope of the current rule. We previously responded to comments on these topics in the FY 2017 SNF PPS final rule (81 FR 52021 through 52029); we refer the commenters to the FY 2017 SNF PPS final rule for a detailed response on these issues. We also note that in the FY 2018 SNF PPS proposed rule (81 FR 21058), we sought comment on the exclusion of baseline nursing facility residents as a potential future modification of the Discharge to Community-PAC SNF QRP measure. We refer readers to section III.D.2.i.1 of this final rule for a discussion of this issue. We also refer readers to section III.D.2.b.1. of this final rule for responses to comments received related to accounting for social risk factors for the Discharge to Community-PAC SNF QRP measure.

g. SNF QRP Quality Measures Beginning With the FY 2020 SNF QRP

In the FY 2018 SNF PPS proposed rule (82 FR 21045 through 21057), beginning with the FY 2020 SNF QRP, in addition to the quality measures we are retaining under our policy described in section III.D.2.f. of this final rule, we proposed to remove the current pressure ulcer measure entitled Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (Short Stay) (NQF #0678) and to replace it with a modified version of the measure entitled Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury and to adopt four function outcome measures on resident functional status. We also proposed to characterize the data elements described below as standardized resident assessment data under section 1899B(b)(1)(B) of the Act that must be reported by SNFs under the SNF QRP through the MDS.

The measures are as follows:

• Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury.
• Application of IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633).
• Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634).
• Application of IRF Functional Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (NQF #2635).
• Application of IRF Functional Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (NQF #2636).

The measures are described in more detail below.

(1) Replacing the Current Pressure Ulcer Quality Measure, Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (Short Stay) (NQF #0678), with a Modified Pressure Ulcer Measure, Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury

(a) Measure Background

In the FY 2018 SNF PPS proposed rule (82 FR 21045 through 21049), we proposed to remove the current pressure ulcer measure, Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (Short Stay) (NQF #0678) from the SNF QRP measure set and replace it with a modified version of that measure, Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury, beginning with the FY 2020 SNF QRP. The change in the measure name is to reduce confusion about the new modified measure. The modified version differs from the current version of the measure because it includes new or worsened unstageable pressure ulcers, including deep tissue injuries (DTIs), in the measure numerator. The modified version of the measure would satisfy the IMPACT Act domain of skin integrity and changes in skin integrity.

We note that the technical specifications for the pressure ulcer measure were updated in August 2016 through a subregulatory process to ensure technical alignment of the SNF measure specifications with the LTCH, IRF, and HH specifications. The technical updates were added to ensure clarity in how the measure is calculated, and to avoid possible over counting of pressure ulcers in the numerator. We corrected the technical specifications to mitigate the risk of over counting new or worsened pressure ulcers and to reflect the actual unit of analysis as finalized in the rule, which is a stay (Medicare Part A stay) for SNF QRP, consistent with the IRF, and LTCH QRPs, rather than an episode (which could include multiple stays) as is used in the case of Nursing Home Compare. Thus, we updated the SNF measure specifications to reflect all resident stays, rather than the most-recent episode in a quarter, which is comprised of one or more stays in that measure calculation. Also, to ensure alignment, we corrected our specifications to ensure that healed wounds are not incorrectly captured in the measure. Further, we corrected the specifications to ensure the exclusion of residents who expire during their SNF stay. The SNF specifications can be reviewed on our Web site at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html.

(b) Measure Importance

As described in the FY 2016 SNF PPS final rule (80 FR 46433), pressure ulcers are high-cost adverse events and an important measure of quality. For information on the history and rationale for the relevance, importance, and applicability of having a pressure ulcer measure in the SNF QRP, we refer readers to the FY 2016 SNF PPS final rule (80 FR 46433 through 46434).

We proposed to adopt a modified version of the current pressure ulcer measure because unstageable pressure ulcers, including DTIs, are similar to Stage 2, Stage 3, and Stage 4 pressure ulcers in that they represent poor outcomes, are a serious medical condition that can result in death and disability, are debilitating and painful, and are often an avoidable outcome of medical care. Studies show that most pressure ulcers can be avoided and can also be healed in acute, post-acute, and long-term care settings with appropriate medical care.

Furthermore, some studies indicate that DTIs, if managed using appropriate care, can be resolved without deteriorating into a worsened pressure ulcer. While DTIs are a subset of unstageable pressure ulcers, we collect DTI data elements separately and analyze them both separately and with other unstageable pressure ulcer item categories in our analysis below. We note that DTIs are categorized as a type of unstageable pressure ulcer on the MDS and other post-acute care item sets.

While there are few studies that provide information regarding the incidence of unstageable pressure ulcers in PAC settings, an analysis conducted by a contractor suggests the incidence of unstageable pressure ulcers varies according to the type of unstageable pressure ulcer and setting. This analysis examined the national incidence of new unstageable pressure ulcers in SNFs at discharge compared with admission using SNF discharges from January through December 2015. The contractor found a national incidence of 0.40 percent of new unstageable pressure ulcers due to slough and/or eschar, 0.02 percent of new unstageable pressure ulcers due to non-removable dressing/device, and 0.57 percent of new DTIs. In addition, an international study spanning the time period 2006 to 2009, provides some evidence to suggest that the proportion of pressure ulcers identified as DTI has increased over time.

The inclusion of unstageable pressure ulcers, including DTIs, in the numerator of this measure is expected to increase measure scores and variability in measure scores, thereby improving the ability to discriminate among poor- and high-performing SNFs. In the currently implemented pressure ulcer measure, Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (Short Stay) (NQF #0678), analysis using data from Quarter 4 2015 through Quarter 3 2016 reveals that the SNF mean score is 1.75 percent; the 25th and 75th percentiles are 0.0 percent and 2.53 percent, respectively; and 29.11 percent of facilities have perfect scores. In the measure, Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury, during the same timeframe, the SNF mean score is 2.58 percent; the 25th and 75th percentiles are 0.65 percent and 3.70 percent, respectively; and 20.32 percent of facilities have perfect scores.

(c) Stakeholder Feedback

Our measure development contractor sought input from subject matter experts, including Technical Expert Panels (TEPs), over the course of several years on various skin integrity topics and specifically those associated with the inclusion of unstageable pressure ulcers, including DTIs. Most recently, on July 18, 2016, a TEP convened by our measure development contractor provided input on the technical specifications of this quality measure, including the feasibility of implementing the proposed measure's updates related to the inclusion of unstageable ulcers, including DTIs, across PAC settings. The TEP supported the updates to the measure across PAC settings, including the inclusion in the numerator of unstageable pressure ulcers due to slough and/or eschar that are new or worsened, new unstageable pressure ulcers due to a non-removable dressing or device, and new DTIs. The TEP recommended supplying additional guidance to providers regarding each type of unstageable pressure ulcer. This support was in agreement with earlier TEP meetings, held on June 13, and November 15, 2013, which had recommended that CMS update the specifications for the pressure ulcer measure to include unstageable pressure ulcers in the numerator. Exploratory data analysis conducted by our measure development contractor suggests that the addition of unstageable pressure ulcers, including DTIs, will increase the observed incidence and variation in the rate of new or worsened pressure ulcers at the facility level, which may improve the ability of the proposed quality measure to discriminate between poor- and high-performing facilities.

We solicited stakeholder feedback on this proposed measure by means of a public comment period held from October 17 through November 17, 2016. In general, we received considerable support for the proposed measure. A few commenters supported all of the changes to the current pressure ulcer measure that resulted in the measure, with one commenter noting the significance of the work to align the pressure ulcer quality measure specifications across the PAC settings. Many commenters supported the inclusion of unstageable pressure ulcers due to slough/eschar, due to non-removable dressing/device, and DTIs in the quality measure. Other commenters did not support the inclusion of DTIs in the quality measure because they stated that there is no universally accepted definition for this type of skin injury.

The public comment summary report for the proposed measure is available on the CMS Web site at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html. This summary includes further detail about our responses to various concerns and ideas stakeholders raised at that time.

The NQF-convened Measures Application Partnership (MAP) Post-Acute Care/Long-Term Care (PAC/LTC) Workgroup met on December 14 and 15, 2016, and provided input to us about this measure. The workgroup provided a recommendation of “support for rulemaking” for use of the measure in the SNF QRP. The MAP Coordinating Committee met on January 24 and 25, 2017, and provided a recommendation of “conditional support for rulemaking” for use of the proposed measure in the SNF QRP. The MAP's conditions of support include that, as a part of measure implementation, CMS provide guidance on the correct collection and calculation of the measure result, as well as guidance on public reporting Web sites explaining the impact of the specification changes on the measure result. The MAP's conditions also specify that CMS continue analyzing the proposed measure to investigate unexpected results reported in public comment. We intend to fulfill these conditions by offering additional training opportunities and educational materials in advance of public reporting, and by continuing to monitor and analyze the proposed measure. More information about the MAP's recommendations for this measure is available at http://www.qualityforum.org/WorkArea/linkit.aspx?LinkIdentifier=id&ItemID=84452.

We reviewed the NQF's consensus endorsed measures and were unable to identify any NQF-endorsed pressure ulcer quality measures for PAC settings that are inclusive of unstageable pressure ulcers. There are related measures, but after careful review, we determined these measures are not applicable for use in SNFs based on the populations addressed or other aspects of the specifications. We are unaware of any other such quality measures that have been endorsed or adopted by another consensus organization for the SNF setting. Therefore, based on the evidence discussed above, we proposed to adopt the quality measure entitled, Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury, for the SNF QRP beginning with the FY 2020 SNF QRP. We plan to submit the proposed measure to the NQF for endorsement consideration as soon as feasible.

(d) Data Collection

The data for this quality measure would be collected using the MDS, which is currently submitted by SNFs through the Quality Improvement and Evaluation System (QIES) Assessment Submission and Processing (ASAP) System. The proposed standardized resident assessment data applicable to this measure that must be reported by SNFs for admissions as well as discharges occurring on or after October 1, 2018 is described in section III.D.2. of this final rule. SNFs are already required to complete unstageable pressure ulcer data elements on the MDS. While the inclusion of unstageable wounds in the proposed measure results in a measure calculation methodology that is different from the methodology used to calculate the current pressure ulcer measure, the data elements needed to calculate the proposed measure are already included in the MDS. In addition, this proposed measure will further standardize the data elements used in risk adjustment of this measure. Our proposal to eliminate duplicative data elements will result in an overall reduced reporting burden for SNFs for the proposed measure.

To view the updated MDS, with the proposed changes, we refer to the reader to https://www.cms.gov/medicare/quality-initiatives-patient-assessment-instruments/nursinghomequalityinits/mds30raimanual.html . For more information on MDS submission using the QIES ASAP System, we refer readers to https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/NHQIMDS30TechnicalInformation.html .

For technical information about this proposed measure, including information about the measure calculation and the standardized resident assessment data elements used to calculate this measure, we refer readers to the document titled, Final Measure Specifications for SNF QRP Quality Measures and Standardized Resident Assessment Data Elements, available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html .

We proposed that SNFs begin reporting the proposed pressure ulcer measure, Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury, which will replace the current pressure ulcer measure, with data collection beginning October 1, 2018 for admissions as well as discharges.

We sought public comment on our proposal to replace the current pressure ulcer measure, Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (Short Stay) (NQF #0678), with a modified version of that measure, entitled Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury, beginning with the FY 2020 SNF QRP. A discussion of these comments, along with our responses, appears below.

Comment: Many commenters supported the proposed replacement of the current pressure ulcer measure, the Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (Short Stay) (NQF #0678), with a modified version of that measure, entitled Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury. Commenters recognized that the proposed measure will meet the requirements of the IMPACT Act for the Skin Integrity and Changes in Skin Integrity domain. Commenters believed that the revisions identified in the proposed rule will improve on the existing pressure ulcer measure and ensure that the data collected accurately reflects the care and conditions of the SNF patient population. One commenter supported the use of data elements that are already in use in the MDS to reduce reporting burden for providers. Another commenter noted that revisions to quality measures are an important part of ensuring accurate information that is reflective of advances in knowledge and technology, and ensuring that the data reflect the patient population.

Response: We appreciate the commenters' support to replace the current pressure ulcer measure, Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (Short Stay) (NQF #0678), with a modified version of the measure, Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury to fulfill the requirements of the IMPACT Act. We agree that this proposal will limit regulatory burden and promote high quality care, as the commenters describe.

Comment: A few commenters expressed concerns that the variation in measure scores between facilities could reflect differences in the interpretation of definitions for unstageable pressure ulcers or DTIs, rather than actual differences in quality or care practices. One commenter cautioned that a measure should not be changed to create performance variation, but rather to be consistent with current science or to provide clarity and consistent data collection. The commenters encouraged additional testing of the measure to ensure that it collects accurate data.

Response: We have performed testing to compare the performance of the proposed measure with the existing pressure ulcer/injury measure. Current findings indicate that the measure is both valid and reliable in the SNF, LTCH, and IRF settings.

The reliability and validity of the data elements used to calculate this quality measure have been tested in several ways. Rigorous testing on both reliability and validity of the data elements in the MDS 3.0 provides evidence for the data elements used in the SNF, LTCH, and IRF settings. The MDS 3.0 pilot test showed good reliability, and the results are applicable to the IRF-PAI as well as the LTCH CARE Data Set because the data elements tested are the same as those used in the IRF-PAI and LTCH CARE Data Set. Across pressure ulcer data elements, average gold-standard to gold-standard kappa statistic was 0.905. The average gold-standard to facility-nurse kappa statistic was 0.937. These kappa scores indicate “almost perfect” agreement using the Landis and Koch standard for strength of agreement.

To assess the construct validity of this measure, or the degree to which the measure construct measures what it claims or purports to be measuring, our measure contractor sought input from TEPs over the course of several years. Most recently, on July 18, 2016, a TEP supported the inclusion in the numerator of unstageable pressure ulcers due to slough and/or eschar that are new or worsened, new unstageable pressure ulcers/injuries due to a non-removable dressing or device, and new DTIs. The measure testing activities were presented to TEP members for their input on the reliability, validity, and feasibility of this measure change. The TEP members supported the measure construct.

The proposed measure also increased the variability of measures scores between providers, as noted by some commenters. We would like to clarify that the goal of the proposed measure is not to create performance variation where none exists, but rather to better measure existing performance variation. This increased variability of scores between facilities will improve the ability of the measure to distinguish between high- and low-performing facilities.

We will continue to perform reliability and validity testing in compliance with NQF guidelines and the Blueprint for the CMS Measures Management System to ensure that that the measure demonstrates scientific acceptability (including reliability and validity) and meets the goals of the QRP. Finally, as with all measure development and implementation, we will provide training and guidance prior to implementation of the measure to promote consistency in the interpretation of the measure.

Comment: Commenters requested further training and guidance in completing the M0300 data element that will be used to calculate the proposed quality measure. One commenter stated that confusion exists related to worsening of pressure ulcers, unstageable pressure ulcers due to slough or eschar, and the concept of “present on admission”. One commenter stated that the use of these data elements would require SNFs to calculate the number of new or worsened pressure ulcers by subtracting those present on admission. Some commenters stated that the modified measure may be difficult for providers to capture because they are being asked to report on a different data element.

Response: The measure will be calculated using data reported on the M0300 data element collected at discharge, which only requires SNFs to report the number of pressure ulcers for each stage (including stages 2, 3, and 4, unstageable due to slough and/or eschar, unstageable due to non-removable dressing/device, and DTIs), and of those, the number that were present on admission. The M0300 data element currently exists on the MDS, and the current MDS RAI Manual, as well as prior versions of the Manual, include guidance about how to complete the data element, including unstageable pressure ulcers and pressure ulcers that are present on admission. The MDS RAI Manual can be found at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursinghomeQualityInits/MDS30RAIManual.html.

Comment: We received several comments regarding the inclusion of unstageable pressure ulcers in the proposed measure. One commenter specifically supported the inclusion of these types of pressure ulcers. Other commenters did not support the inclusion of unstageable pressure ulcers, in the quality measure as proposed, and encouraged further testing. Some commenters stated that there is a lack of clear definition of pressure ulcers included in this measure, and that those definitions may be too subjective to get reliable data. Commenters also requested that we provide training opportunities and educational materials prior to the implementation of this measure.

Response: We appreciate the support we have received regarding the inclusion of unstageable pressure ulcers, including DTIs, in the proposed quality measure. We believe that the inclusion of unstageable pressure ulcers in the measure will result in a fuller picture of quality to residents and families, and lead to further quality improvement efforts that will advance patient safety by reducing the rate of facility acquired pressure ulcers at any stage. We would like to clarify that the definitions of pressure ulcers are adapted from the National Pressure Ulcer Advisory Panel (NPUAP), and are standardized across all PAC settings. These definitions are universally accepted, objective, and considered to be the gold-standard definition by national and international stakeholders such as the NPUAP, European Pressure Ulcer Advisory Panel (EPUAP), Wound, Ostomy and Continence Nurses Society (WOCN), amongst others. As a result, the use of these universally accepted definitions of pressure ulcers furthers our commitment to ensuring that all quality measures implemented in the QRP meet the testing goals of the QRP.

To provide greater clarity about the definitions of different types of unstageable pressure ulcers and how to code them on the MDS, we are currently engaged in multiple educational efforts. These include training events, updates to the manuals and training materials, and responses to Help Desk questions to promote understanding and proper coding of these data elements. We will continue to engage in these training activities prior to implementation of the proposed measure.

Comment: One commenter specifically supported the new measure and the specific inclusion of DTIs, and stressed the importance and impact of such change in increasing the number of pressure ulcers captured. The commenter stated that it would be important to note the impact on the Five Star Quality Rating System. This commenter also noted that some DTIs can also evolve or worsen, despite being managed with appropriate care. Other commenters did not support the inclusion of DTIs in the measure. These commenters stated that there is not a universally accepted definition of DTIs, and that DTIs are commonly misdiagnosed, which could lead to surveillance bias.

Response: We appreciate the comments regarding the inclusion of DTIs in the proposed quality measure. DTIs are often an avoidable outcome of medical care, are debilitating and painful, and can result in death and/or disability, similar to Stage 2, Stage 3 and Stage 4 pressure ulcers. While some DTIs may worsen, studies indicate that many DTIs, if managed using appropriate care, can be resolved without deteriorating into a worsened pressure ulcer. Therefore, we believe that the inclusion of DTIs in the proposed quality measure is essential to be able to accurately reflect the number of these types of pressure injuries and to provide the appropriate patient care. Further, we believe that it is important to do a thorough assessment on every patient in each PAC setting, including a thorough skin assessment documenting the presence of any pressure ulcers or injuries of any kind, including DTIs. We agree that it is important to conduct thorough and consistent assessments to avoid the possibility of surveillance bias.

When considering the addition of DTIs to the measure numerator, we convened cross-setting TEPs in June and November 2013, and obtained input from clinicians, experts, and other stakeholders. An additional cross-setting TEP convened by our measure development contractor in July 2016 also supported the recommendation to include unstageable pressure ulcers, including DTIs, in the numerator of the quality measure. Given DTIs' potential impact on mortality, morbidity, and quality of life, it may be detrimental to the quality of care to exclude DTIs from a pressure ulcer quality measure.

We do not intend to include the proposed measure in the Five Star Quality Rating System calculations.

Comment: Several commenters recommended that we attain NQF endorsement of the Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury measure prior to implementation.

Response: While this measure is not currently NQF-endorsed, we recognize that the NQF endorsement process is an important part of measure development and plan to submit this measure for NQF endorsement consideration as soon as feasible.

Comment: Several commenters noted that there is a difference in the denominator across settings in terms of which payer sources (Medicare Part A or Medicare Advantage) are included in the measure. Commenters recommended that we ensure that common denominators are used when displaying this measure for quality comparison purposes. One commenter stated that there is an IMPACT Act mandate to implement “interoperable measures” across PAC settings.

Response: We recognize that data is currently collected from different payer sources for each PAC setting. We believe that quality care is best assessed through the collection of data from all patients, and strive to include the largest possible patient population in the measure denominator. For this reason, we do not seek to limit the denominator in each setting based on the data currently available in other settings (that is, limiting every setting denominator to Medicare Part A patients). Regarding the concern that different patient population denominators are misleading to consumers and providers, we seek to clarify the intent and use of this quality measure through rulemaking, provider training, and ongoing communication with stakeholders. Ongoing communication includes the posting of measure specifications and communication accompanying public reporting. Further, we will take into consideration the expansion of the SNF QRP to include all payer sources through future rulemaking.

The Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury measure is harmonized across all PAC settings and uses standardized resident assessment data as required by the IMPACT Act. Further, we would like to clarify that the M0300 data element used to calculate this measure is standardized across all PAC settings, enabling interoperability. This standardization and interoperability of data elements allows for the exchange of information among PAC providers and other providers to whom this data is applicable. We refer readers to the measure specifications, which describe the specifications for the measure in PAC settings, Final Specifications for SNF QRP Quality Measures and Standardized Resident Assessment Data Elements, available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html .

Comment: One commenter indicated support for our efforts to standardize data elements across PAC settings and encouraged further standardization of coding instructions across settings. The commenter specifically noted that coding guidance surrounding Kennedy Ulcers seems to differ between the LTCH and SNF manuals. The commenter urged us to thoroughly review all manuals to ensure standardization of coding guidance and instructions.

Response: The LTCH QRP Manual Version 3.0 instructs LTCHS to not count Kennedy ulcers in the pressure ulcer data elements. The MDS RAI Manual Version 1.14 provides guidance regarding the etiology of ulcers that should be reported in the data elements, but does not provide specific guidance on Kennedy ulcers. The guidance in the two manuals differs in order to be specific to each setting. Although the guidance is tailored to be most applicable to each setting, the data elements are standardized. Therefore, we do not expect this tailored guidance to add variation to the measure outcome or to the standardized resident assessment data.

Comment: A few commenters noted that SNF performance scores on the proposed measure are likely to differ from performance scores on the currently implemented pressure ulcer measure, Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (Short Stay) (NQF #0678). They recommended development of educational materials for the public to explain the perceived shifts in performance.

Response: We appreciate commenters' concerns about differences in performance scores between the two measures and the possibility of misinterpretation. While the proposed measure will not be directly comparable to the existing measure, it is expected to provide an improved measure of quality moving forward since it will more accurately capture the number of new and worsened pressure ulcers and include unstageable pressure ulcers. Further information and training will be provided to providers as well as consumers regarding how to interpret scores on the proposed measure, to avoid any possible confusion between the proposed measure and the existing measure.

Comment: One commenter suggested that we include additional risk factors in the proposed measure for populations that may be compromised physically, such as the ventilator-dependent population, and to include factors such as whether the resident experienced a hospital stay, was in the emergency department for an extended period of time, was on a stretcher for an extended period of time, was receiving palliative care, and other hospital factors that may lead to the development of pressure ulcers. The commenter also recommended that social risk factors be accounted for in the quality measure. One commenter stated that the proposed measure should be properly risk adjusted.

Response: The proposed quality measure would be risk adjusted for functional mobility admission performance, bowel continence, diabetes mellitus or peripheral vascular disease/peripheral arterial disease, and low body mass index in each of the four settings. This risk adjustment methodology is described further in the Final Specifications for SNF QRP Quality Measures and Standardized Resident Assessment Data Elements, available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html . As with our measure modification and evaluation processes, we will continue to analyze this measure, specifically assessing the addition of variables to the risk adjustment model, and testing the inclusion of other risk factors as additional risk adjustors. This continued refinement of the risk adjustment models will ensure that the measure remains valid and reliable to inform quality improvement within and across each PAC setting, and to fulfill the public reporting goals of quality reporting programs. Our approach to using social risk factors for risk adjustment is further described in section III.D.2.B.1 of this final rule.

Comment: One commenter requested clarification regarding the proposed measure and the population it is applied to, stating that the long stay pressure ulcer quality measure and short stay pressure ulcer quality measure appear to be combined into a single measure.

Response: The proposed measure, Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury, is distinct from both the Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (Short-Stay) Measure (NQF #0678) and the Percent of High Risk Residents with Pressure Ulcers (Long Stay) Measure (NQF #0679). There are several key differences between these measures and the programs they are used in. The long-stay measure, Percent of High-Risk Residents with Pressure Ulcers (NQF #0679), measures the percent of residents with one or more conditions indicating high risk to develop pressure ulcers (impaired bed mobility or transfer, comatose, or malnutrition/risk of malnutrition) with any pressure ulcers. This measure is used in the Nursing Home Quality Initiative (NHQI) and reported on Nursing Home Compare. Conversely, the short-stay measure, Percent of Residents with Pressure Ulcers that are New or Worsened (short-stay) (NQF #0678), currently used in used in the SNF QRP, assesses the percentage of residents who develop new pressure ulcers or have existing pressure ulcers worsen over their course of stay in a PAC facility.

The short stay measure does not include unstageable pressure ulcers in the numerator. The measure is used in the NHQI and reported on Nursing Home Compare, and is also currently applied to SNF residents for the SNF QRP.

We reviewed both the short stay and long stay measures for suitability, but the short stay measure does not include unstageable pressure ulcers in the numerator, as described above, and the long stay measure was determined to not be applicable for use in SNFs due to the populations addressed. The proposed measure is to be applied to the SNF population, which comprises residents who are receiving skilled nursing services. This measure includes new or worsened pressure ulcers that are numerically staged or unstageable, and is standardized across the PAC settings. Further information about the specifications of this measure can be found at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html .

Final Decision: After consideration of the public comments we received, we are finalizing our proposal to remove the current pressure ulcer measure, Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (Short Stay) (NQF #0678), from the SNF QRP measure set and to replace it with a modified version of that measure, entitled Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury, for the SNF QRP with an implementation date of October 1, 2018.

(2) Functional Outcome Measures

In the FY 2018 SNF PPS proposed rule (82 FR 21047 through 21057) we proposed for the SNF QRP four measures that we are specifying under section 1899B(c)(1) of the Act for the purposes of meeting the functional status, cognitive function, and changes in function and cognitive function domain: (1) Application of the IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633); (2) Application of the IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634); (3) Application of the IRF Functional Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (NQF #2635); and (4) Application of the IRF Functional Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (NQF #2636). We finalized the same functional outcome measures for the IRF QRP in the FY 2016 IRF PPS final rule (80 FR 47111 through 47117). These measures are: (1) IRF Functional Outcome Measure: Change in Self-Care for Medical Rehabilitation Patients (NQF #2633); (2) IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation (NQF #2634); (3) IRF Functional Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (NQF #2635); and (4) IRF Functional Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (NQF #2636). We believe these measures satisfy section 1899B(c)(1)(A) of the Act because they address functional status, cognitive function, and changes in function and cognitive function domain. We intend to propose functional outcome measures for the home health and long-term care hospital settings in the future.

In developing these SNF functional outcome quality measures, we sought to build on our cross-setting function work by leveraging data elements currently collected in the MDS section GG, which would minimize additional data collection burden while increasing the feasibility of cross-setting item comparisons.

SNFs provide skilled services, such as skilled nursing or therapy services. Residents receiving care in SNFs include those whose illness, injury, or condition has resulted in a loss of function, and for whom rehabilitative care is expected to help regain that function. Treatment goals may include fostering residents' ability to manage their daily activities so that they can complete self-care and mobility activities as independently as possible, and, if feasible, return to a safe, active, and productive life in a community-based setting. Given that the primary goal of many SNF residents is improvement in function, SNF clinicians assess and document residents' functional status at admission and at discharge to evaluate the effectiveness of the rehabilitation care provided to individual residents and the SNF's effectiveness.

Examination of SNF data shows that SNF treatment practices directly influence resident outcomes. For example, therapy services provided to SNF residents have been found to be correlated with the functional improvement that SNF residents achieve (that is, functional outcomes). Several studies found patients' functional outcomes vary based on treatment by physical and occupational therapists. Specifically, therapy was associated with significantly greater odds of improving mobility and self-care functional independence, shorter length of stay, and a greater likelihood of discharge to community. Furthermore, Jung et al. found that an additional hour of therapy treatment per week was associated with approximately a 3.1 percentage-point increase in the likelihood of returning to the community among residents with a hip fracture. Achieving these targeted resident outcomes, including improved self-care and mobility functional independence, reduced length of stay, and increased discharges to the community, is a core goal of SNFs.

Among SNF residents receiving rehabilitation services, the amount of treatment received can vary. For example, the amount of therapy treatment provided varies by type (that is, for-profit versus not-for-profit) and facility location (that is, urban versus rural).

Measuring residents' functional improvement across all SNFs on an ongoing basis would permit identification of SNF characteristics, such as ownership types or locations, associated with better or worse resident risk adjusted outcomes and thus help SNFs optimally target quality improvement efforts.

MedPAC noted that while there was an overall increase in the share of intensive therapy days between 2002 and 2012, the for-profit and urban facilities had higher shares of intensive therapy than not-for-profit facilities and those located in rural areas. Data from 2011 to 2014 indicate that this variation is not explained by patient characteristics, such as activities of daily living, comorbidities and age, as SNF residents with stays in 2011 were more independent on average than the average SNF resident with stays in 2014. Because more intense therapy is associated with more functional improvement for certain beneficiaries, this variation in rehabilitation services supports the need to monitor SNF residents' functional outcomes. Therefore, we believe there is an opportunity for improvement in this area.

In addition, a recent analysis that examined the incidence, prevalence, and costs of common rehabilitation conditions found that back pain, osteoarthritis, and rheumatoid arthritis are the most common and costly conditions affecting more than 100 million individuals and costing more than $200 billion per year. Persons with these medical conditions are admitted to SNFs for rehabilitation treatment.

The use of standardized mobility and self-care data elements would standardize the collection of functional status data, which could improve communication when residents are transferred between providers. Most SNF residents receive care in an acute care hospital prior to the SNF stay, and many SNF residents receive care from another provider after the SNF stay.

Recent research provides empirical support for the risk adjustment variables for these quality measures. In a study of resident functional improvement in SNFs, Wysocki et al. found that several resident conditions were significantly related to resident functional improvement, including cognitive impairment, delirium, dementia, heart failure, and stroke. Also, Cary et al. found that several resident characteristics were significantly related to resident functional improvement, including age, cognitive function, self-care function at admission, and comorbidities.

These outcome-based quality measures could inform SNFs about opportunities to improve care in the area of function and strengthen incentives for quality improvement related to resident function.

We describe each of the four functional outcome quality measures below, and then follow with a discussion of the comments we received.

(a) Application of IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633)

The outcome quality measure, Application of IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633), is an application of the outcome measure finalized in the IRF QRP entitled, IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633). The quality measure estimates the mean risk-adjusted improvement in self-care score between admission and discharge among SNF residents. A summary of the NQF-endorsed quality measure specifications can be accessed on the NQF Web site: http://www.qualityforum.org/qps/2633. Detailed specifications for the NQF-endorsed quality measure can be accessed at http://www.qualityforum.org/ProjectTemplateDownload.aspx?SubmissionID=2633.

The functional outcome measure, the Application of IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633), requires the collection of admission and discharge functional status data by trained clinicians using standardized patient data elements that assess specific functional self-care activities such as shower/bathe self, dressing upper body and dressing lower body. These self-care items are daily activities that clinicians typically assess at the time of admission and/or discharge to determine residents' needs, evaluate resident progress, and/or prepare residents and families for a transition to home or to another provider. The standardized self-care function data elements are coded using a 6-level rating scale that indicates the resident's level of independence with the activity; higher scores indicate more independence. The outcome quality measure also requires the collection of risk factor data, such as resident functioning prior to the current reason for admission, bladder continence, communication ability and cognitive function, at the time of admission.

The data elements included in the quality measure were originally developed and tested as part of the PAC PRD version of the Continuity Assessment Record and Evaluation (CARE) Item Set, which was designed to standardize assessment of patients' and residents' status across acute and post-acute providers, including IRFs, SNFs, HHAs and LTCHs. The development of the CARE Item Set and a description and rationale for each item is described in a report entitled “The Development and Testing of the Continuity Assessment Record and Evaluation (CARE) Item Set: Final Report on the Development of the CARE Item Set: Volume 1 of 3.” Reliability and validity testing were conducted as part of CMS' Post-Acute Care Payment Reform Demonstration, and we concluded that the functional status items have acceptable reliability and validity. A description of the testing methodology and results are available in several reports, including the report entitled “The Development and Testing of the Continuity Assessment Record And Evaluation (CARE) Item Set: Final Report On Reliability Testing: Volume 2 of 3” and the report entitled “The Development and Testing of The Continuity Assessment Record And Evaluation (CARE) Item Set: Final Report on Care Item Set and Current Assessment Comparisons: Volume 3 of 3.” The reports are available on CMS' Post-Acute Care Quality Initiatives Web page at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/CARE-Item-Set-and-B-CARE.html.

(i) Stakeholder Input

A cross-setting function TEP convened by our measure development contractor on September 9, 2013 provided input on the initial technical specifications of this quality measure, Application of IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633). The TEP was supportive of the implementation of this measure and supported CMS's efforts to standardize patient/resident assessment data elements. The TEP summary report is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

The MAP met on December 14 and 15, 2015, and provided input on the measure, Application of IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633) for use in the SNF QRP. The MAP recognized that this quality outcome measure is an adaptation of a currently endorsed measure for the IRF population, and encouraged continued development to ensure alignment of this measure across PAC settings. The MAP noted there should be some caution in the interpretation of measure results due to resident differentiation between facilities. The MAP also noted possible duplication as the MDS already includes function data elements. We note that the data elements for the measure are similar, but not the same as the existing MDS Section G function data elements. The data elements for the measure include those that are the standardized patient assessment data for functional status under section 1899B(b)(1)(B)(i) of the Act. The MAP also stressed the importance of considering burden on providers when measures are considered for implementation. The MAP's overall recommendation was for “encourage further development.” More information about the MAP's recommendations for this measure is available at http://www.qualityforum.org/WorkArea/linkit.aspx?LinkIdentifier=id&ItemID=81593.

Since the MAP's review and recommendation for further development, we have continued to develop this measure by soliciting input via a TEP, providing a public comment opportunity, and providing an update on measure development to the MAP via the feedback loop. More specifically, our measure development contractor convened a SNF-specific function TEP on May 5, 2016, to provide further input on the technical specifications of this quality measure by reviewing the IRF specifications and the specifications of competing and related function quality measures. Overall, the TEP was supportive of the measure and supported our efforts to standardize patient assessment data elements. The SNF-specific function TEP summary report is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

We also solicited stakeholder feedback on the development of this measure by means of a public comment period that was open from October 7, 2016, until November 4, 2016. There was general support of the measure concept and the importance of functional improvement. Comments on the measure varied, with some commenters supportive of the measure, while others were either not in favor of the measure, or in favor of suggested potential modifications to the measure specifications. The public comment summary report for the measure is available on the CMS Web site at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

Further, we engaged with stakeholders when we presented an update on the development of this quality measure to the MAP on October 19, 2016, during a MAP feedback loop meeting. Slides from that meeting are available at http://www.qualityforum.org/WorkArea/linkit.aspx?LinkIdentifier=id&ItemID=83640.

(ii) Competing and Related Measures and Measure Justification

During the development of this proposed functional outcome measure, we have monitored and reviewed NQF-endorsed measures that are competing and/or related to the proposed quality measures. We identified six competing and related quality measures focused on self-care functional improvement for residents in the SNF setting entitled: (1) CARE: Improvement in Self Care (NQF #2613); (2) Functional Change: Change in Self-Care Score for Skilled Nursing Facilities (NQF #2769); (3) Functional Status Change for Patients with Shoulder Impairments (NQF #0426); (4) Functional Status Change for Patients with Elbow, Wrist and Hand Impairments (NQF #0427); (5) Functional Status Change for Patients with General Orthopedic Impairments (NQF #0428); and (6) Change in Daily Activity Function as Measures by the AM-PAC (NQF #0430). We reviewed the technical specifications for these six quality measures and compared these specifications to those of our outcome-based quality measure, the Application of IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633), and have noted the following differences in the technical specifications: (1) The number of risk adjustors and variance explained by these risk adjustors in the regression models; (2) the use of functional assessment items that were developed and tested for cross-setting use; (3) the use of items that are already on the MDS 3.0 and what this means for burden; (4) the handling of missing functional status data; and (5) the use of exclusion criteria that are baseline clinical conditions. We describe these key specifications of the proposed outcome measure, Application of IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633), in detail below.

Our literature review, input from technical expert panels, public comment feedback, and data analyses demonstrated the importance of adequate risk adjustment of admission case mix factors for functional outcome measures. Inadequate risk adjustment of admission case mix factors may lead to erroneous conclusions about the quality of care delivered within the facility, and thus is a potential threat to the validity of a quality measure that examines outcomes of care, such as functional outcomes. The quality measure, the Application of IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633) risk adjusts for more than 60 risk factors, explaining approximately 25 percent of the variance in change in function, and includes all of the following risk factors: prior functioning, prior device use, age, functional status at admission, primary diagnosis, and comorbidities. These risk factors are key predictors of functional performance and should be accounted for in any facility-level comparison of functional outcomes.

Another key feature of the measure, the Application of IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633), is that it uses the functional assessment data elements and the associated rating scale that were developed and tested for cross-setting use. The measure uses functional assessment items from the CARE Item Set, which were developed and tested as part of the PAC-PRD between 2006 and 2010. The items were designed to build on the existing science for functional assessment instruments, and included a review of the strengths and limitations of existing functional assessment instruments. An important strength of the standardized function items from the CARE instrument is that they allow comparison and tracking of patients' and residents' functional outcomes as they move across post-acute settings. Specifically, the CARE Item Set was designed to standardize assessment of patients' status across acute and post-acute settings, including SNFs, IRFs, LTCHs, and HHAs. The risk-adjustors for various setting-specific versions of this measure differ by the inclusion of adjustors such as comorbidities in the IRF measure. However, we believe that the differences in risk adjustment will not hinder future comparability across settings. Agencies such as MedPAC have supported a coordinated approach to measurement across settings using standardized patient data elements.

A third important consideration is that some of the data elements associated with the measure are already included on the MDS in section GG, because we adopted a cross-setting function process measure in the SNF QRP FY 2016 Final Rule (FR 80 46444 through 46453). Three of the self-care data elements necessary to calculate that quality measure, an Application of the Percent of Long-Term Care Hospital Patient with a Functional Assessment and a Care Plan that Addresses Function (NQF #2631) are used to calculate the quality measure. Provider burden of reporting on multiple items was a key consideration discussed by stakeholders in our recent TEP is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

We believe it is important to include the records of residents with missing functional assessment data when calculating a facility-level functional outcome quality measure for SNFs. The proposed measure, the Application of IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633), incorporates a method to address missing functional assessment data.

We believe certain clinically-defined exclusion criteria are important to specify in a functional outcome quality measure to maintain the validity of the quality measure. Exclusions for the quality measure, Application of IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633), were selected through a review of the literature, input from Technical Expert Panels, and input from the public comment process. The quality measure, Application of IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633) is intended to capture improvement in self-care function from admission to discharge for residents who are admitted with an expectation of functional improvement. Therefore, we exclude residents with certain conditions, for example progressive neurologic conditions, because these residents are typically not expected to improve on self-care skills for activities such as lower body dressing. Furthermore, we exclude residents who are independent on all self-care items at the time of admission, because no improvement in self-care can be measured with the selected set of items by discharge. Including residents with limited expectation for improvement could introduce incentives for SNFs to restrict access to these residents.

We would like to note that our measure developer presented and discussed these technical specification differentiations with TEP members during the May 6, 2016 TEP meeting to obtain TEP input on preferred specifications for valid functional outcome quality measures. The differences in measure specifications and the TEP feedback are presented in the TEP Summary Report, which is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html. Overall, the TEP supported the use of a risk adjustment model that addressed all of the following risk factors: Prior functioning, admission functioning, prior diagnosis and comorbidities. In addition, they supported exclusion criteria that would address functional improvement expectations of residents.

(iii) Data Collection Mechanism

Data for the quality measure, the Application of IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633), would be collected using the MDS, with the submission through the QIES ASAP system. For more information on SNF QRP reporting through the QIES ASAP system, refer to CMS Web site at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html. The calculation of the quality measure would be based on the data collection of standardized items to be included in the MDS. The function items used to calculate this measure are the same set of functional status data items that have been added to the IRF-PAI version 1.4, for the purpose of providing standardized resident assessment data elements under the domain of functional status, which is required by the IMPACT Act.

If finalized for implementation into the SNF QRP, the MDS would be modified so as to enable us to calculate this quality measure using additional data elements that are standardized with the IRF-PAI and such data would be obtained at the time of admission and discharge for all SNF residents covered under a Part A stay. The standardized items used to calculate this proposed quality measure do not duplicate existing Section G items currently used for data collection within the MDS. The quality measure and standardized data element specifications for the Application of IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633) can be found on the SNF QRP Measures and Technical Information Web site at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html.

(b) Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634)

This quality measure is an application of the outcome measure finalized in the IRF QRP entitled, IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634). This quality measure estimates the risk-adjusted mean improvement in mobility score between admission and discharge among SNF residents. A summary of this quality measure can be accessed on the NQF Web site: http://www.qualityforum.org/qps/2634. Detailed specifications for this quality measure can be accessed at http://www.qualityforum.org/ProjectTemplateDownload.aspx?SubmissionID=2634.

As previously noted, residents seeking care in SNFs include those whose illness, injury, or condition has resulted in a loss of function, and for whom rehabilitative care is expected to help regain that function. Several studies found patients' functional outcomes vary based on treatment. Physical and occupational therapy treatment was associated with greater functional gains, shorter stays, and a greater likelihood of a discharge to a community. Among SNF residents receiving rehabilitation services, the amount of therapy prescribed can vary widely, and this variation is not always associated with resident characteristics. This variation in rehabilitation services supports the need to monitor SNF resident's functional outcomes, as we believe there is an opportunity for improvement in this area.

The functional outcome measure, the Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634), requires the collection of admission and discharge functional status data by trained clinicians using standardized resident data elements that assess specific functional mobility activities such as toilet transfer and walking. These mobility items are daily activities that clinicians typically assess at the time of admission and/or discharge to determine resident's needs, evaluate resident progress, and prepare residents and families for a transition to home or to another care provider. The standardized mobility function items are coded using a 6-level rating scale that indicates the resident's level of independence with the activity; higher scores indicate more independence.

The functional assessment items included in the outcome quality measures were originally developed and tested as part of the Post-Acute Care Payment Reform Demonstration version of the CARE Item Set, which was designed to standardize assessment of patients' status across acute and post-acute providers, including SNFs, HHAs, IRFs, and LTCHs.

This outcome quality measure also requires the collection of risk factors data, such as resident functioning prior to the current reason for admission, history of falls, bladder continence, communication ability and cognitive function, at the time of admission.

A cross-setting function TEP convened by our measure development contractor on September 9, 2013 provided input on the initial technical specifications of this proposed quality measure, the Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634). The TEP was supportive of the implementation of this measure and supported our efforts to standardize patient/resident assessment data elements. The TEP summary report is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

The list of measures under consideration for the SNF QRP, including this quality measure, was released to the public on November 27, 2015, and early comments were submitted between December 1 and December 7, 2015. The MAP met on December 14 and 15, 2015, sought public comment on this measure from December 23, 2015, to January 13, 2015, and met on January 26 and 27, 2016. The NQF provided the MAP's input to us as required under section 1890A(a)(3) of the Act in the final report, MAP 2016 Considerations for Implementing Measures for Federal Programs: Post-Acute and Long-Term Care, which is available at http://www.qualityforum.org/Setting_Priorities/Partnership/MAP_Final_Reports.aspx. The MAP recognized that this measure is an adaptation of currently endorsed measures for the IRF population, and encouraged continued development to ensure alignment across PAC settings. They also noted there should be some caution in the interpretation of measure results due to patient/resident differentiation between facilities. To alignment across PAC settings, the self-care items included in the proposed quality measure are the same self-care items that are included in the IRF-PAI Version 1.4. We agree with the MAP that patient/resident populations can vary across IRFs and SNFs, and we have taken this issue into consideration while selecting and testing the risk adjustors, which include medical conditions, admission function, prior functioning and comorbidities. The risk-adjustors for the IRF and the SNF versions of this measure differ by the inclusion of adjustors such as comorbidities in the IRF measure. As noted, though there are differences between the measures we believe that the differences in risk adjustment will not hinder future comparability across measures.

The MAP also noted possible duplication as the MDS already includes function data elements. The data elements for the measure are similar, but not the same as the existing MDS Section G function data elements. The data elements for the measures include those that are the proposed standardized resident assessment data elements for function. The MAP also stressed the importance of considering burden on providers when measures are considered for implementation. We appreciate the issue of burden and have taken that into consideration in developing the measure. Please refer to the FY 2016 SNF PPS final rule (80 FR 46428) for more information on the MAP.

The MAP's overall recommendation was for “encourage further development.” More information about the MAP's recommendations for this proposed measure is available at http://www.qualityforum.org/WorkArea/linkit.aspx?LinkIdentifier=id&ItemID=81593.

Since the MAP's review and recommendation for further development, we have continued to develop this measure including soliciting input from a TEP, providing a public comment opportunity, and providing an update on measure development to the MAP via the feedback loop. More specifically, our measure development contractor convened a SNF-specific TEP on May 5, 2016 to provide further input on the technical specifications of this proposed quality measure by reviewing the IRF specifications and the specifications of competing and related function quality measures. Overall, the TEP was supportive of the measure and supported our efforts to standardize patient/resident assessment data elements. The SNF-specific function TEP summary report is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

We also solicited stakeholder feedback on the development of this measure by means of a public comment period open from October 7, until November 4, 2016. There was general support of the measure concept and the importance of functional improvement. Comments on the measure varied, with some commenters supportive of the measure, while others were either not in favor of the measure, or in favor of suggested potential modifications to the measure specifications. The public comment summary report for the proposed measure is available on the CMS Web site at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

We also engaged with the NQF convened MAP when we presented an update on the development of this quality measure on October 19, 2016, during a MAP feedback loop meeting. Slides from that meeting are available at http://www.qualityforum.org/WorkArea/linkit.aspx?LinkIdentifier=idamp;ItemID=83640.

During the development of this measure, we have monitored and reviewed NQF-endorsed measures that are competing and related. We identified seven competing and related quality measures focused on improvement in mobility for residents in the SNF setting entitled: (1) CARE: Improvement in Mobility (NQF #2612); (2) Functional Change: Change in Mobility Score (NQF 2774); (3) Functional Status Change for Patients with Knee Impairments (NQF #0422); (4) Functional Status Change for Patients with Hip Impairments (NQF #0423); (5) Functional Status Change for Patients with Foot and Ankle Impairments (NQF #0424); (6) Functional Status Change for Patients with Lumbar Impairments (NQF #0425); and (7) Change in Basic Mobility as Measures by the AM-PAC (NQF #0429). We reviewed the technical specifications for these seven measures carefully and compared them with the specifications of the proposed quality measure, the Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634) and have noted the following differences in the technical specifications: (1) The number of risk adjustors and variance explained by these risk adjustors in the regression models; (2) the use of functional assessment items that were developed and tested for cross-setting use; (3) the use of items that are already on the MDS 3.0 and what this means for burden; (4) the handling of missing functional status data; and (5) the use of exclusion criteria that are baseline clinical conditions. We describe these key specifications of the proposed outcome measure, the Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634), below in more detail.

Our literature review, input from technical expert panels, public comment feedback, and analyses demonstrated the importance of adequate risk adjustment of admission case mix factors for functional outcome measures. Inadequate risk adjustment of admission case mix factors may lead to erroneous conclusions about the quality of care delivered within the facility, and thus is a potential threat to the validity of a quality measure that examines outcomes of care, such as functional status. The quality measure, the Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634) risk adjusts for more than 60 risk factors, explaining approximately 23 percent of the variance in change in function, and includes all of the following risk adjusters: Prior functioning, prior device use, age, functional status at admission, primary diagnosis and comorbidities. These are key predictors of functional performance and need to be accounted for in any facility-level functional outcome quality measure.

Another key feature of the proposed measure, Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634), is that it uses the functional assessment data elements and the associated rating scale that were developed and tested for cross-setting use. The measure uses functional assessment items from the CARE Item Set, which were developed and tested as part of the PAC PRD between 2006 and 2010.

The items were designed to build on the existing science for functional assessment instruments, and included a review of the strengths and limitations of existing functional assessment instruments. An important strength of the cross-setting function items from the CARE instrument is that they allow tracking of patients' and residents' functional outcomes as they move across post-acute settings. Specifically, the CARE Item Set was designed to standardize assessment of patients' and residents' status across acute and post-acute settings, including SNFs, IRFs, LTCHs, and HHAs. MedPAC has publicly supported a coordinated approach to measurement across settings using standardized resident assessment data elements.

A third important consideration is that some of the data elements associated with the measure, Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634), are already included on the MDS in section GG, because we adopted a cross-setting function process measure in the SNF QRP FY 2016 Final Rule (FR 80 46444 through 46453), and seven of the mobility data elements necessary to calculate that quality measure, an Application of the Percent of Long-Term Care Hospital Patient with a Functional Assessment and a Care Plan that Addresses Function (NQF #2631) are used to calculate the proposed quality measure. Provider burden of reporting on multiple measures was a key consideration discussed by stakeholders in our recent TEP: https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

We believe it is important to include the records of residents with missing functional assessment data when calculating a facility-level functional outcome quality measure for SNFs. The measure, Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634), incorporates a method to address missing functional assessment data.

We believe certain clinically-defined exclusion criteria are important to specify in a functional outcome quality measure to maintain the validity of the quality measure. Exclusions for the proposed quality measure, Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634), were selected through a literature review, input from TEPs, and input from the public comment process. The Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634) is intended to capture improvement in mobility from admission to discharge for residents who are admitted with an expectation of functional improvement. Therefore, we exclude residents with certain conditions, for example progressive neurologic conditions, because these residents are typically not expected to improve on mobility skills for activities such as walking. Furthermore, we exclude residents who are independent on all mobility items at the time of admission, because no improvement can be measured with the selected set of items by discharge. Inclusion of residents with limited expectation for improvement could introduce incentives for SNF providers to limited access to these residents.

Our measure developer contractor presented and discussed these technical specification differentiations during the May 6, 2016 TEP meeting to obtain TEP input on preferred specifications for valid functional outcome quality measures. The differences in measure specifications and the TEP feedback are presented in the TEP Summary Report, which is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/ IMPACT-Act-Downloads-and-Videos.html.

Data for the quality measure, the Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634), would be collected using the MDS, with the submission through the QIES ASAP system. For more information on SNF QRP reporting through the QIES ASAP system, refer to https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html.

The calculation of the quality measure would be based on the data collection of standardized items to be included in the MDS. The function items used to calculate this measure are the same set of functional status data items that have been added to the IRF-PAI version 1.4, for the purpose of providing standardized resident assessment data elements under the domain of functional status. If this quality measure is finalized for implementation in the SNF QRP, the MDS would be modified so as to enable the calculation of these standardized items that are used to calculate this proposed quality measure. The collection of data by means of the standardized items would be obtained at admission and discharge. The standardized items used to calculate this quality measure do not duplicate existing items currently used for data collection within the MDS. The quality measure and standardized data element specifications for the Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634) is available on the SNF QRP Measures and Technical Information Web site at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html.

(c) Application of IRF Functional Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (NQF #2635)

This quality measure is an application of the outcome quality measure finalized in the IRF QRP entitled, IRF Functional Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (NQF #2635). The quality measure estimates the percentage of SNF residents who meet or exceed an expected discharge self-care score. A summary of this quality measure can be accessed on the NQF Web site at http://www.qualityforum.org/qps/2635. Detailed specifications for the quality measure can be accessed at http://www.qualityforum.org/ProjectTemplateDownload.aspx?SubmissionID=2635.

As previously noted, residents seeking care in SNFs include individuals whose illness, injury, or condition has resulted in a loss of function, and for whom rehabilitative care is expected to help regain that function. Several studies found patients' functional outcomes vary based on treatment by physical and occupational therapists. Therapy was associated with greater functional gains, shorter stays, and a greater likelihood of discharge to community. Among SNF residents receiving rehabilitation services, the amount of treatment prescribed can vary widely, and this variation is not associated with resident characteristics. This variation in rehabilitation services supports the need to monitor SNF resident's functional outcomes, as we believe there is an opportunity for improvement in this area.

The outcome quality measure, Application of IRF Functional Outcome Measure: Discharge Self-Care Score or Medical Rehabilitation Patients (NQF #2635), requires the collection of functional status data at admission and discharge by trained clinicians using standardized resident assessment data elements such as eating, oral hygiene, and lower body dressing. These self-care items are daily activities that clinicians typically assess at the time of admission and discharge to determine residents' needs, evaluate resident progress, and prepare residents and families for a transition to home or to another provider. The self-care function data elements are coded using a 6-level rating scale that indicates the resident's level of independence with the activity; higher scores indicate more independence.

The functional assessment items included in the outcome quality measures were originally developed and tested as part of the Post-Acute Care Payment Reform Demonstration version of the CARE Item Set, which was designed to standardize assessment of patients' status across acute and post-acute providers, including SNFs, HHAs, IRFs, and LTCHs.

This outcome quality measure also requires the collection of risk factors data, such as resident functioning prior to the current reason for admission, bladder continence, communication ability, and cognitive function at the time of admission.

A cross-setting function TEP convened by our measure development contractor on September 9, 2013 provided input on the initial technical specifications of this proposed quality measure, the Application of IRF Functional Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (NQF #2635). The TEP was supportive of the implementation of this measure and supported CMS's efforts to standardize patient/resident assessment data elements. The TEP summary report is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

The MAP met on December 14 and 15, 2015, and provided input on the proposed measure, Application of IRF Functional Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (NQF #2635) for use in the SNF QRP. The MAP recognized that this quality measure is an adaptation of a currently endorsed measure for the IRF population, and encouraged continued development to ensure alignment of this measure across PAC settings. The MAP also noted there should be some caution in the interpretation of measure results due to patient/resident differentiation between facilities. The MAP also stressed the importance of considering burden on providers when measures are considered for implementation. The MAP also noted possible duplication as the MDS already includes function data elements. The data elements for the proposed measure are similar, but not the same as the existing MDS function data elements. The data elements for the measures include those that are the proposed standardized assessment data elements for function. The MAP's overall recommendation was to “encourage further development.” More information about the MAP's recommendations for this measure is available at http://www.qualityforum.org/WorkArea/linkit.aspx?LinkIdentifier=id&ItemID=81593.

Since the 2015 MAP's review and recommendation for further development, we have continued to develop this measure including soliciting input via a TEP, proving a public comment opportunity and providing an update on measure development to the MAP via the feedback loop. More specifically, our measure development contractor convened a SNF-specific TEP on May 5, 2016 to provide further input on the technical specifications of this quality measure by reviewing the IRF specifications and the specifications of competing and related function quality measures. Overall, the TEP was supportive of the measure. Specifically, they supported the risk adjustors, suggested some additional risk adjustors, supported the exclusion criteria and supported CMS's efforts to standardize patient/resident assessment data elements. The SNF-specific function TEP summary report is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

We also solicited stakeholder feedback on the development of this measure by means of a public comment period open from October 7, 2016 until November 4, 2016. There was general support of the measure concept and the importance of functional improvement. Comments on the measure varied, with some commenters supportive of the measure, while others were either not in favor of the measure, or in favor of suggested potential modifications to the measure specifications. Some comments focused on suggestions for additional risk adjustors, and the data elements. The public comment summary report for the measure is available on the CMS Web site at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

We also engaged with stakeholders when we presented an update on the development of this quality measure to the MAP on October 19, 2016, during a MAP feedback loop meeting. Slides from that meeting are available at http://www.qualityforum.org/WorkArea/linkit.aspx?LinkIdentifier=id&ItemID=83640.

During the development of this measure, we monitored and reviewed NQF-endorsed measures that are competing and related. We identified six competing and related quality measures focused on self-care functional improvement for residents in the SNF setting entitled: (1) CARE: Improvement in Self Care (NQF #2613); (2) Functional Change: Change in Self-Care Score (NQF #2286); (3) Functional Status Change for Patients with Shoulder Impairments (NQF #0426); (4) Functional Status Change for Patients with Elbow, Wrist and Hand Impairments (NQF #0427); (5) Functional Status Change for Patients with General Orthopedic Impairments (NQF #0428); and (6) Change in Daily Activity Function as Measures by the AM-PAC (NQF #0430).

As described above, we reviewed the technical specifications for these six measures and compared them with the specifications for the quality measure, Application of IRF Functional Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (NQF #2635) and, as described in detail above, we noted the following differences in the technical specifications: (1) The number of risk adjustors and variance explained by these risk adjustors in the regression models; (2) the use of functional assessment items that were developed and tested for cross-setting use; (3) the use of items that are already on the MDS 3.0 and what this means for burden; (4) the handling of missing functional status data; and (5) the use of exclusion criteria that are baseline clinical conditions.

Consistent with the other functional outcome measures, the specifications for this quality measure, Application of IRF Functional Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (NQF #2635), were developed based on our literature review, input from technical expert panels, public comment feedback and data analyses. The details about the specifications for the measures described above also apply to this quality measure. Overall, the TEP supported the use of a risk adjustment model that addressed prior functioning, admission functioning, prior diagnosis and comorbidities. In addition, they supported exclusion criteria that would address functional improvement expectations of residents.

Our measure developer contractor presented and discussed these technical specification differentiations during the May 6, 2016 TEP meeting to obtain TEP input on preferred specifications for valid functional outcome quality measures. The differences in measure specifications and the TEP feedback are presented in the TEP Summary Report, which is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

Data for the quality measure, the Application of IRF Functional Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (NQF #2635), would be collected using the MDS, with the submission through the QIES ASAP system. For more information on SNF QRP reporting through the QIES ASAP system, refer to CMS Web site at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html.

The calculation of the proposed quality measure would be based on the data collection of standardized items to be included in the MDS. The function items used to calculate this measure are the same set of functional status data items that have been added to the IRF-PAI version 1.4, for the purpose of providing standardized resident assessment data elements under the domain of functional status.

The collection of data by means of the standardized items would be obtained at admission and discharge. The standardized items used to calculate this quality measure do not duplicate existing items currently used for data collection within the MDS. The quality measure and standardized data element specifications for the Application of IRF Functional Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (NQF #2635) can be found on the SNF QRP Measures and Technical Information Web site at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html.

If finalized for implementation into the SNF QRP, the MDS would be modified so as to enable us to calculate the proposed measure using additional data elements that are standardized with the IRF-PAI and such data would be obtained at the time of admission and discharge for all SNF residents covered under a Part A stay.

(d) Application of IRF Functional Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (NQF #2636)

This quality measure is an application of the outcome quality measure finalized in the IRF QRP entitled, IRF Functional Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (NQF #2636). This quality measure estimates the percentage of SNF residents who meet or exceed an expected discharge mobility score. A summary of this quality measure can be accessed on the NQF Web site: http://www.qualityforum.org/qps/2636. Detailed specifications for this quality measure can be accessed at http://www.qualityforum.org/ProjectTemplateDownload.aspx?SubmissionID=2636.

As previously noted, residents seeking care in SNFs include individuals whose illness, injury, or condition has resulted in a loss of function, and for whom rehabilitative care is expected to help regain that function. Several studies found patients' functional outcomes vary based on treatment by physical and occupational therapists. Therapy was associated with greater functional gains, shorter stays, and a greater likelihood of discharge to community. Among SNF residents receiving rehabilitation services, the amount of treatment prescribed can vary widely, and this variation is not associated with resident characteristics. This variation in rehabilitation services supports the need to monitor SNF resident's functional outcomes, as we believe there is an opportunity for improvement in this area.

The functional outcome measure, Application of IRF Functional Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (NQF #2636), requires the collection of admission and discharge functional status data by trained clinicians using standardized resident data elements that assess specific functional mobility activities such as bed mobility and walking. These standardized mobility items are daily activities that clinicians typically assess at the time of admission and/or discharge to determine residents' needs, evaluate resident progress and prepare residents and families for a transition to home or to another care provider. The standardized mobility function items are coded using a 6-level rating scale that indicates the resident's level of independence with the activity; higher scores indicate more independence.

The functional assessment items included in the outcome quality measures were originally developed and tested as part of the Post-Acute Care Payment Reform Demonstration version of the CARE Item Set, which was designed to standardize assessment of patient or resident status across acute and post-acute providers, including SNFs, HHAs, IRFs, and LTCHs.

This quality measure requires the collection of risk factors data, such as resident functioning prior to the current reason for admission, history of falls, bladder continence, communication ability and cognitive function, at the time of admission.

A cross-setting function TEP convened by our measure development contractor on September 9, 2013 provided input on the initial technical specifications of this quality measure, Application of IRF Functional Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (NQF #2636). The TEP was supportive of the implementation of this measure and supported our efforts to standardize patient assessment data elements. The TEP summary report is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

The MAP met on December 14 and 15, 2015, and provided input on the measure, Application of IRF Functional Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (NQF #2636), for use in the SNF QRP. The MAP recognized that this quality measure is an adaptation of a currently endorsed measure for the IRF population, and encouraged continued development to ensure alignment of this measure across PAC settings. The MAP noted there should be some caution in the interpretation of measure results due to patient/resident differentiation between facilities. The MAP also stressed the importance of considering burden on providers when measures are considered for implementation. The MAP also noted possible duplication as the MDS already includes function data elements. The data elements for the proposed measure are similar, but not the same as the existing MDS function data elements. The data elements for the measure include those that are the standardized patient data elements for function. The MAP's overall recommendation was to “encourage further development.” More information about the MAP's recommendations for this proposed measure is available at http://www.qualityforum.org/WorkArea/linkit.aspx?LinkIdentifier=id&ItemID=81593.

Since the MAP's review and recommendation for further development, we have continued to develop this measure including soliciting input via a TEP, proving a public comment opportunity and providing an update on measure development to the MAP via the feedback loop. More specifically, our measure development contractor convened a SNF-specific TEP on May 5, 2016, to provide further input on the technical specifications of this quality measure by reviewing the IRF specifications and the specifications of competing and related function quality measures. Overall, the TEP was supportive of the measure and supported our efforts to standardize patient/resident assessment data elements. The SNF-specific function TEP summary report is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

We also solicited stakeholder feedback on the development of this measure by means of a public comment period open from October 7, 2016, until November 4, 2016. There was general support of the measure concept and the importance of functional improvement. Comments on the measure varied, with some commenters supportive of the measure, while others were either not in favor of the measure, or suggested potential modifications to the measure specifications.

The public comment summary report for the proposed measure is available on the CMS Web site at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

We also engaged with stakeholders when we presented an update on the development of this quality measure to the MAP on October 19, 2016, during a MAP feedback loop meeting. Slides from that meeting are available at http://www.qualityforum.org/WorkArea/linkit.aspx?LinkIdentifier=id&ItemID=83640.

During the development of this measure, we have monitored and reviewed the NQF-endorsed measures that are competing and related. We identified seven competing and related quality measures focused on mobility functional improvement for residents in the SNF setting entitled: (1) CARE: Improvement in Mobility (NQF #2612); (2) Functional Change: Change in Mobility Score (NQF #2774); (3) Functional Status Change for Patients with Knee Impairments (NQF #0422); (4) Functional Status Change for Patients with Hip Impairments (NQF #0423); (5) Functional Status Change for Patients with Foot and Ankle Impairments (NQF #0424); (6) Functional Status Change for Patients with Lumbar Impairments (NQF #0425); and (7) Change in Basic Mobility as Measures by the AM-PAC (NQF #0429). As described above, we reviewed the technical specifications for these seven measures carefully and compared them with the specifications of the proposed quality measure, Application of IRF Functional Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (NQF #2636) and have noted the following differences in the technical specifications: (1) The number of risk adjustors and variance explained by these risk adjustors in the regression models; (2) the use of functional assessment items that were developed and tested for cross-setting use; (3) the use of items that are already on the MDS 3.0 and what this means for burden; (4) the handling of missing functional status data; and (5) the use of exclusion criteria that are baseline clinical conditions.

Consistent with the other functional outcome measures, the specifications for this quality measure, Application of IRF Functional Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (NQF #2636), were developed based on our literature review, input from technical expert panels, public comment feedback and data analyses. The details about how the specifications for the measures differ as described in the previous functional outcome measure sections, also apply to this quality measure.

Our measure developer contractor presented and discussed these technical specification differentiations during the May 6, 2016 TEP meeting to obtain TEP input on preferred specifications for valid functional outcome quality measures. The differences in measure specifications and the TEP feedback are presented in the TEP Summary Report, which is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

Data for the quality measure, the Application of IRF Functional Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (NQF #2636), would be collected using the MDS, with the submission through the QIES ASAP system. Additional information on SNF QRP reporting through the QIES ASAP system can be found on the CMS Web site at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html.

The calculation of the quality measure would be based on the data collection of standardized items to be included in the MDS. The function items used to calculate this measure are the same set of functional status data items that have been added to the IRF-PAI version 1.4, for the purpose of providing standardized resident assessment data elements under the domain of functional status.

The collection of data by means of the standardized items would be obtained at admission and discharge. The standardized items used to calculate this quality measure do not duplicate existing items currently used for data collection within the MDS. The quality measure and standardized resident data element specifications for the Application of IRF Functional Outcome Measure: Discharge Change in Mobility Score for Medical Rehabilitation Patients (NQF #2636) can be found on the SNF QRP Measures and Technical Information Web site at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html.

If finalized for implementation into the SNF QRP, the MDS would be modified so as to enable us to calculate the measure using additional data elements that are standardized with the IRF-PAI and such data would be obtained at the time of admission and discharge for all SNF residents covered under a Part A stay.

We sought public comments on our proposal to adopt the four functional outcome quality measures, entitled Application of IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633); Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634);, Application of IRF Functional Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (NQF #2635); and Application of IRF Functional Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (NQF #2636), beginning with the FY 2020 SNF QRP. All of the comments we received addressed all four measures, and our discussion of them follows.

Comment: Several stakeholders supported the adoption of all four functional status quality measures into the SNF QRP. One commenter noted that self-care and mobility are of particular concern for persons with advanced illness. This commenter further noted that function affects daily life and quality of life for both persons and caregivers, and that tracking this information during a SNF stay and at discharge would improve transitions. The commenter encouraged us to increase measurement of functional status for all patients in all settings. Another commenter who supported the measures noted that valid and reliable measures of functional outcomes are important for informing treatment planning. Two commenters supported all 4 functional status quality measures in the SNF setting, and noted their general support for quality measures in all PAC settings that assess functional status and the real-life needs of beneficiaries. These two commenters believe that these four functional outcome measures move the SNF QRP in this direction. Another commenter stated that having a core set of data elements will allow for tracking of function across the continuum of care and is in alignment with the goals of the IMPACT Act. Another commenter supported our efforts to improve quality of care and ensure appropriate resource allocation among PAC settings, and specifically voiced agreement for adapting the NQF-endorsed functional outcome measures from the IRF setting to the SNF setting to align measures noting the intent of the IMPACT Act. This commenter stated that measures should be clinically relevant, representative for a given setting and patient population, and meaningful to patients and families.

Response: We appreciate the commenters' support for the four functional status outcome quality measures that we proposed to adopt for the SNF QRP. We agree that patient and resident functioning in the areas of self-care care and mobility are clinically relevant and are an important area of quality in post-acute care (PAC) settings. In addition, we believe that examining resident functioning during the SNF stay will help SNFs focus on optimizing residents' functioning and discharge planning and support residents' transitions from the SNF to home or another setting. Finally, we agree that valid and reliable measures of functional outcomes will assist SNFs in planning treatment aimed at increasing or maintaining functional status.

Comment: One commenter offered support for these measures in concept, but expressed concern that the proposed measures have not been tested in the SNF setting. The commenter recommended that testing across population types take place prior to any public reporting to avoid confusion among providers and consumers.

Response: CMS strongly agrees that item and quality measure validity and reliability are important. The self-care and mobility items underwent several types of testing across post-acute care settings, including SNFs, as part of the Post-Acute Care Payment Reform Demonstration (PAC PRD). This testing, which included data from 60 SNFs (contributing almost 4,000 CARE assessments) examined the items' feasibility, reliability, and validity. Overall, these results indicate moderate to substantial agreement on these items. Details regarding the reliability and validity testing, can be found in reports entitled The Development and Testing of the Continuity Assessment Record and Evaluation (CARE) Item Set, Volumes 1 through 3, Continuity Assessment Record and Evaluation (CARE) Item Set: Video Reliability Testing, and Continuity Assessment Record and Evaluation (CARE) Item Set: Additional provider-Type Specific Interater Reliability Analyses. These reports are available at http://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/CARE-Item-Set-and-B-CARE.html.

As part of our quality measure development work, we conducted additional reliability and validity testing, including Rasch analysis, which showed acceptable reliability and validity, and these results were discussed during the May 2016 TEP meeting and are summarized in the SNF Function TEP Summary Report, which is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html. Therefore, given the overall findings of these reliability and validity analyses, we believe that the proposed functional outcome measures are sufficiently reliable for the SNF QRP.

In addition, beginning October 1, 2016, SNFs are reporting several of the self-care and mobility data elements that are needed to calculate these measures. The quality measure, an Application of the Percent of LTCH Patients with a Functional Assessment and a Care Plan that Addresses Function (NQF #2631), was finalized for use in the SNF QRP in FY 2016 (80 FR 46444 through 46453). This process measure includes several of the self-care and mobility items included in the SNF functional outcome measures, and we are conducting tests of the reliability and validity of that data. We conduct ongoing analysis of reliability and validity of adopted measures.

Comment: One commenter did not support the proposed function measures because the NQF has not endorsed them for the SNF setting and the Measure Applications Partnership (MAP) recommended continued development. Two commenters recommended that we seek rapid NQF endorsement for the four outcome measures to remove the “application of” and “IRF” wording from the measure titles and to prevent confusion among consumers, policymakers, and payers when displayed. One of these commenters stated that quality performance outcomes reported by an NQF endorsed measure in one setting may not necessarily be comparable to an “application” of the same measure in another setting due to differences in patient populations, payment policy, and specific measure calculation details, case mix adjustors such as co-morbidities, and other measure details. Another commenter recommended that the official name of the proposed measure distinguish them as SNF quality measures, which would decrease the public confusion when viewing them on Nursing Home Compare.

Response: While these measures are not currently NQF-endorsed for SNFs, we recognize that the NQF endorsement process is an important part of measure development and plan to submit these four measures for consideration of NQF endorsement after one full year of data collection. We initially presented the four SNF outcome measures to the MAP in December 2015. After the MAP meeting, we continued development as recommended. Our measure developer contractor convened a SNF Function TEP in May 2016 and we then requested and received public comment via the CMS Measures Management Web site. In October 2016, we presented a review of our additional measure development work to the MAP as part of the feedback loop to give an update on the measure development activities.

We appreciate the comments pertaining to NQF endorsement of the measures before they are publicly displayed and comments on the titling of the proposed functional outcome measures. With regard to the measure title, we recognize the confusion of leveraging the words “IRF” in our title application when we are collecting for a SNF population, and we will reassess the titling for these outcome measures to decrease confusion among all stakeholders.

Comment: Several commenters expressed concern about the added burden of collecting data for the functional outcome measures. One commenter noted that the addition of the section GG items needed for the function outcome measures will increase the time providers need to complete residents' assessments. A few commenters stated that changes in the MDS as a result of these measures will involve additional staff time and resources for training and monitoring compliance. One commenter suggested that we provide financial support for the additional reporting burden.

Response: We appreciate the commenters' concerns associated with the proposed functional outcome measures. We recognize that any new data collection is associated with burden and take such concerns under consideration when developing and selecting quality measures. As we develop quality measures, we review existing items and consider the appropriateness of adding or deleting any items. We note that some of the data elements associated with the measure are already included on the MDS in section GG, because we adopted a cross-setting function process measure in the SNF QRP FY 2016 Final Rule (80 FR 46444 through 46453). Three of the self-care data elements and seven mobility data elements necessary to calculate that quality measure, an Application of the Percent of Long-Term Care Hospital Patient with a Functional Assessment and a Care Plan that Addresses Function (NQF #2631) are used to calculate the quality measure and are finalized in this rule as standardized resident assessment data elements.

Comment: Three commenters noted that the requirement to assess residents while utilizing both the section G—Functional Status and section GG—Functional Abilities and Goals items on the MDS is burdensome. One of the commenters explained that to address the same functional activities in two different sections of the MDS, with different item definitions, and with different look-back periods, is excessively burdensome, and introduces unnecessary risk for reporting errors. The two other commenters further suggested that we analyze the section G mobility and self-care items that address the same or similar domains in section GG to identify opportunities to eliminate the redundant and non-compliant mobility and self-care items from section G.

Response: We recognize that the items in section G and section GG address similar domains of mobility and self-care. However, for the SNF QRP, we believe that the section GG items and the associated 6-level scale will allow us to better distinguish change at the highest and lowest levels of functioning by documenting minimal change from no change at the low end of the scale. This is important for measuring progress in some of the most complex cases treated in PAC. The items in section GG were developed with input from the clinical therapy communities to better measure the change in function, regardless of the severity of the individual's functional limitations. To reduce the potential burden associated with collecting additional items, we have included several mechanisms in the section GG to reduce the number of items that apply to any one resident. First, in section GG, there are skip patterns pertaining to walking and wheelchair mobility that allow the clinician to skip items if the resident does not walk or does not use a wheelchair, respectively. The skip patterns mean that only a subset of section GG items are needed for most residents. Second, section GG items will only be collected at admission and discharge.

Comment: Two comments requested more detailed information about how the functional outcome measures could be used to improve quality and how we expect to use the information.

Response: We believe that examining residents' functional outcomes will help SNF staff focus on optimizing patients' functioning and supporting patients' transition from the SNF to home or another setting. Furthermore, we believe that the feedback we provide to SNFs on these measures will allow providers to monitor their performance on key rehabilitation outcomes, relative to other facilities, and identify opportunities to improve their quality of care.

Comment: One commenter voiced concern about the proposal to include functional outcome measures that focus on functional improvement without also proposing measures that cover SNF residents who are in the facility for functional maintenance or the prevention or slowing of functional decline. The commenter stated that the standards of care and goals for patients in an IRF cannot be adopted for SNFs unless an additional measure that focuses on residents covered under functional maintenance is also adopted. The commenter further noted that adoption of the four functional outcome measures will send the wrong message to SNFs and indicate they are being judged solely on whether they improve residents' functioning. The commenter recommends delaying implementation of these measures until a maintenance measure can also be implemented simultaneously. This commenter disagreed that the exclusion of patients not receiving physical therapy or occupational therapy is an appropriate proxy for SNF residents for whom there is no expectation of functional improvement and suggested we consider another measure that does not penalize SNFs that provide maintenance therapy.

Response: We agree that our measures should address maintenance and the prevention or slowing of functional decline, and we note that the functional process measure, Application of Percent of LTCH Patients with an Admission and Discharge Functional Assessment and a Care Plan that Addresses Function (NQF #2631), which is already included in the SNF QRP measure set, addresses this topic. The functional process measure requires that a SNF conduct a functional assessment at both admission and discharge and that such assessment include at least one goal related to function. Such functional status goals may focus on maintenance of function, slowing decline in function or functional improvement. Likewise, the proposed discharge functional outcome measures, Application of the IRF Function Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (NQF #2635) and Application of the IRF Function Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (NQF #2636), calculate the residents' observed and expected discharge functional status. Maintenance of function or slowed decline in function may be expected based on the resident's characteristics and this would be captured in these measures. We also support future quality measurement work that will assess the development of other measures that focus on maintaining function and the slowing of functional decline.

Finally, we would like to note that the Nursing Home Quality Initiative includes two quality measures focused on functional maintenance and slowing decline. These measures are reported to the public on the Nursing Home Compare Web site and are calculated using MDS Section G data elements. We intend to develop similar quality measures focused on maintenance of function and decline in function that would be calculated using section GG Self-Care and Mobility data elements. With regard to unintended consequences, we will monitor potential unintended consequences of this exclusion criterion, and take these suggestions into consideration during our ongoing efforts to improve our quality measures.

Comment: One commenter agreed with the exclusion of residents who do not have an expectation of functional improvement for the 2 change functional outcome measures (Application of IRF Functional Outcome Measure: Change in Self-Care for Medical Rehabilitation Patients (NQF #2633) and Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634)) and requested clarification as to how we would identify these residents. The commenter requested additional detail regarding residents who qualify for this exclusion at admission and for residents whose status changes during the SNF stay. The commenter noted that to ensure accurate and appropriate identification of beneficiaries who qualify for this exclusion, CMS needs to provide more detail regarding it. One commenter stated that we should provide additional information regarding how SNFs will be held accountable if the goal changes from expecting functional improvement in a resident to not expecting functional improvement during the resident's stay. Another commenter also voiced concern that changes in residents' goals between admission and discharge are common and would impact outcomes.

Response: For this exclusion criterion, we provide the list of medical conditions that we will use in the Final Rule Specifications for SNF QRP Quality Measures document, which is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html.

We recognize that a resident's status or goals may change during the SNF stay, and the measures include several exclusions that are applied based on the resident's status at discharge to reflect this change prior to the end of the stay. For example, a resident may experience an incomplete stay due to an urgent medical condition and is discharged to an acute care hospital. We recognize that it is challenging to collect discharge functional assessment data under these circumstances. For this reason, these residents are excluded from the four functional outcome measures. We would also like to clarify that the collection of a patient's goal is simply to track whether a patient's goal was established on admission rather than to track the expectation of function improvement.

Another exclusion criterion in the 4 functional outcome measures relates to residents who are discharged to hospice. This may be a circumstance where a resident's status changed during the stay due to a new medical diagnosis or an unexpected worsening of a resident's condition. The list of all measure exclusions and the specifications for each of these exclusion criteria are provided in the Final Rule Specifications for SNF QRP Quality Measures document. We will continue to monitor for other examples as part of our ongoing quality measure development work.

Comment: Several commenters disagreed with one proposed exclusion criteria, Residents who do not receive physical or occupational therapy services. Two commenters suggested that we adopt more person-centered criteria that reflect functional improvement expectations in addition to or to replace the current proposed exclusion that focuses on therapy services. The two commenters stated that providers who administer therapy services to residents to maintain, but not improve function, would have lower functional improvement scores and the criterion “creates a significant disincentive to provide any physical therapy (PT) or occupational therapy (OT) to SNF residents that require skilled services to maintain or delay decline in function.” One of the two commenters stated this may be a disincentive to provide therapy to residents who fit into the Jimmo class of beneficiaries who may not improve but still need SNF services. One of these commenters recommended that CMS exclude residents whose aggregate “Admission Performance” mobility (GG01701) or self-care (GG01301) score (see Step 1 of the CMS proposed quality measures algorithms) is greater than or equal to their “Discharge Goal” mobility (GG01702) or self-care (GG01302) score. Another commenter opposed excluding from the functional outcome measures residents who do not receive occupational therapy or physical therapy.

One commenter who disagreed with the proposed exclusions criterion further noted that the exclusion of “residents who do not receive physical or occupational therapy services,” for the 4 functional outcome measures is substantively different than the May 2016 SNF Function TEP discussion, and the 2016 CMS Measurement Management Public Comment document. This commenter recognized that the exclusion did refer to “Residents who do not have an expectation of functional improvement,” which was subsequently clarified to exclude “Residents who do not receive physical or occupational therapy services.” The commentator expressed that no explanation or data analysis was provided to justify the change in the exclusion definition.

Response: We thank the commenters for their feedback and suggestions. We acknowledge the commenters' concern about excluding residents who do not receive physical or occupational therapy services. As noted in the SNF Function TEP Report, our measure development contractor did solicit suggestions from TEP members about methods to operationalize exclusion criteria so that the quality measure would include only residents who were expected to improve functional status, and TEP members did not offer a specific recommendation to address this issue. For residents who are expected to improve their functional abilities, physical and/or occupational therapy would be part of the resident's care plan to assist the resident to relearn how to perform the activity or to learn a new way to perform the activity. With regard to the commenter's suggestion to exclude residents whose aggregate “Admission Performance” is greater than or equal to their “Discharge Goal,” we would like to clarify that the Function Process Measure requires SNFs to code at least one Discharge Goal item on the 5-day admission assessment. The suggestion would require SNFs to code all function Discharge Goal items, which is not currently required, and this would incur a significant burden on SNFs.

Comment: MedPAC noted the importance of monitoring the accuracy of data that is reported on measures that assess functional status.

Response: We agree with MedPAC on the importance of monitoring the accuracy of functional status data that is reported to CMS, as data accuracy is necessary to calculate reliable and valid quality measures. To that end, we conduct ongoing analyses of the assessment data submitted from PAC providers to ensure accuracy by examining the reliability and validity of the data elements on a quarterly basis.

Comment: One commenter cautioned that the education level and professional expertise of personnel collecting SNF functional outcome measure data are important to consider when analyzing and drawing conclusions about the data.

Response: We recognize that each SNF may have unique workflow issues, which may mean that data collection protocols are not exactly alike. However, we require that SNFs submit accurate data, and we provide training and other resources.

Comment: One commenter supported the general numerator and denominator definitions proposed for the four proposed SNF functional outcome measures.

Response: We appreciate the commenter's support.

Comment: One commenter expressed support for the denominator exclusion criteria proposed for the four proposed SNF functional outcome measures.

Response: We appreciate the commenter's support.

Comment: One commenter expressed concern regarding the exclusion, “residents who are scored as independent upon admission,” from the change in self-care score measure and the inclusion of these residents in the self-care discharge score measure. The commenter explained that this will cause confusion among providers, and recommended that further education be offered to providers.

Response: This exclusion criterion only applies to the two change quality measures (Application of IRF Functional Outcome Measure: Change in Self-Care for Medical Rehabilitation Patients (NQF #2633) and Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634)), and is related to a measurement issue. A resident who is independent with each of the self-care or mobility activities in section GG at the time of admission would be coded a 6 on each of those items, and any improvement in self-care or mobility skills the resident achieved during the stay could not be measured with the same set of function data elements and rating scale at discharge. Therefore, residents who are at the “ceiling” of the self-care or mobility scale at the start of a SNF stay are excluded from the respective change in self-care or change in mobility quality measure. Including these residents in a change quality measure may disadvantage providers serving these residents, as the change in self-care or mobility could not be mathematically higher than zero. We would like to note that residents who are independent with all self-care or mobility activities are included in the discharge self-care and the discharge mobility quality measures, and for the discharge quality measures, maintaining independence with all the self-care or mobility activities is the expected outcome. With regard to provider knowledge about this topic, we recognize the importance of comprehensive training and we intend to provide such training.

Comment: Two commenters noted that the calculation of the 4 functional outcome quality measures requires recoding of “activity did not occur” codes. These commenters expressed concern about recoding the “activity did not occur” codes (that is, codes 07, 09, 88) to 01—Dependent, and one of the two commenters did not support recoding of missing data as the method was not clear. [The other commenter expressed concern that recoding the activity not attempted codes to 01 will not accurately reflect resident status or change, and that mobility and self-care tasks being refused, not applicable, or not attempted due to medical or safety concerns, does not necessarily mean the resident is dependent.

Another commenter noted that this recoding can result in different statistical and clinical inferences compared to not recoding items to 01. The commenter recommended further detail regarding the use of “activity did not occur” codes and that an analysis be conducted that compares the recoding method to excluding any or all the four “activity did not occur” item responses, and provide the percentage of patient stays impacted. The commenter requested that these results be shared with stakeholders for comment before adopting these four proposed functional outcomes measures.

Response: We appreciate the concerns presented by commenters about handling missing data and the “activity not attempted” codes. “Activity did not occur” codes and missing data are recoded to 01. Dependent to calculate the quality measure. The rationale for this recoding relates to the likelihood that when a resident cannot attempt an activity due to a medical condition or safety concern, that the resident often would have required significant assistance from one or more helpers to complete the activity had the activity been attempted. Thus, the resident would have been considered dependent with the activity. Likewise, the code 09, “Not applicable,” is used to indicate that the activity was not attempted, and that the resident did not perform the activity prior to the current illness, injury or exacerbation. We believe our re-coding approach is better than excluding any resident stays that include one or more items coded as “activity not attempted,” because excluding these residents would exclude residents who, in general, are lower functioning. That said, we are exploring other methods of recoding items when an activity was not attempted. We believe it is important to continue to monitor the reliability and validity of the functional outcome measures, including issues such as this one. Ongoing analyses of these items and outcomes may provide support for an alternative approach to item recoding in the future.

Comment: One commenter conditionally supported the inclusion of only Medicare Part A residents, but requested that we consider revising this criterion in the future to include SNF Medicare Advantage enrollees. The commenter noted that with growing enrollment in the Medicare Advantage program, excluding these beneficiaries may result in the outcome measure not adequately representing quality of care for the entire SNF. The commenter recommended that we pursue the regulatory and/or statutory approaches necessary to make data reporting and analysis possible include the Medicare Advantage population, and that this was essential so that functional outcomes of all Medicare beneficiaries (Part A or Medicare Advantage) reported by these proposed measures would more accurately represent the quality of care provided by a SNF. Two commenters commented that the description of the proposed measures should specify that the measure estimates outcomes for the Medicare Part A coverage benefit, as opposed to the admission and discharge from a nursing home. The commenter noted this was important because a Medicare Part A resident may remain in the nursing facility at the end of the Part A coverage period, so while the resident may be “discharged” from Part A benefits, he/she is not “discharged” from the nursing home.

Response: The commenter is correct that the functional outcome measures apply only to Medicare Part A SNF residents. The assessment data for the functional outcome measures would be collected at the start of the SNF Part A stay and the end of the Part A stay. We appreciate the suggestion to expand the proposed measure collection to a Medicare Advantage population. We will take the recommendation to expand the measure population into consideration in future measure development efforts. Additional discussion of the expansion of quality measures to include all residents regardless of payer status can be found in section III.D.2.k.5

Comment: One commenter noted there are meaningful SES, clinical, or other differences between traditional Medicare versus Medicare Advantage (MA) enrollees that could affect comparisons between facilities with different proportion of Medicare Advantage and Part A stays. The commenter further requested that this possibility should be investigated.

Response: For a discussion of social risk factors in the SNF QRP, please see the discussion in section III.D.2.b.1 of this rule.

Comment: One commenter stated that the calculation of the four proposed measures is complex, particularly with respect to the calculation of the expected discharge functional status score using a formula, which may result in providers not understanding the precise target outcome. The commenter further noted that the measure scores might be inappropriately compared across PAC settings even though they are calculated differently using different risk adjustor coefficients. The commenter stated that significant education and ongoing feedback for providers will be necessary when these measures are implemented to improve quality of care and suggested that we simplify the calculations for the functional outcome measures.

Another commenter voiced concern that the calculated “Expected score” for the function outcome measures would be an inaccurate point of comparison if the risk adjustors were not accurate. The commenter suggested that we fully evaluate the risk adjustors in a large data sample to ensure they are appropriate prior to implementation. The commenter also suggested that we should have a transparent process that is clearly communicated with stakeholders to clarify and refine risk adjustors for the functional outcome measures. The commenter noted that if there is not a refinement period of the risk adjustors, providers will be penalized for their performance on these measures at the same time that we are examining the risk adjustors' accuracy and possibly modifying them.

Response: We continuously examine the performance of quality measures and revise measures, including risk adjustment, to optimize measurement of quality ensuring that our measures and their components are accurate. We also continue to seek stakeholder input as we conduct our internal measure maintenance work. Further, we agree that education is important and necessary to help SNFs, as well as other PAC settings, understand how the four proposed functional outcome measures will be calculated. To that end, we intend to provide training materials through the CMS webinars, open door forums, and help desk support. The expected scores are calculated using the results of our risk-adjustment models. During our May 2016 TEP, we discussed the risk adjustment models extensively, and these discussions included a review of our analyses of the mean admission, discharge and change for the self-care and mobility scores for each risk adjustor. We also reviewed the risk adjustors for competing measures. These discussions are summarized in the SNF Function Summary TEP report, which is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical- Information.html . We believe the risk adjustment model is methodologically strong.

Comment: One commenter generally supported the proposed risk adjustment approach for the four proposed functional outcome measures, but requested additional items to address social risk, such as Medicare-Medicaid status. The commenter recommended testing of the risk adjustment methodology to ensure it adjusts for meaningful differences. Another commenter suggested that we risk adjust the four proposed functional outcome measures for social and environmental factors, such as social support and an accessible home environment. The commenter stated that by not adjusting for social and environmental risk factors, we might be creating conflicting incentives between functional improvement and resource use measures. Another commenter supported the use of other assessment data, such as mode of communication and gateway processes. One commenter expressed support for the proposed risk adjustors for the functional outcome measures, but recommended that we reassess all risk adjusters once the new MDS data are submitted.

Response: We selected the risk factors based on literature review, clinical relevance, TEP input, and empirical findings from the PAC-PRD analyses. For a discussion of social risk factors in the SNF QRP, we refer the commenter to section III.D.2.b.1. of this rule. We agree with the importance of testing and continuously monitoring the risk adjustment models so that the functional outcome quality measures reflect true differences in the effectiveness of treatments provided by SNFs. We will continue to examine the performance of our quality measures and revise risk adjustment approaches as necessary to optimize quality measurement.

Comment: Several commenters supported the use of selected risk adjustors and specifically noted that they support risk adjustors in the areas of age, admission function score, medical conditions, and impairments. One commenter stated that the proposed list of comorbidities used for risk adjustment of the functional outcome measures appears comprehensive but requested further detail of the source of the comorbidities data and the proposed look-back period for including the comorbidities. One commenter supported the inclusion of prior functioning and prior device use items for risk adjustment in the functional outcome measures but was concerned that the collection of this data will add administrative burden. Some commenters noted that coding for addition risk adjusters might cause additional provider burden. One commenter supported the inclusion of new data elements for risk adjustment, specifically the prior functioning, prior device use, primary medical condition category and prior surgery items, but under the condition that we appropriately account for the additional reporting burden within the SNF PPS rates. Another commenter expressed concern about the accuracy and burden of collecting the items that refer to a time period outside the defined period of the SNF stay. One commenter stated that SNFs would not know what determines the model estimate, and proposed that we provide the benchmark for comparison prior to the fiscal year. In addition, this commenter questioned the use of a statistical model since section GG includes the establishment of goals, arguing outcomes could be compared to the SNF's own established goals. Other commenters requested that we use the median discharge scores instead of the mean values as a way to avoid the impact of outliers on the expected score. Another commenter expressed that poor risk adjustment would penalize SNFs that provide care to medically-complex and socioeconomically disadvantaged residents, and threaten access to care.

Response: We agree with commenters on the importance of risk adjustment as functional outcomes can vary based on residents' demographic and admission clinical status. Risk adjustment allows for the comparison of functional outcomes across SNFs. As with other risk adjustors, both prior functioning and prior device use were identified as important risk adjustors for the functional outcome measures through data analyses. In development of the quality measures, we selected risk-adjustors including comorbidities, and other health and prior functioning items, based on evidence in the literature, stakeholder comments during TEPs, public comment opportunities statistical findings, and input from subject matter experts. As we develop and refine quality measures, we review existing items, listen to feedback from providers, and consider the appropriateness of adding or deleting any items to the MDS. Reduction of burden is an important consideration as we develop and refine quality measures, which includes risk adjustors for outcome measures. We would like to emphasize the importance of risk adjustment as functional outcomes can vary based on residents' demographic and clinical factors. Prior functioning is an important predictor of functional improvement and this is data routinely collected by therapists when developing a resident's care plan.

We agree with the commenter that it is important for risk adjustment of quality measures to be reliable and valid. As mentioned previously, the risk adjustors were determined based on data analysis, stakeholder input, literature review, clinical relevance and public comment. As noted above, we agree with the commenter for the need to re-examine the risk adjustment model when additional data become available. In addition, we appreciate the continued involvement of stakeholders in all phases of measure development and implementation.

We refer the commenter to the Specifications for SNF QRP Quality Measures and Standardized Resident Assessment Data Elements document available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html for additional details about the risk adjustment approach.

With regard to the use of the discharge goals, we would like to note that the quality measure, Application of Percent of Long-Term Care Hospital Patients with an Admission and Discharge Functional Assessment and a Care Plan that Addresses Function (NQF #2631), requires documentation of only one goal. Using goals to determine outcomes would require SNFs to complete all goals in section GG, which would add significant burden. With regard to the suggestion of using the median rather than the mean value, we will examine this approach as we examine additional data to determine how it affects quality measure scores.

We would like to note that the risk adjustment model for these outcomes includes up to 60 risk-adjusters, and includes more clinically and statistically relevant adjusters for function than other risk-adjusted functional outcomes measures. We will pursue ongoing monitoring and analysis of these proposed functional outcome measures to identify any potential disparities across patient and facility characteristics.

Comment: One commenter was concerned that the PAC PRD data from 34 nursing facilities and other providers used to develop the risk adjustors for the functional outcome measures for SNFs were inadequate. The commenter felt that a larger volume of data is necessary to verify the current risk adjustors. The commenter recommended that we reevaluate these risk adjustors on a regular basis to ensure their accuracy and to ensure that SNF providers are not evaluated and penalized in the future based on inadequate risk adjustment. The commenter also stated that suggestions offered during a Technical Expert Panel should be tested with data before becoming part of the quality measure and payment system.

Response: As previously discussed, the risk adjustors were selected based on literature review, clinical relevance, Technical Expert Panel input, public comment opportunities, and empirical findings from the data analyses from 60 SNFs and approximately 4,000 resident assessments. Based on our comprehensive approach to developing the models and the alignment between these models and the IRF models, we believe that our models are adequate for risk adjustment for the four SNF functional outcome measures. As part of measure maintenance and evaluation, we routinely analyze data to monitor the performance of implemented quality measures, including risk adjustment models, and thus we agree with the commenter that we should re-examine the risk adjustment model when national data become available. We aim to develop accurate and fair measures and we continuously examine the performance of quality measures and revise measures, including risk adjustment, to optimize measurement of quality.

Comment: Some commenters requested that additional risk adjusters be included in the proposed outcome measures' statistical models, and that each model includes a similar set of risk adjusters. One commenter requested that cognition and age be included in the model, while other commenters were concerned that “prior functioning: functional cognition”, “fall history”, and “prior functioning: mobility” were not included in the self-care model. Another commenter disagreed with the specification “independent” as the reference category since it appeared this also included residents with an unknown prior functional status. The commenter explained that in PAC settings, it is more likely that a patient who cannot report their prior functional status was more dependent rather than more independent before being admitted, so should not be grouped into the “independent” reference category.

Response: The majority of risk adjusters are the same in both the self-care and mobility functional outcome models. With regard to the variables included in the mobility models, but not included in the self-care models, these variables were all tested in the self-care model, but they were not statistically significant predictors of the change in self-care scores or the discharge self-care scores. As noted above, we will continue to examine the risk adjustment models when more data become available. We would also like to clarify that cognition and age are included in risk adjustment models and that the Brief Interview for Mental Status (BIMS) specifically accounts for functional variation associated with cognition status. Regarding the reference group “independent” for the prior functional status risk adjustors, we appreciate the commenter's suggestion and will take it into consideration.

Comment: Several commenters requested additional information regarding coding of some of the risk adjustment variables. One commenter requested additional detail about how a SNF would identify the appropriate primary medical condition category for the proposed new MDS item I0020, which is used for risk adjustment of the functional outcome measures. The commenter stated that the current approach of requiring the provider to identify one of the 13 primary medical diagnoses or list an ICD-10 code is burdensome and suggested rather a provider should enter the applicable ICD-10 code onto the MDS, which would then be mapped by the MDS grouper software to identify the applicable condition. The commenter further stated that the admitting diagnosis for admission to a SNF may not be directly relevant to the diagnosis associated with mobility and self-care treatment plans and goals, unlike with IRFs, and recommended that we revise this section of the MDS to request providers report the primary medical condition associated with mobility and self-care treatment. Another commenter requested more clarification on the use ICD-10 codes in defining the primary medical condition category, and further noted concern that these codes are more prevalent in the IRF setting, compared to the SNF setting. This commenter expressed concern about where the diagnosis group information will come from and explained that ICD-10 coding is complete and requires multiple levels of consideration and clinical input. Another commenter requested information on how “medically complex” is defined. Other commenters requested further clarification on where information for items such as mechanical ventilation will be acquired, how “major surgery” is defined and how the interaction between primary diagnosis and SNF admission functional status is determined in risk adjustment.

Response: We appreciate the commenters' concerns regarding coding of the primary medical conditions as well as the coding of mechanical ventilation and major surgery for risk adjustment. As previously noted, we intend to provide guidance on these issues as part of our comprehensive training. Some of these variables were added to the IRF-PAI Version 1.4 when the functional outcome measures were adopted in the IRF QRP, and since these primary medical conditions will be aligned across the IRF and SNF settings, providers can get a preview of the coding guidance and definitions in the IRF PAI Training Manual on page J-5, which is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/IRF-Quality-Reporting/IRF-PAI-and-IRF-QRP-Manual.html . The RAI manual will also be updated with all timely and accurate information. With regard to the primary medical condition diagnosis, which are risk adjustors for the four functional outcome measures, the proposed MDS effective October 1, 2018 does include primary diagnosis as a data element.

Comment: One commenter noted that the use of the term “Primary rehabilitation diagnosis” does not recognize that not all patients are admitted for rehabilitation.

Response: We would like to clarify that the term “Primary rehabilitation diagnosis” is not used as part of the four proposed functional outcome measures.

Comment: One commenter supported the use of the BIMS for risk adjustment of the functional outcome measures, stating that learning and memory deficits can significantly impact the rehabilitation of residents with functional impairments. However, the commenter stated that the BIMS is designed as a resident interview and that the use of the BIMS alone as risk adjustment in the SNF setting would be problematic due to the high percentage of residents unable to complete the BIMS as a result of severe cognitive or physical impairments. The commenter stated that a SNF resident's inability to complete the BIMS is often associated with slower rates and lesser degrees of functional improvement than those residents that can complete the BIMS. This commenter requested clarification as to how we will address risk adjustment for these residents and suggested excluding SNF residents that cannot complete the BIMS items if they are not accounted for in the current risk adjustment model. The commenter also suggested development of standardized patient assessment data for clinician observation of cognitive function and mental status in the future to account for residents who are unable to complete the BIMS.

Response: We appreciate the commenter's feedback regarding the use of the BIMS in risk adjustment for the functional outcome measures. We would like to clarify that in the MDS 3.0, if a resident is unable to complete the BIMS, the provider is directed to administer the Staff Assessment for Mental Status (C0700-C1000), and the data from the staff assessment for mental status is used for cognitive status risk adjustment when the BIMS score is not available. With regard to the residents who are unable to be interviewed for the BIMS due to communication disorders, the BIMS can also be administered in writing. Further, we note that communication impairment is also a risk adjusters the self-care and mobility models. With regard to the residents who are unable to be interviewed for the BIMS due to communication disorders, we note that communication impairment is also a risk adjusters the self-care and mobility models.

Comment: MedPAC noted the importance of using a consistent definition for “at admission” to enable accurate comparisons across PAC providers. The commenter stated that we should require that the assessment be completed within 3 days of admission and stated that the Day-5 assessment in SNFs is problematic since it can be conducted between Day 1 and Day 8.

Response: We appreciate the importance of data collection within consistent assessment time frames and we maintain a consistent approach to collecting information on or as close to the time of admission as possible. For example, on the 5-day assessment in SNF, the assessment time frame for the section GG Self-Care and Mobility data items on the MDS is 3 calendar days at the time of admission (first 3 calendar days) and discharge (day of discharge and the 2 days prior to the day of discharge). Therefore, across all PAC assessment instruments, we are collecting on a patient's usual performance within that three-day time period. That is, the 3-day assessment time frame for the section GG Self-Care and Mobility data elements is standardized across the three institutional PAC settings, SNFs, IRFs and LTCHs.

Comment: Two commenters requested that we ensure that the four quality measures are consistently reviewed for reliability, accuracy, and applicability to patients in different PAC settings to develop standards to compare quality across PAC settings. The commenters requested that we consider whether variation in training and practices among providers in various PAC settings affects data entry processes for the MDS and other PAC instruments, and whether this undermines the comparability of the proposed functional outcome measures. Another commenter requested that we provide clear language that cross-setting applications are not valid at this time due to differences in patient populations, payment policy, and specific measure calculation details. One commenter voiced concern that additional time, testing, and training may be necessary to ensure measures are implemented consistently across different settings that use very different processes, scales, definitions, and time frames, to allow data to be comparable across settings.

One commenter requested that we use the same set of definitions for standardized and interoperable functional assessment data in each PAC setting. The commenter further stated that this would mitigate providers collecting and calculating data for these measures differently across settings. The commenter was concerned discrepancies could result in unintended consequences with regard to payment and public reporting.

Response: We agree with the commenters that the accurate collection of functional assessment data is important across all PAC settings. Providers are required to submit accurate data to us, and we provide training and other resources. Providers should collect data in a manner that fits with the clinical workflow within their facility. With regard to the concern that reporting variability may impact comparability across facilities, we agree that comprehensive training is needed to ensure accuracy of data collection and interpretation as well as successful implementation of new measures. As with previous measures, we will provide training sessions, training manuals, Webinars, open door forums, help desk support, and a Web site that hosts training information ( http://www.youtube.com/user/CMSHHSgov ). At this time, we are adopting these measures into the SNF QRP, which is a pay-for-reporting program, and have not specified a timeframe for public reporting of these measures for SNFs.

With regard to the request for standardized and interoperable functional assessment data in each PAC setting, we agree with the commenter about the importance of accurate collection of standardized patient assessment data across the PAC settings. The item definitions are the same across PAC settings, and we continue to work to harmonize the coding guidance for the standardized assessment data elements as we believe that this is key to the collection of accurate data.

Comment: One commenter supported our proposal to collect data on the proposed function quality measures through the MDS using the QIES ASAP system.

Response: We appreciate the commenter's support.

Final Decision: After careful consideration of the public comments received, we are finalizing our proposal to adopt the four functional outcome measures, Application of IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633), Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634), the Application of IRF Functional Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (NQF #2635), the Application of IRF Functional Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (NQF #2636), beginning with the FY 2020 SNF QRP.

h. Modifications to Potentially Preventable 30-Day Post-Discharge Readmission Measure for Skilled Nursing Facility (SNF) Quality Reporting Program (QRP)

In the FY 2017 SNF PPS final rule (81 FR 52030 through 52034), we adopted the Potentially Preventable 30-Day Post-Discharge Readmission Measure for SNF QRP. This measure was developed to meet section 1899B(d)(1)(C) of the Act, which calls for measures to reflect all-condition risk-adjusted potentially preventable hospital readmission rates for PAC providers, including SNFs.

This measure was specified to be calculated using 1 year of Medicare FFS claims data; however, in the FY 2018 SNF PPS proposed rule (82 FR 21057) we proposed to increase the measurement period to 2 years of claims data. The rationale for this change is to expand the number of SNFs with 25 stays or more, which is the minimum number of stays that we require for public reporting. Furthermore, this modification will align the SNF measure more closely with other potentially preventable hospital readmission measures developed to meet the IMPACT Act requirements and adopted for the IRF and LTCH QRPs, which are calculated using 2 consecutive years of data.

We also proposed to update the dates associated with public reporting of SNF performance on this measure. In the FY 2017 SNF PPS final rule (81 FR 52030 through 52034), we finalized initial confidential feedback reports by October 2017 for this measure based on 1 calendar year of claims data from discharges during CY 2016 and public reporting by October 2018 based on data from CY 2017. However, to make these measure data publicly available by October 2018, we proposed to shift this measure from calendar year to fiscal year, beginning with publicly reporting on claims data for discharges in fiscal years 2016 and 2017.

Additional information regarding the Potentially Preventable 30-Day Post-Discharge Readmission Measure for SNF QRP can be found at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html .

We sought public comment on our proposal to increase the length of the measurement period and to update the public reporting dates for this measure. A discussion of these comments, along with our responses, appears below.

Comment: We received several comments on our proposal to expand the data reporting period for SNFs from one year to 2 years for the Potentially Preventable 30-Day Post-Discharge Readmission Measure for SNF QRP. MedPAC and other commenters supported this proposal because it would increase the number of SNFs included in public reporting. Other commenters expressed support for aligning the SNF measure with the potentially preventable hospital readmission measures we have adopted for the IRF and LTCH QRPs, which also use 2 years of data.

Some commenters were concerned that the greater lag associated with expanding the reporting period to 2 years would make the measure less valuable or sensitive to quality improvement. One commenter was concerned that publicly reporting performance data based on 2 years of data may not accurately reflect the quality of care that SNFs are currently furnishing. Some commenters were opposed to the proposal because it would not align with measurement periods used in other SNF quality measures. One commenter was specifically opposed to shifting this measure to a fiscal year cycle because most SNF data are based on calendar years, noting that inconsistent time periods may create confusion. Another commenter did not oppose the shift to fiscal year as long as confidential feedback reports and review and correction timelines would not be negatively impacted.

Response: We appreciate commenters' concerns that increasing the measurement period from one year to 2 years would create a greater delay between data collection and public reporting of this measure. However, we agree with those commenters that noted the benefit of increasing the number of SNFs for public reporting purposes outweighs the concerns associated with the data delays. We also agree with commenters that this change would better align the SNF measure with the other PPR measures developed to meet the requirements of the IMPACT Act. We also note that changing the public reporting dates for this measure from calendar to fiscal year will not impact providers' confidential feedback reports or the length of time they have to review and correct the data to be made publicly available.

Final Decision: After careful consideration of the public comments, we are finalizing our proposal to increase the measurement period from 1 year to 2 years for the calculation of the Potentially Preventable 30-day Post-Post Discharge Readmission Measure for SNF QRP measure. We are also finalizing our proposal to shift from calendar to fiscal years for public reporting of this measure.

i. SNF QRP Quality Measures Under Consideration for Future Years

In the FY 2018 SNF PPS proposed rule (82 FR 21058), we invited public comment on the importance, relevance, appropriateness, and applicability of each of the quality measures listed in Table 19 for future years in the SNF QRP.

We are also considering a measure focused on pain that relies on the collection of patient-reported pain data, and another measure regarding the Percent of Residents Who Were Assessed and Appropriately Given the Seasonal Influenza Vaccine. Finally, we are considering a measure related to patient safety, that is, Patients Who Received an Antipsychotic Medication.

Commenters submitted the following comments related to the proposed rule's discussion of the SNF QRP Quality Measures Under Consideration for Future Years. A discussion of these comments, along with our responses, appears below.

Comment: One commenter supporting the future measure concept of the percent of residents who self-report moderate to severe pain, suggested inclusion of this measure by FY 2019 at the latest. Another commenter suggested that they do not believe that pain experience alone should be a quality measure, expressing that the presence of pain does not provide enough information to help an individual's overall quality of life improve.

One commenter suggested that a measure be developed that reflects patient-centered care pain management regardless of ability to self-report as a significant portion of SNF residents are not able to self-report pain and suggested using reliable and valid observational assessment items such as those in the current MDS 3.0 Section J0800 and J0850. The commenter encouraged us to consider incorporating the standardized observational pain assessment data elements that are currently being developed and tested to fulfill the requirements of the IMPACT Act. The commenter also urged us to seek NQF endorsement for any new measures to be incorporated into the SNF QRP program. Another commenter encouraged assessment for communication about pain rather than experience of pain without inadvertently incentivizing the use of opioid medications in alignment with proposed changes to HCAHPS. Another commenter suggested modifying this measure to reflect the proportion of residents for which moderate to severe pain interferes with or prevents important daily functional tasks and drive improvements in quality of life.

Response: We appreciate the comments pertaining to the Application of Percent of Residents Who Self-Report Moderate to Severe Pain (Short Stay) (NQF #0676) measure under consideration for future implementation in the SNF QRP. We note that appropriately assessing pain as an outcome is important, acknowledge the importance of avoiding unintended consequences that may arise from such assessments, and will take into consideration the commenters' recommendations. We would like to note that our goal is to submit all fully developed measures to NQF for consideration of endorsement.

Comment: We received several comments supporting the development of a seasonal influenza vaccine measure appropriate for the SNF population. One commenter stated that the incidence and impact of influenza disease is severe within the population of older adults in a SNF setting, and stated that as a result, there is a need for this measure. One commenter further suggested that a measure of this type presents an important opportunity to promote higher quality and more efficient health care for Medicare beneficiaries. One commenter recommended that we give due consideration to the cost of these services when the costs (for example, the purchase of the vaccine) of these services are bundled into the SNF Part A payment rates. This commenter supported alignment with ongoing efforts to collect and report this measure in the Long-Term Care Hospital Quality Reporting Program (LTCH QRP). Further, this commenter suggested CMS may want to add a pneumococcal vaccine measure in addition to an influenza measure.

Response: We acknowledge the commenters' support of inclusion of a seasonal influenza vaccine measure. We will take all recommendations into consideration in our ongoing efforts to identify and propose appropriate measures for the SNF QRP.

Comment: We received general support for development of an antipsychotic medication measure appropriate for the SNF population. One commenter expressed support for this measure concept and suggested inclusion of the measure by FY 2019 at the latest. One commenter expressed support for including most individuals in the measure regardless of dementia diagnoses. However, this commenter further suggested that Food and Drug Administration (FDA) approved indications of the medications should be excluded from this measure. Another commenter suggested further development of the measure as there is no existing baseline measurement. Another commenter suggested that any future measure should account for informed choices by persons with behavioral and psychotic symptoms of dementia (BPSD) and their families regarding the use of antipsychotic medications for appropriately-used antipsychotics, even if the medication does not have an indication approved by the FDA for their symptoms.

Response: We acknowledge the support of inclusion of an antipsychotic measure and note the suggestion pertaining to the exclusions as well as the measure accounting for persons with BPSD. Recommendations will be taken into consideration in our ongoing efforts to identify and propose appropriate measures for the SNF QRP in the future.

Comment: MedPAC suggested that we consider the adoption of future measures that can assess providers' ability to maintain function and prevent functional decline. MedPAC noted that the two quality measures for change in function do not capture whether a provider can maintain function as residents with conditions who are not expected to improve or who are already independent are excluded from the four measures that we are finalizing.

Response: We agree with MedPAC that future quality measurement work should include the development of quality measures that focus on maintaining function and prevention of functional decline. We appreciate MedPAC's concern regarding the exclusion of residents who are not expected to improve due to certain medical conditions or who are independent. We would like to point out that two of the measures we are adopting in this final rule for the SNF QRP, Application of the IRF Function Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (NQF #2635) and Application of the IRF Function Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (NQF #2636), capture residents who are independent with function at admission. In that situation, maintenance of independence for the section GG self-care or mobility activities would apply to these residents.

Comment: One commenter recommended the addition of a quality measure of maintenance of functional status to the SNF QRP to address requirements of the Jimmo Settlement. The commenter noted that functional improvement is not a goal for all residents receiving rehabilitation; for some residents, maintaining or slowing functional decline is a goal.

Response: We appreciate the commenter's suggestions, and we will consider this recommendation in future measure development.

Comment: One commenter encouraged us to consider the importance of instrumental activities of daily living as a measurement construct for assessing patient need, monitoring quality, and affecting care and payment, stating that instrumental activities of daily living performance is critical to maintaining safety and avoiding readmissions.

Response: We appreciate the commenter's suggestions for future measures and we will consider this recommendation in future measure development.

Comment: MedPAC commented that while the proposed future measures capture important dimensions of SNF care, MedPAC prefers that Medicare hold providers accountable for claims-based outcome measures. Several commenters suggested further development and standardization of outcome measures to compare and contrast between PAC settings and to assess short- and long-term patient status post injury or illness. One commenter suggested moving away from an emphasis on process measures toward more outcome-related measures. Another commenter added that any additional vaccination measure give due consideration to the cost of these services. Others suggested measures related to consumer satisfaction following short stay rehabilitation and discharge home. One commenter suggested that any patient experience of care survey for SNFs be economical in its approach and carefully aligned with other surveys to reduce duplicative collection activities. Other commenters suggested a number of additional measures for inclusion in the SNF QRP. One commenter suggested that we consider developing measures to assess quality of life and long-term functional outcomes such as community-oriented factors including ability to live independently, return to work (where appropriate), community participation and social interaction. Another commenter suggested workforce related measures such as staffing quality metrics from payroll-based journal staffing and collection such as staff turnover, nursing staff hours per resident stay and CNA hours per resident stay. The commenter further recommended measures that include language related to initiating palliative care and making ethical considerations regarding continuing or terminating complex medical care. The commenter also suggested incorporating coordination and collaboration on patient, family, and medical goals of care as well as assessment of family members' and caregivers' capacity to assume patient care post-discharge. Another commenter further recommended that measures such as those currently reported on Nursing Home Compare be used in the interim until more post-acute care cross-setting measures are developed.

Response: We appreciate the input from MedPAC and other commenters for their suggestions on future measure concepts as well as on the interim use of measures currently reported on Nursing Home Compare. With all measures, we seek to fulfill the mandate of the IMPACT Act to align across settings and will take these comments into consideration as we further develop measures for use in the SNF QRP.

(1) IMPACT Act Measure—Possible Future Update to Measure Specifications

In the FY 2017 SNF PPS final rule (81 FR 52021 through 52029), we finalized the Discharge to Community-Post Acute Care (PAC) Skilled Nursing Facility (SNF) Quality Reporting Program (QRP) measure, which assesses successful discharge to the community from a SNF setting, with successful discharge to the community including no unplanned rehospitalizations and no death in the 31 days following discharge from the SNF. We received public comments (see 81 FR 52025 through 52026) recommending exclusion of baseline nursing facility residents from the measure, as these residents did not live in the community prior to their SNF stay. At that time, we highlighted that using Medicare FFS claims alone, we were unable to accurately identify baseline nursing facility residents. We stated that potential future modifications of the measure could include assessment of the feasibility and impact of excluding baseline nursing facility residents from the measure through the addition of patient assessment-based data. In response to these public comments, we are considering a future modification of the Discharge to Community-PAC SNF QRP measure, which would exclude baseline nursing facility residents from the measure. Further, this measure is specified to be calculated using one year of Medicare FFS claims data. We are considering expanding the measurement period in the future to two consecutive years of data to increase SNF sample sizes and reduce the number of SNFs with fewer than 25 stays that would otherwise be excluded from public reporting. This modification would also align the measurement period with that of the discharge to community measures adopted for the IRF and LTCH Quality Reporting Programs to meet the IMPACT Act requirements; both the IRF and LTCH measures have measurement periods of two consecutive years.

We sought public comment on these considerations for Discharge to Community-PAC SNF QRP measure in future years of the SNF QRP. A discussion of these comments, along with our responses, appears below.

Comment: Multiple commenters expressed support for excluding baseline nursing facility residents from the discharge to community measure as a potential future measure modification. Commenters stated that this exclusion would result in the measure more accurately portraying quality of care provided by SNFs, while controlling for factors outside of SNF control.

Response: We acknowledge the commenters' support for the potential exclusion of baseline nursing facility residents as a future measure modification. We will consider their views and determine whether to propose to exclude baseline nursing facility residents from the Discharge to Community-PAC SNF QRP measure in future years of the SNF QRP.

Comment: MedPAC supported expanding the Discharge to Community-PAC SNF QRP measurement period from 1 year to 2 years, acknowledging that it is important to include as many providers in public reporting as possible and that expansion to 2 years is a good strategy to help include more low-volume providers in public reporting. A few commenters opposed expansion of the measurement period to 2 years, expressing concern that it decreased the timeliness of the data and actionability for providers to drive change in quality or process improvement. One commenter expressed concern that the expansion would misalign the measurement period with that of other SNF measures in use, and that inclusion of older data would decrease sensitivity to change in quality, particularly for high volume SNFs. This commenter stated that a 2-year window would not accurately reflect recent improvement or decline in discharge planning practices, resulting in inaccurate portrayal of the current quality of care furnished by a SNF. Another commenter expressed concern that a two-year measurement period penalized facilities with adverse ratings for longer periods of time.

Response: We acknowledge MedPAC for its support for possible expansion of the Discharge to Community-PAC SNF QRP measurement period to 2 years in future years of the SNF QRP. We would like to clarify that we did not propose this change, but are considering it for future years. We also acknowledge commenters' concerns about expanding the measurement period to 2 years. We will consider these views and determine whether to propose expanding the Discharge to Community-PAC SNF QRP measurement period from 1 year to 2 years in future years of the SNF QRP.

(2) IMPACT Act Implementation Update

As a result of the input and suggestions provided by technical experts at the TEPs held by our measure developer, and through public comment, we are engaging in additional development work for two measures that would satisfy the domain of accurately communicating the existence of and providing for the transfer of health information and care preferences when the individual transitions, in section 1899B(c)(1)(E) of the Act, including performing additional testing. The measures under development are: Transfer of Information at Post-Acute Care Admission, Start or Resumption of Care from other Providers/Settings; and Transfer of Information at Post-Acute Care Discharge, and End of Care to other Providers/Settings. We intend to specify these measures under section 1899B(c)(1)(E) of the Act no later than October 1, 2018 and we intend to propose to adopt them for the FY 2021 SNF QRP, with data collection beginning on or about October 1, 2019.

Commenters submitted the following comments related to the proposed rule's discussion of the IMPACT Act Implementation Update. A discussion of these comments, along with our responses, appears below.

Comment: One commenter suggested that we be cautious in our development of the Transfer of Health Information measure set and only proceed to propose and adopt measures that receive NQF endorsement. This commenter cited concerns about the measure development, citing the 2016 MAP PAC/LTC meeting. A commenter supported our efforts to promote coordination of care across the care continuum, and commented that the transfer of accurate health information—including resident preferences, care plan, and other information—is essential to quality outcomes for residents. A commenter expressed appreciation that we are developing measures that will help facilitate the accurate communication of a person's health information and care preferences across the continuum of care and believes that these measures will facilitate better care coordination and outcomes. The commenter also appreciated that we have engaged providers and consumers in the development of these measures and encourages us to develop measures that represent a balance between the volume and detail of information exchanged and reported, and the underlying administrative burdens the measures may create. The commenter noted that the burden is particularly important for small and rural providers that may have more challenges with technology-driven information exchange because health information technology incentive programs for hospitals and physicians have not been extended to SNF providers.

Response: We appreciate the comments and feedback on the Transfer of Health Information measures that are currently under development. We also appreciate the recognition that we have engaged providers and consumers in the development of these measures. As we continue to develop these measures, we will consider this feedback. We would like to clarify that the measure under development does not currently require the adoption of health IT and electronic means of information transfer. We intend to re-submit these measures, once fully specified and tested, for review to the MAP PAC/LTC Workgroup. Further, we plan to submit the measures to the NQF for consideration for endorsement when we believe the measures are ready for NQF review.

j. Standardized Resident Assessment Data Reporting for the SNF QRP

(1) Standardized Resident Assessment Data Reporting for the FY 2019 SNF QRP

Section 1888(e)(6)(B)(i)(III) of the Act requires that for fiscal year 2019 and each subsequent year, SNFs report standardized resident assessment data required under section 1899B(b)(1) of the Act. As we describe in section III.D.2.g.(1) above, we are finalizing in this final rule that the current pressure ulcer measure, Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (Short Stay) (NQF #0678), will be replaced with the proposed pressure ulcer measure, Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury, beginning with the FY 2020 SNF QRP. The current pressure ulcer measure will remain in the SNF QRP until that time. Accordingly, for the requirement that SNFs report standardized resident assessment data for the FY 2019 SNF QRP, we proposed that the data elements used to calculate that measure meet the definition of standardized resident assessment data for medical conditions and co-morbidities under section 1899B(b)(1)(B)(iv) and that the successful reporting of that data under section 1888(e)(6)(B)(i)(II) for admissions as well as discharges occurring during fourth quarter CY 2017 would also satisfy the requirement to report standardized resident assessment data for the FY 2019 SNF QRP.

The collection of assessment data pertaining to skin integrity, specifically pressure related wounds, is important for multiple reasons. Clinical decision support, care planning, and quality improvement all depend on reliable assessment data collection. Pressure related wounds represent poor outcomes, are a serious medical condition that can result in death and disability, are debilitating, painful and are often an avoidable outcome of medical care. Pressure related wounds are considered health care acquired conditions.

As we note above, the data elements needed to calculate the current pressure ulcer measure are already included on the MDS and reported for SNFs, and exhibit validity and reliability for use across PAC providers. Item reliability for these data elements was also tested for the nursing home setting during implementation of MDS 3.0. Testing results are from the RAND Development and Validation of MDS 3.0 project. The RAND pilot test of the MDS 3.0 data elements showed good reliability and is also applicable to both the IRF-PAI and the LTCH CARE Data Set because the data elements tested are the same. Across the pressure ulcer data elements, the average gold-standard nurse to gold-standard nurse kappa statistic was 0.905. The average gold-standard nurse to facility-nurse kappa statistic was 0.937. Data elements used to risk adjust this quality measure were also tested under this same pilot test, and the gold-standard to gold-standard kappa statistic, or percent agreement (where kappa statistic not available), ranged from 0.91 to 0.99 for these data elements. These kappa scores indicate “almost perfect” agreement using the Landis and Koch standard for strength of agreement.

The data elements used to calculate the current pressure ulcer measure received public comment on several occasions, including when that measure was proposed in the FY 2012 IRF PPS (76 FR 47876) and IPPS/LTCH PPS proposed rules (76 FR 51754). Further, they were discussed in the past by TEPs held by our measure development contractor on June 13 and November 15, 2013, and recently by a TEP on July 18, 2016. TEP members supported the measure and its cross-setting use in PAC. The report, Technical Expert Panel Summary Report: Refinement of the Percent of Patients or Residents with Pressure Ulcers that are New or Worsened (Short-Stay) (NQF #0678) Quality Measure for Skilled Nursing Facilities (SNFs), Inpatient Rehabilitation Facilities (IRFs), Long-Term Care Hospitals (LTCHs), and Home Health Agencies (HHAs), is available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/IMPACT-Act-of-2014/IMPACT-Act-Downloads-and-Videos.html.

We sought public comment on this proposal. A discussion of these comments, along with our responses, appears below.

Comment: We received many comments in support of reporting the data elements already implemented in the SNF QRP to fulfill the requirement to report standardized resident assessment data for the FY 2019 SNF QRP. Specifically, many commenters supported the use of data elements used in calculation of the Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (Short Stay) (NQF #0678) to fulfill this requirement.

Response: We appreciate the commenter's support of the proposal.

Final Decision: After consideration of the public comments received, we are finalizing the proposal that the data elements currently reported by SNFs to calculate the current measure, Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (Short Stay) (NQF #0678), meet the definition of standardized resident assessment data for medical conditions and co-morbidities under section 1899B(b)(1)(B)(iv) of the Act, and that the successful reporting of that data under section 1888(e)(6)(B)(i)(II) of the Act would also satisfy the requirement to report standardized resident assessment data under section 1888(e)(6)(B)(i)(III) of the Act.

(2) Standardized Resident Assessment Data Reporting Beginning With the FY 2020 SNF QRP

In the FY 2018 SNF PPS proposed rule (82 FR 21059 through 21076), we described our proposals for the reporting of standardized resident assessment data by SNFs beginning with the FY 2020 SNF QRP. SNFs would be required to report these data for SNF admissions at the start of the Medicare Part A stay and SNF discharges at the end of the Medicare Part A stay that occur between October 1, 2018 and December 31, 2018, with the exception of two data elements (Hearing and Vision), which would be required for SNF admissions at the start of the Medicare Part A stay only that occur between October 1, 2018, and December 31, 2018. Following the initial reporting year for the FY 2020 SNF QRP, subsequent years for the SNF QRP would be based on a full calendar year of such data reporting.

In selecting the data elements, we carefully weighed the balance of burden in assessment-based data collection and aimed to minimize additional burden through the utilization of existing data in the assessment instruments. We also note that the resident assessment instruments are considered part of the medical record, and sought the inclusion of data elements relevant to resident care. We also took into consideration the following factors for each data element: overall clinical relevance; ability to support clinical decisions, care planning and interoperable exchange to facilitate care coordination during transitions in care; and the ability to capture medical complexity and risk factors that can inform both payment and quality. Additionally, the data elements had to have strong scientific reliability and validity; be meaningful enough to inform longitudinal analysis by providers; had to have received general consensus agreement for its usability; and had to have the ability to collect such data once but support multiple uses. Further, to inform the final set of data elements for proposal, we took into account technical and clinical subject matter expert review, public comment and consensus input in which such principles were applied. We also took into account the consensus work and empirical findings from the PAC PRD. We acknowledge that during the development process that led to these proposals, some providers expressed concern that changes to the MDS to accommodate standardized resident assessment data reporting would lead to an overall increased reporting burden. However, we note that there is no additional data collection burden for standardized data already collected and submitted on the quality measures.

Comment: Many commenters expressed significant concerns with respect to our standardized resident assessment data proposals. Several commenters stated that the new standardized resident assessment data reporting requirements will impose significant burden on providers, given the volume of new standardized resident assessment data elements that were proposed to be added to the MDS. Several commenters noted that the addition of the proposed standardized resident assessment data elements would require hiring more staff, retraining staff on revised questions or coding guidance, and reconfiguring internal databases and EHRs. Other commenters expressed concerns about the gradual but significant past and future expansion of the MDS through the addition of standardized resident assessment data elements and quality measures, noting the challenge of coping with ongoing additions and changes.

Several commenters expressed concern related to the implementation timeline in the proposed rule, which would require SNFs to begin collecting the proposed standardized resident assessment data elements in the timeframe stated in the proposed rule. A few commenters noted that CMS had not yet provided sufficient specifications or educational materials to support implementation of the new resident assessments in the proposed timeline.

Several commenters urged CMS to delay the reporting of new standardized resident assessment data elements by at least one year, and to carefully assess whether all of the proposed standardized resident assessment data elements are necessary under the IMPACT Act. Commenters suggested ways to delay the proposals for standardized resident assessment data elements in the categories of Cognitive Function and Mental Status; Special Services, Treatments, and Interventions; and Impairments, including allowing voluntary or limited reporting for a period of time before making comprehensive reporting mandatory, and delaying the beginning of mandatory data collection for a period of time. Some commenters recommended that during the delay, CMS re-evaluate whether it can require the reporting of standardized resident assessment data in a less burdensome manner.

Response: We understand the concerns raised by commenters that the finalization of our standardized resident assessment data proposals would require SNFs to spend a significant amount of resources preparing to report the data, including updating relevant protocols and systems and training appropriate staff. We also recognize that we can meet our obligation to require the reporting of standardized resident assessment data with respect to the categories described in section 1899B(b)(1)(B) of the Act while simultaneously being responsive to these concerns. Therefore, after consideration of the public comments we received on these issues, we have decided that at this time, we will not finalize the standardized resident assessment data elements we proposed for three of the five categories under section 1899B(b)(1)(B) of the Act: Cognitive Function and Mental Status; Special Services, Treatments, and Interventions; and Impairments. Although we believe that the proposed standardized resident assessment data elements would promote transparency around quality of care and price as we continue to explore reforms to PAC payment system, the data elements that we proposed for each of these categories would have imposed a new reporting burden on SNFs. We agree that it would be useful to evaluate further how to best identify the standardized resident assessment data that would satisfy each of these categories; would be most appropriate for our intended purposes including payment and measure standardization; and can be reported by SNFs in the least burdensome manner. As part of this effort, we intend to conduct a national field test that allows for stakeholder feedback and to consider how to maximize the time SNFs have to prepare for the reporting of standardized resident assessment data in these categories. We intend to make new proposals with respect to the categories described in sections 1899B(b)(1)(B)(ii), (iii) and (v) of the Act no later than in the FY 2020 SNF PPS proposed rule.

In this final rule, we are finalizing the standardized resident assessment data elements that we proposed to adopt for the IMPACT Act categories of Functional Status and Medical Conditions and Co-Morbidities. Unlike the standardized resident assessment data that we are not finalizing, the standardized resident assessment data that we proposed for these categories are already required to calculate the Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (NQF #0678) quality measure, the Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury quality measure (which we are finalizing in this final rule), and the Application of Percent of Long-Term Care Hospital Patients with an Admission and Discharge Functional Assessment and a Care Plan That Addresses Function (NQF #2631) quality measure (which we finalized in the FY 2016 SNF PPS final rule). As a result, we do not believe that finalizing these proposals creates a new reporting burden for SNFs or otherwise necessitates a delay.

Comment: Several commenters expressed support for the adoption of standardized resident assessment data elements. A few commenters expressed support for standardizing the definitions as well as the implementation of the data collection effort. Several commenters also supported CMS' goal of standardizing the questions and responses across all PAC settings to help “enable the data to be interoperable, allowing it to be shared electronically, or otherwise between PAC provider types.” Another commenter noted full support of the IMPACT Act's goals and objectives and appreciated CMS' efforts to regularly communicate with stakeholders through various national provider calls, convening of stakeholders, and meetings with individual organizations.

Response: We appreciate the support of these proposals, but note that for the reasons explained above, we have decided at this time to not finalize the proposals for three of the five categories under section 1899B(b)(1)(B) of the Act: Cognitive Function and Mental Status; Special Services, Treatments, and Interventions; and Impairments.

Comment: Several commenters stated that there is insufficient evidence demonstrating the reliability and validity of the proposed standardized resident assessment data elements. Some commenters stated that the expanded standardized resident assessment data reporting requirements have not yet been adequately tested to ensure they collect accurate and useful data in this setting. A few commenters stated that six of the items that are currently reported in the MDS would be expanded to include additional sub-elements that SNFs would be required to complete. One of these commenters stated that CMS' conclusion that the collection of these standardized resident assessment data elements in the SNF setting would be feasible and the standardized resident assessment data elements would result in valid and reliable data was based on the current use of these data elements in the MDS and the testing of these data elements in the PAC PRD. One commenter stated that several of the proposed standardized resident assessment data elements that had not been adequately tested were deemed close enough to an item that had been tested in the PAC PRD or in other PAC settings and thus appropriate for implementation.

Response: Our standardized resident assessment data elements were selected based on a rigorous multi-stage process described in the FY 2018 SNF PPS proposed rule (82 FR 21044). In addition, we believe that the PAC PRD testing of many of these data elements provides good evidence from a large, national sample of patients and residents in PAC settings to support the use of these standardized patient/resident assessment data elements in and across PAC settings. However, as noted above, we have decided at this time to not finalize the proposals for three of the five categories under section 1899B(b)(1)(B) of the Act: Cognitive Function and Mental Status; Special Services, Treatments, and Interventions; and Impairments. Prior to making new proposals for these categories, we intend to conduct extensive testing to ensure that the standardized resident assessment data elements we select are reliable, valid and appropriate for their intended use.

Comment: MedPAC supported the addition of standardized resident assessment data elements, but cautioned that measures, when used for risk-adjustment, may be susceptible to inappropriate manipulation by providers. MedPAC believed that CMS may want to consider requiring a physician signature to attest that the reported service was reasonable and necessary and including a statement adjacent to the signature line warning that filing a false claim is subject to treble damages under the False Claims Act.

Response: We acknowledge MedPAC's feedback, and agree with the importance of data integrity within resident assessments. We will explore the suggestions made by MedPAC.

Comment: One commenter noted that the absence of a single source document that identifies the MDS data element, assessment type, allowable item responses, and item responses that could negatively impact SNF QRP performance scores and creates administrative challenges in keeping up to date with measure and item changes. This commenter urged us to provide a single resource for SNF providers to identify each individual MDS 3.0 data element identified by CMS and applicable to the various measures and standardized cross-setting data elements that apply to the SNF QRP. Another commenter urged us to provide detailed guidance and training documents that includes prescriptive coding, similar to what was done for the MDS. Another commenter stressed the importance of timely, appropriate education and training for providers to ensure that there is interoperability following full implementation. Another commenter also believed that standardized resident assessment data collected may be affected by educational level and professional expertise of the evaluator and advocated for fully developed risk-adjusters.

Response: We acknowledge the commenters' feedback with respect to administrative challenges and the desire for detailed guidance and training. In ongoing standardized resident assessment data element development work, we will continue to be mindful of the administrative challenges that new mandated assessment items will place on providers. We agree with the commenter about the importance of providing clear coding guidelines for the proposed standardized resident assessment data elements for a range of education levels. We are also committed to providing comprehensive training and guidance to providers, for any new data elements, including standardized resident assessment data elements, to ensure the fidelity of the assessment.

Comment: A few commenters sought clarification on interoperability requirements, if and how SNF providers will be required to demonstrate interoperability, and described potential challenges to interoperable data exchange, such as timeframes related to data submission (for example, 14 days after discharge for SNFs) and inconsistencies in how data are captured. One commenter encouraged CMS to consider interoperability standards that promote information exchange utilizing EHRs and to specify which data standards are to be used and how they are to be implemented to ensure consistency across providers. The same commenter recommended that CMS work with EHR vendors and other IT developers to implement changes and to consider the time required for implementing changes adopted in the final rule, which may require adopting timelines that are more extended than what was originally required. Further, two commenters urged CMS to develop methods to incentivize providers who are “stepping up” and adopting health information technology (HIT), despite the costs and the absence of a regulatory requirement to do so.

Response: We acknowledge commenters' concerns regarding standardization and interoperability of the proposed standardized resident assessment data elements to meet section 1899B(a)(1)(B) of the Act requirements. We wish to clarify that implementation of the proposed standardized resident assessment data elements is intended to facilitate interoperability. We acknowledge that the provision requires that we make certain resident assessment data standardized and interoperable to allow for the exchange of data among PAC settings and other providers in order to access longitudinal information which will facilitate coordinated care and improved outcomes. While the IMPACT Act requires that the post-acute resident assessment instruments be modified so that certain resident assessment data are standardized and interoperable, it does not require the exchange of electronic health information by such providers. We appreciate the comments surrounding the need for more time for providers to implement the changes necessary in response to such modifications, and have addressed this topic in our proposals within this section.

A full discussion of the standardized resident assessment data elements that we proposed to adopt for the categories described in sections 1899B(b)(1)(B)(ii), (iii) and (v) can be found in the FY 2018 SNF PPS proposed rule (82 FR 21060 through 21076). In light of our decision to not finalize our proposals with respect to these categories, we are not going to address in this final rule the specific technical comments that we received on these proposed data elements. However, we appreciate the many technical comments we did receive specific to each of these data elements, and we will take them into consideration as we develop new proposals for these categories. Below we discuss the comments we received specific to the standardized resident assessment data we proposed to adopt, and are finalizing in this final rule, for the categories of Functional Status and Medical Conditions and Co-Morbidities.

a. Standardized Resident Assessment Data by Category

(1) Functional Status Data

We proposed that the data elements currently reported by SNFs to calculate the measure, Application of Percent of Long-Term Care Hospital Patients with an Admission and Discharge Functional Assessment and a Care Plan That Addresses Function (NQF #2631), would also meet the definition of standardized resident assessment data for functional status under section 1899B(b)(1)(B)(i) of the Act, and that the successful reporting of that data under section 1886(m)(5)(F)(i) of the Act would also satisfy the requirement to report standardized resident assessment data under section 1886(m)(5)(F)(ii) of the Act.

These patient assessment data for functional status are from the CARE Item Set. The development of the CARE Item Set and a description and rationale for each item is described in a report entitled “The Development and Testing of the Continuity Assessment Record and Evaluation (CARE) Item Set: Final Report on the Development of the CARE Item Set: Volume 1 of 3.” Reliability and validity testing were conducted as part of CMS' Post-Acute Care Payment Reform Demonstration, and we concluded that the functional status items have acceptable reliability and validity. A description of the testing methodology and results are available in several reports, including the report entitled “The Development and Testing of the Continuity Assessment Record And Evaluation (CARE) Item Set: Final Report On Reliability Testing: Volume 2 of 3” and the report entitled “The Development and Testing of The Continuity Assessment Record And Evaluation (CARE) Item Set: Final Report on Care Item Set and Current Assessment Comparisons: Volume 3 of 3.” The reports are available on CMS' Post-Acute Care Quality Initiatives Web page at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Post-Acute-Care-Quality-Initiatives/CARE-Item-Set-and-B-CARE.html. For more information about this quality measure, we refer readers to the FY 2016 SNF PPS final rule (80 FR 46444 through 46453).

We sought public comment on this proposal. A discussion of these comments, along with our responses, appears below.

Comment: Several commenters supported the collection of standardized resident assessment data across PAC settings to satisfy the IMPACT Act's functional status data reporting requirement. Some commenters specifically expressed support for our proposal that data elements used to calculate Application of Percent of Long-Term Care Hospital Patients with an Admission and Discharge Functional Assessment and a Care Plan that Addresses Function (NQF #2631) be used to meet the definition of standardized resident assessment data for functional status. One commenter noted that their support of standardized resident assessment data was contingent on not adding to facilities' costs or burden.

Response: We appreciate the commenters' support of the functional status standardized resident assessment data for SNFs. These standardized resident assessment data have the potential to facilitate communication among providers and improve care. With regard to burden and cost, we would like to clarify that the data elements from the quality measure Application of Percent of Long-Term Care Hospital Patients with an Admission and Discharge Functional Assessment and a Care Plan that Addresses Function (NQF #2631) are data elements that are currently being collected on the MDS by SNFs, and therefore, there is no additional burden or cost associated with this reporting.

Comment: One commenter requested that we clarify that reporting on the Discharge Goal items for each mobility and self-care item in the SNF PPS admission assessment is for SNF QRP reporting purposes, and does not require a care plan to be developed for each discharge goal.

Response: The proposal to use the data elements used to calculate the function process quality measure as standardized resident assessment data refers to the admission and discharge performance self-care and mobility items. The adopted measure Application of Percent of Long-Term Care Hospital Patients with an Admission and Discharge Functional Assessment and a Care Plan that Addresses Function (NQF #2631) requires that only one goal be reported for each SNF patient stay, and that the requirement for that quality measure remains unchanged. Reporting one goal on the MDS satisfies the measure numerator care plan criteria. The SNF does not need to provide any further documentation about a resident's care plan.

Final Decision: Based on the evidence provided above, we are finalizing that the data elements currently reported by SNFs to calculate the measure, Application of Percent of Long-Term Care Hospital Patients with an Admission and Discharge Functional Assessment and a Care Plan That Addresses Function (NQF #2631), would also meet the definition of standardized resident assessment data for functional status under section 1899B(b)(1)(B)(i) of the Act, and that the successful reporting of that data under section 1886(m)(5)(F)(i) of the Act would also satisfy the requirement to report standardized resident assessment data under section 1886(m)(5)(F)(ii) of the Act.

(2) Medical Condition and Comorbidity Data

We proposed that the data elements needed to calculate the current measure, Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (Short Stay) (NQF #0678), and the proposed measure, Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury, meet the definition of standardized resident assessment data for medical conditions and co-morbidities under section 1899B(b)(1)(B)(iv) of the Act, and that the successful reporting of that data under section 1888(e)(6)(B)(i)(II) of the Act would also satisfy the requirement to report standardized resident assessment data under section 1888(e)(6)(B)(i)(III) of the Act.

“Medical conditions and comorbidities” and the conditions addressed in the standardized resident assessment data used in the calculation and risk adjustment of these measures, that is, the presence of pressure ulcers, diabetes, incontinence, peripheral vascular disease or peripheral arterial disease, mobility, as well as low body mass index, are all health-related conditions that indicate medical complexity that can be indicative of underlying disease severity and other comorbidities.

Specifically, the data elements used in the measure are important for care planning and provide information pertaining to medical complexity. Pressure ulcers are serious wounds representing poor healthcare outcomes, and can result in sepsis and death. Assessing skin condition, care planning for pressure ulcer prevention and healing, and informing providers about their presence in patient transitions of care is a customary and best practice. Venous and arterial disease and diabetes are associated with low blood flow which may increase the risk of tissue damage. These diseases are indicators of factors that may place individuals at risk for pressure ulcer development and are therefore important for care planning. Low BMI, which may be an indicator of underlying disease severity, may be associated with loss of fat and muscle, resulting in potential risk for pressure ulcers. Bowel incontinence and the possible maceration to the skin associated, can lead to higher risk for pressure ulcers. In addition, the bacteria associated with bowel incontinence can complicate current wounds and cause local infection. Mobility is an indicator of impairment or reduction in mobility and movement which is a major risk factor for the development of pressure ulcers. Taken separately and together, these data elements are important for care planning, transitions in services and identifying medical complexities.

In sections III.D.2.g.1. and III.D.2.j.1. of this final rule, we discuss our rationale for proposing that the data elements used in the measures meet the definition of standardized resident assessment data. In summary, we believe that the collection of such assessment data is important for multiple reasons, including clinical decision support, care planning, and quality improvement, and that the data elements assessing pressure ulcers and the data elements used to risk adjust showed good reliability. We solicited stakeholder feedback on the quality measure, and the data elements from which it is derived, by means of a public comment period and TEPs, as described in section III.D.2.g.1. of this final rule.

We sought public comment on this proposal. A discussion of these comments, along with our responses, appears below.

Comment: We received support for the reporting of data elements already implemented in the SNF QRP to satisfy the requirement to report standardized resident assessment data. Specifically, many commenters supported the use of data elements used in calculation of the current measure, Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (Short Stay) (NQF #0678), or the proposed measure, Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury, to fulfill this requirement.

Response: We appreciate the comments in support of the proposal, and agree that these data elements currently reported by SNFs meet the definition of standardized resident assessment data and satisfy the requirement to report standardized resident assessment data.

Final Decision: After consideration of the public comments we received, we are finalizing as proposed that the data elements currently reported by SNFs to calculate the current measure, Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (Short Stay) (NQF #0678), and the proposed measure, Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury, meet the definition of standardized resident assessment data for medical conditions and co-morbidities under section 1899B(b)(1)(B)(iv) of the Act, and that the successful reporting of that data under section 1888(e)(6)(B)(i)(II) of the Act would also satisfy the requirement to report standardized resident assessment data under section 1888(e)(6)(B)(i)(III) of the Act.

k. Form, Manner, and Timing of Data Submission Under the SNF QRP

(1) Start Date for Standardized Resident Assessment Data Reporting by New SNFs

In the FY 2016 SNF PPS final rule (80 FR 46455), we adopted timing for new SNFs to begin reporting quality data under the SNF QRP beginning with the FY 2018 SNF QRP. We proposed in the FY 2018 SNF PPS proposed rule (82 FR 21076) that new SNFs will be required to begin reporting standardized resident assessment data on the same schedule.

We sought public comment on the proposal that new SNFs will be required to begin reporting standardized resident assessment data on the same schedule. A discussion of these comments, along with our responses, appears below.

Comment: We received a comment in support of maintaining the same start date policy for both standardized resident assessment data and SNF QRP measures as this creates consistency in reporting.

Response: We appreciate the commenter's support for extending this policy to the standardized resident assessment data under the SNF QRP.

Final Decision: We are finalizing that new SNFs will be required to begin reporting standardized resident assessment data on the same schedule that they are currently required to begin reporting other quality data under the SNF QRP.

(2) Mechanism for Reporting Standardized Resident Assessment Data Beginning With the FY 2019 SNF QRP

Under our current policy, SNFs report data by completing applicable sections of the MDS, and submitting the MDS-RAI to CMS through the QIESASAP system. For more information on SNF QRP reporting through the QIES ASAP system, refer to the “Related Links” section at the bottom of https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/index.html?redirect=/NursingHomeQualityInits/30_NHQIMDS30TechnicalInformation.asp#TopOfPage. In addition to the data currently submitted on quality measures as previously finalized and discussed in section III.D.2.f. of this final rule, in the FY 2018 SNF PPS proposed rule (82 FR 21076) we proposed that SNFs would be required to begin submitting the proposed standardized resident assessment data for SNF Medicare resident admissions and discharges that occur on or after October 1, 2018 using the MDS. Details on the modifications and assessment collection for the MDS for the proposed standardized resident assessment data are available at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html.

We sought public comments on this proposal. A discussion of these comments, along with our responses, appears below.

Comment: A commenter expressed support for maintaining the same data submission mechanism policy for submitting both standardized resident assessment data and data on SNF QRP measures, as this facilitates consistency in reporting.

Response: We appreciate the commenter's support.

Final Decision: We are finalizing that beginning with the FY 2019 SNF QRP, SNFs will be required to begin submitting standardized resident assessment data for SNF Medicare resident admissions and discharges that occur on or after October 1, 2018 using the MDS. We note that for the FY 2019 SNF QRP, the standardized resident data elements are already submitted using the same (existing) data submission mechanism.

(3) Schedule for Reporting Standardized Resident Assessment Data Beginning With the FY 2019 SNF QRP

Starting with the FY 2019 SNF QRP, we proposed to apply our current schedule for the reporting of measure data to the reporting of standardized resident assessment data. Under this proposed policy, except for the first program year for which a measure is adopted, SNFs must report data on measures for SNF Medicare admissions that occur during the 12-month calendar year (CY) period that apply to the program year. For the first program year for which a measure is adopted, SNFs are only required to report data on SNF Medicare admissions that occur on or after October 1 and discharged from the SNF up to and including December 31 of the calendar year that applies to that program year. For example, for the FY 2018 SNF QRP, data on measures adopted for earlier program years must be reported for all CY 2016 SNF Medicare admissions that occur on or after October 1, 2016 and discharges that occur on or before December 31, 2016. However, data on newly adopted measures for the FY 2018 SNF QRP program year must only be reported for SNF Medicare admissions and discharges that occur during the last calendar quarter of 2016.

Tables 20 and 21 illustrate this policy using the FY 2019 and FY 2020 SNF QRP as examples.

In the FY 2018 SNF PPS proposed rule (82 FR 21076 through 21077), we proposed that for the SNF QRP starting with the 2019 SNF QRP, we would apply our current schedule for the reporting of measure data to the reporting of standardized resident assessment data. Specifically, we proposed to apply to the submission of standardized resident assessment data our policy that except for the first program year for which a measure is adopted, SNFs must report data on measures for SNF Medicare admissions that occur during the 12 month calendar year period that apply to the program year and that for the first program year for which a measure is adopted, SNFs are only required to report data on SNF Medicare admissions that occur on or after October 1 and are discharged from the SNF up to and including December 31 of the calendar year that applies to the program year. We sought comment on our proposal to extend our current policy governing the schedule for reporting the quality measure data to the reporting of standardized resident assessment data beginning with the FY 2019 SNF QRP. A discussion of these comments, along with our responses, appears below.

Comment: Commenters supported our proposal to adopt the same data reporting schedule for both standardized resident assessment data and SNF QRP measure data as this creates consistency in reporting. Another commenter added that we should allow facilities to become familiar with the assessment and coding requirements associated with the new standardized resident assessment data elements for a period of time before quality measure reporting begins.

Response: We appreciate commenters' support to extend this policy to the standardized resident assessment data submitted under the SNF QRP. We agree that comprehensive training is needed to ensure accurate data collection and to ensure successful reporting on new measures that are constructed using the new data. As with the data collection required on new assessment data collection in the past, we will provide training sessions, training manuals, webinars, open door forums, help desk support, and a Web site that hosts training information and will continue to provide the training providers may need to understand item concepts and coding instructions.

Comment: In light of the additional data elements being proposed for the MDS, one commenter recommended that the reporting data for the purposes of quality measures for the SNF QRP not begin at the same time as new items are added to the MDS, and requested at least a 3-month time frame of data collection with the new items before the data is collected for use in a quality measure.

Response: We interpret the comment to mean that given the new data elements and need for SNFs to become familiar with the coding of the new standardized resident assessment data elements, the commenter believes that we should not use the first three months of data in the calculation of the measures to be publicly reported. We acknowledge that SNFs may need time to transition to new data reporting requirements. As discussed previously, data collection on new measures that are calculated using resident assessment data begins using a schedule that starts on October 1 of a given year, we anticipate using the subsequent calendar year of data for public reporting.

Final Decision: After careful consideration of the public comments, we are finalizing our proposal to extend our current policy governing the schedule for reporting quality measure data to the standardized resident assessment data elements beginning with the FY 2019 SNF QRP.

(4) Schedule for Reporting the Quality Measures Beginning with the FY 2020 SNF QRP

As discussed in section III.D.2.g. of this final rule, we are finalizing the adoption of five quality measures beginning with the FY 2020 SNF QRP: (1) Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury; (2) Application of IRF Functional Outcome Measure: Change in Self-Care for Medical Rehabilitation Patients (NQF #2633); (3) Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634); (4) Application of IRF Functional Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (NQF #2635); (5) and Application of IRF Functional Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (NQF #2636). In the FY 2018 SNF PPS proposed rule (82 FR 21077) we proposed that SNFs would report data on these measures using the MDS that is submitted through the QIES ASAP system. For the FY 2020 SNF QRP, SNFs would be required to report these data for admissions as well as discharges that occur between October 1, 2018 and December 31, 2018. More information on SNF reporting using the QIES ASAP system is located at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/index.html?redirect=/NursingHomeQualityInits/30_NHQIMDS30TechnicalInformation.asp#TopOfPage. Starting in CY 2019, SNFs would be required to submit data for the entire calendar year beginning with the FY 2021 SNF QRP.

We sought public comment on this proposal. A discussion of these comments, along with our responses, appears below.

Comment: Two commenters supported our proposal that SNFs report admission and discharge data for the five quality measures beginning with the FY 2020 SNF QRP using the QIES ASAP system.

Response: We thank the commenters for their support.

Final Decision: We are finalizing our policy as proposed for the Schedule for Reporting the Quality Measures Beginning with the FY 2020 SNF QRP.

(5) Input Sought on Data Reporting Related to Assessment Based Measures

Through various means of public input, including that through previous rules (FY 2016 SNF PPS final rule, 80 FR 46415), public comment on measures, and the MAP, we received input suggesting that we expand the quality measures to include all residents and patients regardless of payer status so as to ensure representation of the quality of the services provided on the population as a whole, rather than a subset limited to Medicare. While we appreciate that many SNF residents are also Medicare beneficiaries, we agree that collecting quality data on all residents in the SNF setting supports our mission to ensure quality care for all individuals, including Medicare beneficiaries. We also agree that collecting data on all patients provides the most robust and accurate reflection of quality in the SNF setting. Accurate representation of quality provided in SNFs is best conveyed using data on all SNF residents, regardless of payer. We also appreciate that collecting quality data on all SNF residents regardless of payer source may create additional burden. However, we also note that the effort to separate out SNF residents covered by other non-FFS Medicare payers could have clinical and work flow implications with an associated burden, and we further appreciate that it is common practice for SNFs to collect MDS data on all residents regardless of payer source. Additionally, we note that data collected through MDS for Medicare beneficiaries should match that beneficiary's claims data in certain key respects (for example, diagnoses and procedures); this makes it easier for us to evaluate the accuracy of reporting in the MDS, such as by comparing diagnoses at hospital discharge to diagnoses at the follow-on SNF admission. However, we would not have access to such claims data for non-Medicare beneficiaries. Thus, we sought input on whether we should require quality data reporting on all SNF residents, regardless of payer, where feasible—noting that Part A claims data are limited to only Medicare beneficiaries.

We sought comments on this topic. A discussion of these comments, along with our responses, appears below.

Comment: We received overwhelming support from commenters including MedPAC and others for the expansion of quality measures to include all residents regardless of payer. Several commenters as well as MedPAC expressed the benefit of enabling comparisons between FFS beneficiaries and other users (including beneficiaries enrolled in Medicare Advantage), expressing that such data would serve to better inform beneficiaries on the broader quality of the entire facility, especially those who are or will become long-term care residents of the same facility. MedPAC also highlighted that while the data collection activity incurs some cost, some providers currently assess all residents routinely. Some commenters conveyed that data collection on all payers is more feasible than having to select only Medicare populations. Several commenters noted that it is advantageous for facilities to focus on quality outcomes for all residents regardless of payer, and several commenters noted that having information on rates for all residents regardless of payor allows providers to utilize these measures in system-based quality improvement initiatives.

One commenter noted a preference for using claims-based data and urged that claims-based SNF QRP measures be re-specified to allow for this inclusion. Another commenter highlighted the value in using readily available MDS assessment-based data to better represent facility performance on measures previously reported using Medicare Part A claims data only.

Response: We acknowledge support for this policy from MedPAC and other commenters. We agree that having such information from all payers adds value to data comparisons, allows enhanced use of assessment data already being collected on all residents, and further supports system-wide quality improvement goals.

(l) Application of the SNF QRP Data Completion Thresholds to the Submission of Standardized Resident Assessment Data Beginning with the FY 2019 SNF QRP

We have received questions surrounding the data completion policy we adopted beginning with the FY 2018 program year, specifically with respect to how that policy applies to patients who reside in the SNF for part of an applicable period, for example, a patient who is admitted to a SNF during one reporting period but discharged in another, or a patient who is assessed upon admission using one version of the MDS but assessed at discharge using another version. We previously finalized in the FY 2016 SNF PPS final rule (80 FR 46458) that SNFs must report all of the data necessary to calculate the measures that apply to that program year on at least 80 percent of the MDS assessments that they submit. The term “measures” refers to quality measures, resource use, and other measures. We also stated, in response to a comment, that we would consider data to have been satisfactorily submitted for a program year if the SNF reported all of the data necessary to calculate the measures if the data actually can be used for purposes of such calculations (as opposed to, for example, the use of a dash [-]).

Some stakeholders interpreted our requirement that data elements be necessary to calculate the measures to mean that if a patient is assessed, for example, using one version of the MDS at admission and another version of the MDS at discharge, the two assessments are included in the pool of assessments used to determine data completion only if the data elements at admission and discharge can be used to calculate the measures. Our intention, however, was not to exclude assessments on this basis. Rather, our intention was solely to clarify that for purposes of determining whether a SNF has met the data completion threshold, we would only look at the completeness of the data elements in the MDS for which reporting is required under the SNF QRP.

To clarify our intended policy, in the FY 2018 SNF PPS proposed rule (82 FR 21077 through 21078), we proposed that for the purposes of determining whether a SNF has met the data completion threshold, we would consider whether the SNF has reported all of the required data elements applicable to the program year on at least 80 percent of the MDS assessments that they submit for that program year. For example, if a resident is admitted on December 20, 2017 but discharged on January 10, 2018: (1) The resident's 5-Day PPS assessment would be used to determine whether the SNF met the data completion threshold for the 2017 reporting period (and associated program year), and (2) the discharge assessment would be used to determine whether the SNF met the data completion threshold for the 2018 reporting period (and associated program year). We also clarified in the FY 2018 SNF PPS proposed rule (82 FR 21078) that some assessment data will not invoke a response; in those circumstances, data are not “missing” or incomplete. For example, in the case of a resident who does not have any of the medical conditions in a check all that apply listing, the absence of a response indicates that the condition is not present, and it would be incorrect to consider the absence of such data as missing in a threshold determination.

We also proposed to apply this policy to the submission of standardized resident assessment data, and to codify it at § 413.360(b) of our regulations. We sought comment on these proposals. A discussion of these comments, along with our responses, appears below.

Comment: We received a comment noting the usefulness of a document we published indicating which data we would be using to determine compliance by SNFs beginning with the FY 2018 SNF QRP. The commenter also requested that we continue providing that resource. The commenter also acknowledged our clarification of which MDS assessments are included in compliance determinations when the resident admission occurs in one reporting period for the SNF QRP, while their discharge occurs in a subsequent reporting period. The commenter further acknowledged our clarification that an MDS item will not be considered as missing data in the circumstances when no response is necessary.

Another commenter requested additional explanation and examples regarding how the threshold compliance calculation is applied. One commenter suggested that the 80 percent data completion threshold finalized in the SNF PPS FY 2016 final rule is set too low and requested that, for the FY 2018 payment determination year and beyond, the data completion threshold be increased to at least ninety percent. We also received a comment suggesting that requiring that SNFs submit data on 100% of all items necessary to calculate quality measures and all additional standardized resident assessment data elements is set too high. They also expressed that the tracking of dash use, which is what is used to determine compliance, is burdensome. Another commenter suggested that we omit the first quarter of required data reporting in our determination of compliance given the newness of the reporting. They further expressed that for FY 2018 SNF QRP, the Review and Correct reports that were proved were unavailable for the SNFs to help them identify if they were successful in meeting the compliance threshold.

One commenter did not support the codification of this proposal in our regulations with respect to the FY 2019 SNF QRP, and requested that we first review the results of the initial implementation of this policy and propose such codification in the future.

Response: We appreciate the commenter's support of the materials we provided to help SNFs identify the required MDS data elements for accurate submission in order to meet the requirements of the SNF QRP. We have published the document, Technical Specifications for Reporting Assessment-Based Measures for FY2018, which identifies item completion specifications for calculation of missing data rates on our Web site at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html and intend to update this resource document as suggested.

We do not believe that the Review and Correct Reports would be an appropriate mechanism for informing SNFs whether they have complied with our data completion threshold. This report is intended to provide SNFs information related to their overall quality measure calculations. It will not provide SNFs with the discrete, data element level information on what response was coded for every resident assessment data element. We refer to the CMS SNF QRP Training Web site for detailed information on the Review and Correct Reports: https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Training.html.

Although the Review and Correct Reports do not enable SNFs to track the coding of dashes which is what can lead to non-compliance, we provide other reports via the Certification and Survey Provider Enhanced Reports Reporting (CASPER) System which SNFs can use to track their dash use in the assessment data they have submitted and other submission information. These reports include: Submitter Validation Reports, Facility Final Validation Reports, Error Detail by Facility Reports, Activity or Submission Activity Reports and Assessment Print Reports. We are also looking into other mechanisms and reports that would serve to further assist SNFs in easily identifying their data completion thresholds.

To illustrate an example as requested, if a provider submitted 100 records in a reporting period and 80% of those records had all of the standardized resident assessment data elements that we require and the data necessary to calculate the measures used in the SNF QRP, the SNF would meet our compliance determination.

We currently believe that the completion of all of the required data elements on at least 80 percent of all required assessments is a fair criterion for a new program and is consistent with other post-acute care programs. Regarding the suggestion that we not consider the initial quarter of data reporting by SNFs on new data that is required, we have analyzed the first quarter of data reporting and found that most SNFs were successful in their data submission. We appreciate that SNFs seek to track their compliance rates and the burden that may be associated with their tracking of such data submission. However, we believe that ensuring the submission of accurate data is an inherent responsibility of the SNF. We note that the use of dashes, which is what can lead to a determination of non-compliance, should be rare in that the assessment data collected is required and the expectation is that SNFs perform these assessments on their residents for not only data reporting purposes for the SNF QRP, but also for other purposes as well. As has been noted, overall dash use by SNFs is already low. That said, the reports we provide can assist in a SNF's tracking of their dash rates and we will evaluate other types of reports that can assist.

Final Decision: We are finalizing our proposal to apply the threshold levels as proposed, to extend this policy to the submission of standardized resident assessment data, and to codify the requirement at § 413.360(b) of our regulations.

m. SNF QRP Data Validation Requirements

We refer readers to the FY 2016 SNF PPS final rule (80 FR 46458 through 46459) for a summary of our approach to the development of data validation process for the SNF QRP. At this time, we are continuing to explore data validation methodology that will limit the amount of burden and cost to SNFs, while allowing us to establish estimations of the accuracy of SNF QRP data.

n. SNF QRP Submission Exception and Extension Requirements

We refer readers to the FY 2016 SNF PPS final rule (80 FR 46459 through 46460) for our finalized policies regarding submission exception and extension requirements for the FY 2018 SNF QRP. We did not propose any changes to the SNF QRP requirements that we adopted in these final rules. However, in the FY 2018 SNF PPS proposed rule (82 FR 21078) we proposed to codify the SNF QRP Submission Exception and Extension Requirements at new § 413.360(c).

We remind readers that, in the FY 2016 SNF PPS final rule (80 FR 46459 through 46460) we stated that SNF's must request an exception or extension by submitting a written request along with all supporting documentation to CMS via email to the SNF Exception and Extension mailbox at SNFQRPReconsiderations@cms.hhs.gov. We further stated that exception or extension requests sent to CMS through any other channel would not be considered as a valid request for an exception or extension from the SNF QRP's reporting requirements for any payment determination. To be considered, a request for an exception or extension must contain all of the requirements as outlined on our Web site at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-QR-Reconsideration-and-Exception-and-Extension.html . We sought public comments on our proposal to codify the SNF QRP submission exception and extension requirements. A discussion of these comments, along with our responses, appears below.

Comment: A few commenters did not support codification of the SNF QRP Submission Exception and Extension Requirements until one SNF QRP program year has been completed.

Response: Our proposal to codify existing policy in our regulations was technical in nature and would have no effect on its existing applicability and enforceability. To the extent that the commenter was asking us to delay the effective date of this policy, we did not propose such a delay, and we believe that SNFs will benefit from having this process available to them in the event that they experience an extraordinary circumstance during the FY 2018 program year.

Final Decision: After considering the comments we received, we are codifying the SNF QRP submission exception and extension requirements at § 413.360(c) of our regulations.

o. SNF QRP Submission Reconsideration and Appeals Procedures

We refer the reader to the FY 2016 SNF PPS final rule (80 FR 46460 through 46461) for a summary of our finalized reconsideration and appeals procedures for the SNF QRP beginning with the FY 2018 SNF QRP. We did not propose any changes to these procedures in the FY 2018 SNF PPS proposed rule (82 FR 21078). However, we proposed to codify the SNF QRP Reconsideration and Appeals procedures at new § 413.360(d). Under these procedures, a SNF must follow a defined process to file a request for reconsideration if it believes that a finding of noncompliance with the reporting requirements for the applicable fiscal year is erroneous, and the SNF can file a request for reconsideration only after it has been found to be noncompliant. To be considered, a request for a reconsideration must contain all of the elements outlined on our Web site at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-QR-Reconsideration-and-Exception-and-Extension.html. We stated that we would not review any reconsideration request that is not accompanied by the necessary documentation and evidence, and that the request should be emailed to CMS at the following email address: SNFQRPReconsiderations@cms.hhs.gov. We further stated that reconsideration requests sent to CMS through any other channel would not be considered.

We sought public comments on our proposal to codify the SNF QRP reconsideration and appeals procedures. A discussion of these comments, along with our responses, appears below.

Comment: Several commenters did not support the codification of SNF QRP Submission Reconsideration and Appeals Procedures until at least the FY 2018 SNF QRP program year has been completed.

Response: Our proposal to codify existing policy in our regulations was technical in nature and would have no effect on its existing applicability and enforceability. To the extent that the commenter was asking us to delay the effective date of this policy, we did not propose such a delay, and we believe that SNFs will benefit from having this process available to them in the event that they wish to seek reconsideration during the FY 2018 program year.

Final Decision: After considering the comments, we are finalizing our decision to codify the SNF QRP submission reconsideration and appeals requirements at new § 413.360(d) of our regulations.

p. Policies Regarding Public Display of Measure Data for the SNF QRP

Section 1899B(g) of the Act requires the Secretary to establish procedures for the public reporting of SNFs' performance, including the performance of individual SNFs, on the quality measures specified under section (c)(1) and resource use and other measures specified under section (d)(1) of the Act (collectively, IMPACT Act measures) beginning not later than 2 years after the specified application date under section 1899B(a)(2)(E) of the Act. This is consistent with the process applied under section 1886(b)(3)(B)(viii)(VII) of the Act, which refers to the public display and review requirements for the Hospital Inpatient Quality Reporting (IQR) Program. For a more detailed discussion about the provider's confidential review process prior to public display of measures, we refer readers to the FY 2017 SNF PPS final rule (81 FR 52045 through 52048).

In the FY 2018 SNF PPS proposed rule, pending the availability of data, we proposed to publicly report data in CY 2018 for the following 3 assessment-based measures: (1) Application of Percent of Long-Term Care Hospital (LTCH) Patients With an Admission and Discharge Functional Assessment and a Care Plan That Addresses Function (NQF #2631); (2) Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (NQF #0678); and (3) Application of Percent of Residents Experiencing One or More Falls with Major Injury (NQF #0674). Data collection for these 3 assessment-based measures began on October 1, 2016. We proposed to display data for the assessment-based measures based on rolling quarters of data, and we would initially use discharges from January 1, 2016 through December 31, 2016.

In addition, we proposed to publicly report 3 claims-based measures for: (1) Medicare Spending Per Beneficiary-PAC SNF QRP; (2) Discharge to Community-PAC SNF QRP; and (3) Potentially Preventable 30-Day Post-Discharge Readmission Measure for SNF QRP.

These measures were adopted for the SNF QRP in the FY 2017 SNF PPS rule to be based on data from one calendar year. As previously adopted in the FY 2017 SNF PPS final rule (81 FR 52045 through 52047), confidential feedback reports for these 3 claims-based measures will be based on data collected for discharges beginning January 1, 2016 through December 31, 2016. However, our current proposal revises the dates for public reporting and we proposed to transition from calendar year to fiscal year to make these measure data publicly available by October 2018.

For the Medicare Spending Per Beneficiary-PAC SNF QRP and Discharge to Community-PAC SNF QRP measures, we proposed public reporting beginning in calendar year 2018 based on data collected from discharges beginning October 1, 2016, through September 30, 2017 and rates will be displayed based on one fiscal year of data. For the Potentially Preventable 30-day Post-Discharge Readmission Measure for SNF QRP, we also proposed to increase the years of data used to calculate this measure from one year to 2 years and to update the associated reporting dates. These proposed revisions to the Potentially Preventable 30-Day Post-Discharge Readmission Measure for SNF QRP will result in the data being publicly reported with discharges beginning October 1, 2015, through September 30, 2017 and rates will be displayed based on two consecutive fiscal years of data.

Also, we proposed to discontinue the public display of data on the assessment-based measure “Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (Short Stay) (NQF #0678)” and to replace it with a modified version of the measure entitled “Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury” from the SNF QRP by October 2020.

For the assessment-based measures, Application of Percent of Long-Term Care Hospital (LTCH) Patients With an Admission and Discharge Functional Assessment and a Care Plan That Addresses Function (NQF #2631); Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (NQF #0678); and Application of Percent of Residents Experiencing One or More Falls with Major Injury (NQF #0674), to ensure the statistical reliability of the measures, we proposed to assign SNFs with fewer than 20 eligible cases during a performance period to a separate category: “The number of cases/resident stays is too small to report”. If a SNF had fewer than 20 eligible cases, then the SNF's performance would not be publicly reported for the measure for that performance period.

For the claims-based measures Medicare Spending Per Beneficiary-PAC SNF QRP; Discharge to Community-PAC SNF QRP; and Potentially Preventable 30-Day Post-Discharge Readmission Measure for SNF QRP, we proposed to assign SNFs with fewer than 25 eligible cases during a performance period to a separate category: “The number of cases/resident stays is too small to report,” to ensure the statistical reliability of the measures. If a SNF had fewer than 25 eligible cases, the SNF's performance would not be publicly reported for the measure for that performance period. For Medicare Spending Per Beneficiary-PAC SNF QRP we proposed to assign SNFs with fewer than 20 eligible cases during a performance period to a separate category: “The number of cases/resident stays is too small to report” to ensure the statistical reliability of the measure. If a SNF has fewer than 20 eligible cases, the SNF's performance would not be publicly reported for the measure for that performance period.

We invited public comment on the proposal for the public display of these three assessment-based measures and three claims-based measures, and the replacement of “Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (Short Stay) (NQF #0678)” with a modified version of the measure, “Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury” described above. A discussion of these comments, along with our responses, appears below.

Comment: A commenter requested that we consider aligning the public reporting periods and provider deadlines across PAC settings and other CMS programs.

Response: We are working to achieve alignment where possible. For example, with respect to the following 3 assessment-based measures: (1) Application of Percent of Long-Term Care Hospital (LTCH) Patients With an Admission and Discharge Functional Assessment and a Care Plan That Addresses Function (NQF #2631); (2) Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (NQF #0678); and (3) Application of Percent of Residents Experiencing One or More Falls with Major Injury (NQF #0674), we intend to initially report data using discharges from January 1, 2017 through December 31, 2017 for the public display of data, which aligns with the IRF and LTCH QRPs.

Comment: A commenter supported the proposed minimum denominator requirements for public display.

Response: We appreciate the commenter's support.

Comment: A few commenters supported the public display of assessment-based measures based on rolling quarters since it reflects more recent SNF quality performance.

Response: We appreciate the commenters' support.

Final Decision: After consideration of the public comments we received, we are finalizing that we intend to begin publicly reporting in 2018 the following assessment-based measures based on the availability of data: (1) “Application of Percent of Long-Term Care Hospital (LTCH) Patients With an Admission and Discharge Functional Assessment and a Care Plan That Addresses Function (NQF #2631); (2) Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (NQF #0678); and (3) Application of Percent of Residents Experiencing One or More Falls with Major Injury (NQF #0674), as well as the following claims-based measures: (1) “Medicare Spending Per Beneficiary-PAC SNF QRP; (2) Discharge to Community-PAC SNF QRP; and (3) Potentially Preventable 30-Day Post-Discharge Readmission Measure for SNF QRP. In addition, we will discontinue the public reporting of data on the assessment-based measure: “Percent of Residents or Patients with Pressure Ulcers That Are New or Worsened (Short Stay) (NQF #0678)” by October 2020.

q. Mechanism for Providing Confidential Feedback Reports to SNFs

Section 1899B(f) of the Act requires the Secretary to provide confidential feedback reports to PAC providers on their performance on the measures specified under subsections (c)(1) and (d)(1) of section 1899B of the Act, beginning 1 year after the specified application date that applies to such measures and PAC providers. In the FY 2017 SNF PPS final rule (81 FR 52046 through 52048), we finalized processes to provide SNFs the opportunity to review their data and information using confidential feedback reports that will enable SNFs to review their performance on the measures required under the SNF QRP. Information on how to obtain these and other reports available to the SNF QRP can be found at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Spotlights-and-Announcements.html. We did not propose any changes to this policy but received comments, which are discussed below.

Comment: A few commenters requested more granular resident-specific data in the reports.

Response: Resident level data will be available in the CASPER QM reports.

Comment: A commenter suggested that we provide confidential feedback reports to SNFs prior to the time that we publicly display their quality measure data.

Response: Before publicly displaying measure scores, providers have several opportunities to review their facility- and resident-level data to ensure the accuracy of quality measure scores. Two separate confidential feedback reports will be provided, in addition to Review and Correct reports, for providers to review their single quarter and aggregate quality measure scores, respectively. The confidential feedback reports are the QM facility- and resident-level reports that will be available to providers beginning in fall 2017, which is prior to public display, and contain quality measure information for a single reporting period. The facility-level QM reports will provide information such as the numerator, denominator, facility observed percent, facility adjusted percent, and national average. The resident-level QM reports will contain individual resident data and provide information related to which residents were included in the quality measures.

The Review and Correct reports, currently available to SNFs, provide aggregate performance for up to the past four full quarters as the data are available. The reports contain information on assessment based measures performance at the facility-level and observed rates. The reports also display data correction deadlines and whether the data correction period is open or closed. Please refer to the SNF QRP Web site for information from the training on the Review and Correct reports: https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Training.html.

Finally, the Provider Preview reports will be available beginning in the summer of 2018. Provider Preview reports are available about 5 months after the end of each reporting period. They contain facility-level quality measure data results and will contain information such as the numerator, denominator, facility observed percent, facility adjusted percent, and national average. Providers will have 30 days upon receiving the Provider Preview reports via their CASPER system folders to review their data. We note at that point in time providers are no longer able to correct the underlying data in these reports. At this point, the data correction period has ended so providers are not able to correct the underlying data in these reports.

3. Skilled Nursing Facility Value-Based Purchasing Program (SNF VBP)

a. Background

Section 215 of the Protecting Access to Medicare Act of 2014 (PAMA) (Pub. L. 113-93) authorized the SNF VBP Program (the “Program”) by adding sections 1888(g) and (h) to the Act. As a prerequisite to implementing the SNF VBP Program, in the FY 2016 SNF PPS final rule (80 FR 46409 through 46426) we adopted an all-cause, all-condition hospital readmission measure, as required by section 1888(g)(1) of the Act. In the FY 2017 SNF PPS final rule (81 FR 51986 through 52009), we adopted an all-condition, risk-adjusted potentially preventable hospital readmission measure for SNFs, as required by section 1888(g)(2) of the Act. In this final rule, we are finalizing proposals related to the Program's implementation.

Section 1888(h)(1)(B) of the Act requires that the SNF VBP Program apply to payments for services furnished on or after October 1, 2018. The SNF VBP Program applies to freestanding SNFs, SNFs affiliated with acute care facilities, and all non-CAH swing-bed rural hospitals. We believe the implementation of the SNF VBP Program is an important step towards transforming how care is paid for, moving increasingly towards rewarding better value, outcomes, and innovations instead of merely volume.

For additional background information on the SNF VBP Program, including an overview of the SNF VBP Report to Congress and a summary of the Program's statutory requirements, we refer readers to the FY 2016 SNF PPS final rule (80 FR 46409 through 46410). We also refer readers to the FY 2017 SNF PPS final rule (81 FR 51986 through 52009) for discussion of the policies that we adopted related to the potentially preventable hospital readmission measure, scoring, and other topics.

In this rule, we are finalizing requirements for the SNF VBP Program, as well as codifying some of those requirements at § 413.338, including certain definitions, the process for making value-based incentive payments, and limitations on review.

We received several general comments on the SNF VBP Program. We note that we did not receive any comments specific to the proposed regulation text. A discussion of the general comments that we received, along with our responses, appears below.

Comment: One commenter urged us to seek the statutory authority to broaden the scope of the SNF VBP Program to include other post-acute care outcome measures beyond measures of readmissions.

Response: We thank the commenter for this suggestion.

Comment: One commenter suggested that we authorize the inclusion of certified peer specialists in value-based, patient-centered treatment, as well as transition teams assigned to nursing home patients with mental illness or substance use disorders who might benefit in recovery from a return to community-based services. The commenter stated that peer support specialists' work could result in savings to the Medicare Program due to reduced rehospitalizations and from reduced medical expenditures for recurring medical conditions.

Response: We appreciate the comment. We will consider whether peer support specialists could play a role providing technical assistance to SNFs to help them reduce avoidable hospital readmissions through our collaboration with the CMS Quality Innovation and Improvement Network.

Comment: One commenter suggested that we analyze the New York State Nursing Home Quality Initiative, which the commenter stated incorporates quality, compliance and efficiency with a focus on potentially avoidable hospitalizations. While the initiative is limited to long-stay Medicaid patients, the commenter stated that it presents several important lessons for the SNF VBP Program. The commenter specifically pointed to the need to structure measures narrowly for participating facilities, regional adjustments, and detailed information that the commenter believes must be provided to participating facilities. The commenter also stated that potentially avoidable hospitalizations are the most important factor, and that incentive payments must be large enough and close enough to the performance period to maximize improvement.

Response: The New York State Nursing Home Quality Initiative “is an annual quality and performance evaluation project to improve the quality of care for residents in Medicaid-certified nursing facilities across New York State.” The initiative scores Medicaid-certified nursing facilities in the state on previous performance and awards up to 100 points for performance on measures of quality, compliance, and efficiency. The initiative also incorporates deficiencies cited during the health inspection survey process and creates an overall score for each facility that forms the basis for a quintile ranking. We appreciate the commenter's suggestion that we consider the New York initiative's results and lessons and we agree that it may be instructive for our continuing SNF VBP Program development. As the commenter noted, its basis in long-stay Medicaid patients differs somewhat from the SNF VBP Program's focus on shorter-stay Medicare patients. However, as the commenter notes, the initiative provides detailed information to participating facilities, a goal that we believe we are now meeting by providing patient-level information to SNF VBP Program participants. We also believe that the SNF VBP Program is, as the commenter suggests, narrowly constructed due to its focus on measures of hospital readmissions, and while we have not considered regional adjustments in the SNF VBP Program to date, we will consider if such adjustments are appropriate in the future.

Comment: One commenter questioned whether the SNF VBP Program's statute actually limits the Program to the specified measures of readmissions, or whether other indicators could be included in performance scoring. The commenter suggested that, at a minimum, we should coordinate our approach and goals between SNF VBP, SNF QRP, and the Staffing Data Collection initiative. Another commenter suggested that we consider additional quality measures for the Program, potentially including measures drawn from Nursing Home Compare, the NH VBP demonstration, or the SNF QRP. The commenter also specifically suggested that we measure turnover as a percentage of nursing staff, total CNA hours per patient day, and total licensed nursing hours per patient day. The commenter stated that these measures can be integrated into SNF VBP because the payroll-based journal staffing information collection system has been operational since July 2016. The commenter also stated that several studies have positively correlated a higher staffing level with higher care quality and outcomes, and stated that such metrics will encourage SNFs to invest in their staffs.

Response: We interpret sections 1888(h)(2)(A) and (B) of the Act to only allow us to include in the Program first the readmission measure specified under section 1888(g)(1), and then in its place, the readmission measure specified under section 1888(g)(2) of the Act. We will continue our collaborative effort with the SNF QRP and Nursing Home Compare programs to align our readmission measure to the fullest extent feasible and practicable. Our collaborative focus area across these programs is to improve the quality of care and reduce hospital readmissions.

We thank the commenters for this feedback.

b. Measures

(1) Background

For background on the measures in the SNF VBP Program, we refer readers to the FY 2016 SNF PPS final rule (80 FR 46419), where we finalized the Skilled Nursing Facility 30-Day All-Cause Readmission Measure (SNFRM) (NQF #2510) that we will use for the SNF VBP Program. We also refer readers to the FY 2017 SNF PPS final rule (81 FR 51987 through 51995), where we finalized the Skilled Nursing Facility 30-Day Potentially Preventable Readmission Measure (SNFPPR) that we will use for the SNF VBP Program instead of the SNFRM as soon as practicable.

(2) Request for Comment on Measure Transition

Section 1886(h)(2)(B) of the Act requires us to apply the SNFPPR to the SNF VBP Program instead of the SNFRM “as soon as practicable.” We intend to propose a timeline for replacing the SNFRM with the SNFPPR in future rulemaking, after we have had a sufficient opportunity to analyze the potential effects of this replacement on SNFs' measured performance. We believe we must approach the decision about when it is practicable to replace the SNFRM thoughtfully, and we continue to welcome public feedback on when it is practicable to replace the SNFRM with the SNFPPR.

In the FY 2017 SNF PPS final rule (81 FR 51995), we summarized the public comments we received in response to our request for when we should begin to measure SNFs on their performance on the SNFPPR instead of the SNFRM. Commenters' views were mixed; one suggested that we replace the SNFRM immediately, while others requested that we wait until the SNFPPR receives NQF endorsement, or that we allow SNFs to receive and understand their SNFPPR data for at least 1 year prior to beginning to use it. Another commenter suggested that we decline to use the SNFPPR until the measure receives additional support from the Measure Application Partnership and is the subject of additional public comment.

We would like to thank stakeholders for their input on this issue. We believe the first opportunity to replace the SNFRM with the SNFPPR would be the FY 2021 program year, which would give SNFs experience with the SNFRM and other measures of readmissions such as those adopted under the SNF QRP. However, we have not yet determined if it would be practicable to replace the SNFRM at that time. We intend to continue to analyze SNF performance on the SNFPPR in comparison to the SNFRM and assess how the replacement of the SNFRM with the SNFPPR will affect the quality of care provided to Medicare beneficiaries.

In the FY 2018 SNF PPS proposed rule, we sought public comments on when we should replace the SNFRM with the SNFPPR, particularly in light of our proposal (discussed further in this section) to adopt performance and baseline periods based on the federal FY rather than on the calendar year. A discussion of these comments, along with our responses, appears below.

Comment: Several commenters supported transitioning to the SNFPPR beginning with the FY 2021 program year as long as the measure has received NQF endorsement. Commenters stated that the measure's importance to the program necessitates thorough vetting, including NQF endorsement, and agreed that waiting until FY 2021 provides SNFs with the opportunity to gain experience with the SNFRM prior to the measure transition. One commenter requested that we provide a timeline for when the measure will replace the SNFRM.

Response: We appreciate the feedback, and we intend to submit the SNFPPR to NQF for consideration of endorsement as soon as possible. We will address the replacement of the SNFRM with the SNFPPR in future rulemaking.

Comment: One commenter expressed continued concern about the SNFPPR, stating that we should conduct additional testing and analysis of the measure before implementing it in the Program. The commenter specifically requested that we await full endorsement by NQF, and if we intend to proceed with its implementation, that we provide SNFPPR performance information in our quarterly reports to SNFs.

Response: As we noted above, we intend to submit the SNFPPR to NQF for consideration of endorsement as soon as possible. We also intend to provide SNFs with SNFPPR performance information in their quarterly reports prior to future replacement of the SNFRM. We intend to update affected stakeholders on timing in future rulemaking.

Comment: One commenter supported adoption of the SNFPPR and did not have any objection to transitioning the Program to the SNFPPR in FY 2021. The commenter also suggested that we consider including additional measures in the Program to cover other relevant quality improvement topics, such as resource use and functional outcomes.

Response: As we discussed above, we interpret sections 1888(h)(2)(A) and (B) of the Act to only allow us to include in the Program first the readmission measure specified under section 1888(g)(1) of the Act, and then in its place, the readmission measure specified under section 1888(g)(2) of the Act. We intend to provide SNF's with SNFPPR rates prior to the replacement for SNF's to learn more about the measure and incorporate into their quality improvement and care transitions efforts to reduce readmissions. We also intend to further analyze the SNFPPR prior to replacing the SNFRM for any association with social risk factors, in collaboration with the Assistant Secretary for Planning and Evaluation. We intend to update stakeholders on this analysis in future rulemaking.

Comment: One commenter supported transitioning the Program to the SNFPPR in FY 2021, if not sooner, and requested additional information on why we believe that FY 2021 is the first opportunity to transition the Program from the SNFRM.

Response: As we discussed in the FY 2018 SNF PPS proposed rule (82 FR 21080), we concluded that FY 2021 would be the first opportunity to replace the SNFRM with the SNFPPR because we believe that giving SNFs two Program years' experience with the SNFRM will provide them with valuable experience with measures of readmissions that will be helpful for their quality improvement efforts generally and with their specific efforts to improve their scores under the SNF VBP Program. To expand on that point, we did not believe it would be helpful to SNFs' quality improvement efforts to adopt a quality measure for a single year, then to replace that measure after that 1 year, particularly because the Program is limited by statute to a single measure at a time. We viewed that instability in the Program's quality metrics as undesirable and unnecessary. We are also concerned that transitioning the Program too quickly could prove confusing for SNFs and for affected patients.

We also intend to provide SNFs with their SNFPPR rates prior to the replacement so that they have an opportunity to learn more about the measure and incorporate that information into their quality improvement and care transitions efforts to reduce readmissions. We also intend to further analyze the SNFPPR prior to replacing the SNFRM for any association with social risk factors, in collaboration with the Assistant Secretary for Planning and Evaluation. We intend to update stakeholders on this analysis in future rulemaking.

Comment: One commenter recommended that we transition the Program to the SNFPPR no sooner than FY 2021 to allow sufficient time for SNFs to adjust to the measure's implementation.

Response: We agree that SNFs need time to adjust to transitions under the Program, which is why we sought comment in the FY 2017 SNF PPS proposed rule on this topic and again sought comment in the FY 2018 SNF PPS proposed rule. We will consider the commenter's feedback as we determine when it is practicable to transition the Program to the SNFPPR.

We thank the commenters for this feedback and will take it into consideration in the future. We also received a number of unsolicited comments on the SNF VBP Program measures. The comments, together with our responses, appear below.

Comment: One commenter expressed concern about our use of measures of readmissions in the Program. The commenter was particularly concerned that these measures place non-profit facilities at a disadvantage compared to their for-profit competitors because non-profits take all patients, including high-risk and high-acuity level patients. The commenter also stated that the measures' risk adjustment methodologies do not fully capture the additional effort needed to treat these patients in the SNF setting, such as the risk of patient non-compliance with medical direction after discharge. The commenter requested that we provide additional transparency into claims-based quality measures in order to improve providers' understanding of their calculations and methodologies.

Response: We thank the commenter for this feedback, but we disagree with their concern. As we discussed in the FY 2016 SNF PPS final rule (80 FR 46418), we believe that the risk adjustment model that we have adopted for the SNFRM will ensure that SNFs serving more complex patient populations will not be penalized inadvertently under the Program. As we discussed in the FY 2017 SNF PPS final rule (81 FR 51993), we have also specified the SNF Potentially Preventable Readmissions Measure for the Program, and that measure estimates the risk-standardized rate of unplanned, potentially preventable hospital readmissions for Medicare FFS beneficiaries. The comprehensive claims-based risk-adjustment model that the measure employs takes into account demographic and eligibility characteristics, principal diagnoses, types of surgery or procedure from the prior short-term hospital stay, comorbidities, length of stay and ICU/CCU utilization from the immediately prior short-term hospital stay, and number of admissions in the year preceding the SNF admission. We continue to believe that the measures' risk adjustment methodologies appropriately adjust for factors beyond SNFs' control. We will carefully monitor the Program's effects on SNFs' measured performance and on care quality, and will work with SNFs to provide as much assistance as possible with their efforts to improve on the Program's measures. For additional information on the SNFRM's calculation and methodology, we refer readers to the SNFRM Technical Report available on our Web site at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Downloads/SNFRM-Technical-Report-3252015.pdf. For additional information on the SNFPPR's calculation and methodology, we refer readers to the SNFPPR Technical Report available on our Web site at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Value-Based-Programs/Other-VBPs/SNFPPR-Technical-Report.pdf.

Comment: Two commenters suggested that we consider removing readmissions from the measure when they are associated with events unrelated to SNF care, such as car accidents or disease outbreaks.

Response: We note that the SNF VBP Program's statute requires that the measure specified under section 1888(g)(1) of the Act must be an “all-cause all-condition hospital readmission” measures, which we specified as the SNFRM (NQF #2510). We previously addressed this issue in detail in the FY 2016 SNF PPS Final Rule (80 FR 46412 through 46413). We explained that the SNFRM has been risk adjusted for case-mix to account for differences in patient populations. The goal of risk adjustment is to account for these differences so that providers who treat sicker or more vulnerable patient populations are not unnecessarily penalized for factors that are outside of their control. Regarding hospitalizations due to other incidents unrelated to SNF care such as car accidents and non-preventable disease outbreaks, we note that these events are random and would not be likely to cluster in certain SNFs over time; thus they would not result in systematic bias in the measure.

Comment: One commenter suggested that we factor the expansion of managed care into our measure development process, noting that many states are rapidly expanding managed care offerings for both Medicare and Medicaid patients. The commenter suggested that we consider consolidating quality measure requirements between Medicare and Medicaid to minimize the burden on participating providers, and suggested that we promote best practices in quality improvement as widely as possible.

Response: The measures that we have adopted for the Program are based on Medicare claims, and are thus restricted to Medicare fee-for-service beneficiaries. We believe that policy to be appropriate given the Program's focus on Medicare fee-for-service payments. From our collaboration with the Quality Innovation and Improvement Networks, we also believe that many of the care transitions and quality improvement strategies used by SNFs are broadly applicable to reduce readmissions for Medicaid and managed-care patients. We will consider methods to monitor managed-care performance in the future, and welcome commenters' input on that topic.

Comment: One commenter urged us to refine and test the SNFPPR further before adopting it for the Program. The commenter was also concerned about our use of differing measures within the same service line, noting that the re-hospitalization measure currently in use in the Nursing Home Five-Star Quality Rating differs from the SNFPPR. The commenter stated that our longer-term goal should be to align the SNF VBP measure with other relevant hospitalization measures such as those used in VBP programs developed under Medicaid waivers.

Response: We thank the commenter for the suggestion. We wish to clarify that we are conducting additional testing on the SNFPPR measure, in preparation to submit that measure to NQF for endorsement consideration. We wish to clarify that the re-hospitalization measure reported on Nursing Home Compare is not a measure of potentially preventable readmissions, as required by PAMA. We agree that aligning measures across Programs, when feasible, may reduce provider confusion.

Comment: One commenter discussed the length of the readmission window for both the SNFRM (NQF #2510) and the SNFPPR. The commenter urged us to extend the readmission window to include the entire SNF stay and a set period after discharge from the SNF.

Response: We believe that the length of the readmission windows for the SNFRM and SNFPPR is appropriate because they are harmonized with measures used in the hospital setting. We note also that a longer readmission window, such as 90-days, would make it difficult to ensure that potentially preventable readmissions occurring up to 90 days after prior hospital discharge are attributable to the SNF care received. We refer readers to the FY 2017 SNF PPS Final Rule (81 FR 51993) for additional details concerning the length of the readmission window for SNF VBP Program measures.

We thank commenters for their feedback.

(3) Updates to the Skilled Nursing Facility 30-Day All-Cause Readmission Measure (NQF#2510)

Since finalizing the SNFRM for use in the SNF VBP Program, we have continued to conduct analyses using more recent data, as well as to make some necessary non-substantive measure refinements. Results of this work and all refinements are detailed in a Technical Report Supplement that is available on the following CMS Web site at https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Value-Based-Programs/Other-VBPs/SNF-VBP.html.

We did not receive any public comments on this topic.

(4) Accounting for Social Risk Factors in the SNF VBP Program

We understand that social risk factors such as income, education, race and ethnicity, employment, disability, community resources, and social support (certain factors of which are also sometimes referred to as socioeconomic status (SES) factors or socio-demographic status (SDS) factors) play a major role in health. One of our core objectives is to improve beneficiary outcomes including reducing health disparities, and we want to ensure that all beneficiaries, including those with social risk factors, receive high quality care. In addition, we sought to ensure that the quality of care furnished by providers and suppliers is assessed as fairly as possible under our programs while ensuring that beneficiaries have adequate access to excellent care.

We have been reviewing reports prepared by the Office of the Assistant Secretary for Planning and Evaluation (ASPE) and the National Academies of Sciences, Engineering, and Medicine on the issue of accounting for social risk factors in CMS's value-based purchasing and quality reporting programs, and considering options on how to address the issue in these programs. On December 21, 2016, ASPE submitted a Report to Congress on a study it was required to conduct under section 2(d) of the Improving Medicare Post-Acute Care Transformation (IMPACT) Act of 2014. The study analyzed the effects of certain social risk factors in Medicare beneficiaries on quality measures and measures of resource use used in one or more of nine Medicare value- based purchasing programs, including the SNF VBP Program. The report also included considerations for strategies to account for social risk factors in these programs. In a January 10, 2017 report released by The National Academies of Sciences, Engineering, and Medicine, that body provided various potential methods for measuring and accounting for social risk factors, including stratified public reporting.

As noted in the FY 2017 IPPS/LTCH PPS final rule, the NQF has undertaken a 2-year trial period in which certain new measures, measures undergoing maintenance review, and measures endorsed with the condition that they enter the trial period can be assessed to determine whether risk adjustment for selected social risk factors is appropriate for these measures. This trial entails temporarily allowing inclusion of social risk factors in the risk-adjustment approach for these measures. At the conclusion of the trial, NQF will issue recommendations on the future inclusion of social risk factors in risk adjustment for these quality measures, and we will closely review its findings.

The SNF VBP section of ASPE's report examined the relationship between social risk factors and performance on the 30-day SNF readmission measure for beneficiaries in SNFs. Findings indicated that beneficiaries with social risk factors were more likely to be re-hospitalized but that this effect was significantly smaller when the measure's risk adjustment variables were applied (including adjustment for age, gender, and comorbidities), and that the effect of dual enrollment disappeared. In addition, being at a SNF with a high proportion of beneficiaries with social risk factors was associated with an increased likelihood of readmissions, regardless of a beneficiary's social risk factors.

As we continue to consider the analyses and recommendations from these reports and await the results of the NQF trial on risk adjustment for quality measures, we are continuing to work with stakeholders in this process. As we have previously communicated, we are concerned about holding providers to different standards for the outcomes of their patients with social risk factors because we do not want to mask potential disparities or minimize incentives to improve the outcomes for disadvantaged populations. Keeping this concern in mind, while we sought input on this topic previously, we again sought public comment on whether we should account for social risk factors in the SNF VBP Program, and if so, what method or combination of methods would be most appropriate for accounting for social risk factors. Examples of methods include: Adjustment of the payment adjustment methodology under the SNF VBP Program; adjustment of provider performance scores (for instance, stratifying providers based on the proportion of their patients who are dual eligible); confidential reporting of stratified measure rates to providers; public reporting of stratified measure rates; risk adjustment of measures as appropriate based on data and evidence; and redesigning payment incentives (for instance, rewarding improvement for providers caring for patients with social risk factors or incentivizing providers to achieve health equity). While we consider whether and to what extent we currently have statutory authority to implement one or more of the above-described methods, we sought comments on whether any of these methods should be considered, and if so, which of these methods or combination of methods would best account for social risk factors in the SNF VBP Program.

In addition, we sought public comment on which social risk factors might be most appropriate for stratifying measure scores and/or potential risk adjustment of a particular measure. Examples of social risk factors include, but are not limited to, dual eligibility/low-income subsidy, race and ethnicity, and geographic area of residence. We are seeking comments on which of these factors, including current data sources where this information would be available, could be used alone or in combination, and whether other data should be collected to better capture the effects of social risk. We will take commenters' input into consideration as we continue to assess the appropriateness and feasibility of accounting for social risk factors in the SNF VBP Program. We note that any such changes would be proposed through future notice-and-comment rulemaking.

We look forward to working with stakeholders as we consider the issue of accounting for social risk factors and reducing health disparities in CMS programs. Of note, implementing any of the above methods would be taken into consideration in the context of how this and other CMS programs operate (for example, data submission methods, availability of data, statistical considerations relating to reliability of data calculations, among others), and we also welcome comment on operational considerations. CMS is committed to ensuring that its beneficiaries have access to and receive excellent care, and that the quality of care furnished by providers and suppliers is assessed fairly in CMS programs.

Commenters submitted the following comments related to the proposed rule's discussion of the Accounting for Social Risk Factors in the SNF VBP Program. A discussion of these comments, along with our responses, appears below.

Comment: Many commenters encouraged us to incorporate social risk factors adjustments in various forms, including stratifying providers into peer groups. Commenters stated that we should require measure developers to incorporate SDS data elements testing in risk adjustment models and suggested that we consider adjusting measures for dual-eligible status as well as education level, limited English proficiency, and living alone, among other possible factors. Some commenters suggested that we examine the Program's effects on specialty populations such as children and residents that are ventilator-dependent, patients receiving dialysis, or patients living with HIV/AIDS. Other commenters suggested that we use IMPACT Act measure data to risk-adjust measures and provider performance scores. One commenter suggested that we consider a stratification approach similar to that proposed for the Hospital Readmissions Reduction Program.

Other commenters encouraged us to incorporate into our future policies the findings both from NQF's sociodemographics trial and from ASPE's report. One commenter noted that the ASPE report found that provider-level factors are more powerful predictors of readmissions than beneficiary-level factors, and that high-dual SNFs were among the best performers on the readmission measure examined. The commenter stated that these results alone do not suggest a need for risk adjustment, but suggested again that we examine NQF's results before determining whether or not risk adjustment is appropriate in the Program, and further suggested that incorporating SES variables into the measures' risk-adjustment model could embed health disparities, create biases in reporting, undermine system-based approaches to providing high-quality care, and create care access problems. Another commenter noted that adjusting for social risk factors could negatively affect providers and facilities in regions where social risk factors are higher, but cautioned that adjusting for such factors may increase health disparities by essentially masking them.

One commenter suggested that we consider developing readmission measures or statistical approaches to report quality performance specifically for beneficiaries with social risk factors. The commenter noted that high social risk beneficiaries are substantially more likely to be re-hospitalized, and that beneficiaries at SNFs serving a high proportion of beneficiaries with social risk factors are also more likely to be re-hospitalized. The commenter stated that these findings suggest that the SNFPPR's outcomes could vary significantly due to factors beyond the SNF's control.

Response: We appreciate all the comments and interest in this topic. As we have previously stated, we are concerned about holding providers to different standards for the outcomes of their patients with social risk factors, because we do not want to mask potential disparities or minimize incentives to improve outcomes for disadvantaged populations. We believe that the path forward should incentivize improvements in health outcomes for disadvantaged populations while ensuring that beneficiaries have access to excellent care. We intend to consider all suggestions as we continue to assess each measure and the overall program. We appreciate that some commenters recommended risk adjustment as a strategy to account for social risk factors, while others stated a concern that risk adjustment could minimize incentives and reduce efforts to address disparities for patients with social risk factors. We intend to conduct further analyses on the impact of strategies such as measure-level risk adjustment and stratifying performance scoring to account for social risk factors including the options suggested by commenters. In addition, we appreciate the recommendations from the commenters about consideration of specific social risk factor variables and will work to determine the feasibility of collecting these patient-level variables. As we consider the feasibility of collecting patient-level data and the impact of strategies to account for social risk factors through further analysis, we will continue to evaluate the reporting burden on providers. Future proposals would be made after further research and continued stakeholder engagement.

We thank commenters for their feedback. We will take it into account in future rulemaking.

c. FY 2020 Performance Standards

We refer readers to the FY 2017 SNF PPS final rule (81 FR 51995 through 51998) for a summary of the statutory provisions governing performance standards under the SNF VBP Program and our finalized performance standards policy, as well as the numerical values for the achievement threshold and benchmark for the FY 2019 program year. We also responded to public comments on these policies in that final rule.

In the proposed rule (82 FR 21081 through 21802), we proposed estimated performance standards for the FY 2020 SNF VBP Program based on the FY 2016 MedPAR files including a 3-month run-out period. We stated our intention to include the final numerical values of the performance standards in the final rule. We have displayed the estimated performance standards' numerical values from the proposed rule in Table 23. As we have done previously, we have inverted the SNFRM rates in Table 23 so that higher values represent better performance.

We sought public comments on these estimated achievement threshold and benchmark values. A discussion of these comments, along with our responses, appears below.

Comment: One commenter supported our performance standards methodology in general. The commenter was concerned, however, that continually rewarding lower readmission rates may not be in the best interests of SNF patients. The commenter suggested that we explore identifying an optimal readmission rate.

Response: Our statistically based benchmark is intended to set an empirically based performance standard of top performing SNFs as an achievable goal for all SNFs during the performance period. We recognize that this benchmark might not be an optimal readmission rate as suggested by the commenter due to performance gaps between current and optimal care, but the intent of the Program's incentives is to encourage SNFs to improve the care they provide. We also caution that establishing a single optimal readmission rate may not be feasible for a nationwide quality program affecting care for millions of Medicare beneficiaries. We intend to carefully monitor the Program's effects on readmission rates and on care quality, and if warranted, will revisit the performance standards methodology in future rulemaking.

In this final rule, we are providing the finalized numerical values of the achievement threshold and the benchmark for the FY 2020 program year. We note that the values have not changed since we published the proposed rule.

Additionally, as discussed further below, we are finalizing baseline and performance periods for the FY 2020 program year based on the federal fiscal year rather than the calendar year as we had finalized for the FY 2019 program year. The numerical values for the achievement threshold and benchmark in Table 24 reflect this final policy by using FY 2016 claims data. As we have done in prior rulemaking, we have inverted the SNFRM rates in Table 24 so that higher values represent better performance.

After consideration of the public comments that we received, we are finalizing the performance standards for the FY 2020 SNF VBP Program as proposed.

d. FY 2020 Performance Period and Baseline Period

(1) Background

We refer readers to the FY 2016 SNF PPS final rule (80 FR 46422) for a discussion of the considerations that we took into account when specifying performance periods for the SNF VBP Program. Based on those considerations, as well as public comments received, we adopted CY 2017 as the performance period for the FY 2019 SNF VBP Program, with a corresponding baseline period of CY 2015.

(2) FY 2020 Policies

As we stated in the proposed rule (82 FR 21082), we continue to believe that a 12-month performance and baseline period are appropriate for the Program, and we are concerned about the operational challenges of linking the 12-month periods to the calendar year. Specifically, the allowance of an approximately 90-day claims run out period following the last date of discharge, coupled with the length of time needed to calculate the measure rates using multiple sources of claims needed for statistical modeling, determine achievement and improvement scores, allow SNFs to review their measure rates, and determine the amount of payment adjustments could risk delay in meeting requirement at section 1888(h)(7) of the Act to notify SNFs of their value-based incentive payment percentages not later than 60 days prior to the fiscal year involved.

We therefore considered what policy options we had to mitigate this risk and ensure that we comply with the statutory deadline to notify SNFs of their payment adjustments under the Program.

We continue to believe that a 12-month performance and baseline period provide a sufficiently reliable and valid data set for the SNF VBP Program. We also continue to believe that, where possible and practicable, the baseline and performance period should be aligned in length and in months included in the selections. Taking those considerations and beliefs into account, we proposed to adopt FY 2018 (October 1, 2017, through September 30, 2018) as the performance period for the FY 2020 SNF VBP Program, with FY 2016 (October 1, 2015, through September 30, 2016) as the baseline period for purposes of calculating performance standards and measuring improvement. We noted that this proposed policy, would, if finalized, give us an additional 3 months between the conclusion of the performance period and the 60-day notification deadline prescribed by section 1888(h)(7) of the Act to complete the activities described above.

We are aware that making this transition from the calendar year to the FY will result in our measuring SNFs on their performance during Q4 of 2017 (October 1, 2017, through December 31, 2017) for both the FY 2019 program year and the FY 2020 program year. During the FY 2019 program year, that quarter will fall at the end of the finalized performance period (January 1, 2017, through December 31, 2017), while during the FY 2020 program year, that quarter will fall at the beginning of the proposed performance period (October 1, 2017, through September 30, 2018). We believe that, on balance, this overlap in data is more beneficial than the alternative. We considered proposing not to use that quarter of measured performance during the FY 2020 program year, but, as a result, we would be left with fewer than 12 months of data with which to score SNFs under the program. As we have stated, we believe it is important to use 12 months of data to avoid seasonality issues and to assess SNFs fairly. We therefore believe that meeting these operational challenges, in total, outweighs any cost to SNFs associated with including a single quarter's SNFRM data in their SNF performance scores twice.

However, as an alternative, we requested comments on whether or not we should instead consider adopting for the FY 2020 Program a one-time, three-quarter performance period of January 1, 2018, through September 30, 2018, and a one-time, three-quarter baseline period of January 1, 2016 through September 30, 2016 to avoid the overlap in performance period quarters that we describe above. We believe this option could provide us with sufficiently reliable SNFRM data for purposes of the Program's scoring while ensuring that SNFs are not scored on the same quality measure data in successive Program years. However, we noted that the shorter measurement period could result in lower denominator counts and seasonal variations in care, as well as disparate effects of cold weather months on SNFs' care could also create variations in quality measurement, and could potentially disproportionately affect SNFs in different areas of the country. Under this alternative, we would resume a 12-month performance and baseline period beginning with the FY 2021 program year.

We sought public comments on our proposal and alternative. In addition, as we continue considering potential policy changes once we replace the SNFRM with the SNFPPR, we also sought comment on whether we should consider other potential performance and baseline periods for that measure. We specifically sought comments on whether we should attempt to align the SNF VBP Program's performance and baseline periods with other CMS value-based purchasing programs, such as the Hospital VBP Program or Hospital Readmissions Reduction Program, which could mean proposing to adopt performance and baseline periods that run from July 1st to June 30th. A discussion of these comments, along with our responses, appears below.

Comment: Some commenters supported our proposed performance and baseline periods for the FY 2020 Program, acknowledging that the one-quarter overlap may be unavoidable and agreeing with us that a three-quarter performance period would not be appropriate. Commenters also stated that it is not necessary to align the SNF VBP Program's performance periods with other VBP programs.

Response: We thank the commenters for their support and feedback.

Comment: Some commenters expressed concern about the SNF VBP Program's shift from calendar year to fiscal year measurement periods while the SNF QRP has proposed the reverse. Commenters were concerned that this lack of alignment between the two programs could be confusing for providers.

Response: As described above, the SNF VBP Program's shift from calendar year to fiscal year measurement periods is logistically necessary to meet the statutory deadlines for the program. CMS will take all necessary steps to minimize any potential confusion among providers.

Comment: One commenter opposed our proposal to maintain 12-month performance and baseline periods while shifting to fiscal year reporting periods, and stated that we should instead use a one-time three-quarter baseline and performance period for the FY 2020 Program year. Another commenter recommended that we use only 9 months for the performance and baseline periods for FY 2019 and FY 2020, and then beginning with FY 2021, consider aligning the reporting periods to other VBP programs that run from July 1 to June 30 of each year. The commenter noted that making this change would result in a six-month overlap as opposed to the 3-month overlap under the proposal, with the result being that the change would occur over 2 years.

Response: We thank the commenters for this feedback. However, as we described in the proposed rule, we are concerned that a shorter performance period than a 12-month period could result in lower denominator counts and seasonal variations in care, which could disproportionately affect SNFs in different regions of the country. Our analysis of 9 and 12 month SNFRM denominator size reveals that these issues are sufficiently mitigated by the commenters' suggestion, and we continue to believe that a one-quarter overlap in performance periods between FY 2019 and FY 2020 is an acceptable compromise to make this transition to performance and baseline periods centered on the federal fiscal year.

Additionally, we believe that using a full year of claims data to calculate performance on the measures ensures that the variation found among SNF performance is due to real differences in care delivery between SNFs, and not within-facility variation due to issues such as seasonality. Based on our SNFRM denominator analysis, we do not believe that using a 9-month performance period would provide us with sufficiently reliable data for a performance year, and given the Program's focus on a single quality measure, we do not believe scoring insufficiently reliable quality measure data to be a practical policy.

After consideration of the public comments that we have received, we are finalizing the performance and baseline period for the FY 2020 SNF VBP Program as proposed.

e. SNF VBP Performance Scoring

We refer readers to the FY 2017 SNF PPS final rule (81 FR 52000 through 52005) for a detailed discussion of the scoring methodology that we have finalized for the Program, along with responses to public comments on our policies and examples of scoring calculations.

(1) Rounding Clarification for SNF VBP Scoring

In the FY 2017 SNF PPS final rule (81 FR 52001), we adopted formulas for scoring SNFs on achievement and improvement. The final step in these calculations is rounding the scores to the nearest whole number.

As we have continued examining SNFRM data, we have identified a concern related to that rounding step. Specifically, we are concerned that rounding SNF performance scores to the nearest whole number is insufficiently precise for purposes of establishing value-based incentive payments under the Program. Rounding scores in this manner has the effect of producing significant numbers of tie scores, since SNFs have between 0 and 100 points available under the Program, and we estimate that more than 15,000 SNFs will participate in the Program. As discussed further in this section, the exchange function methodology that we proposed to adopt is most easily implemented when we are able to differentiate precisely among SNF performance scores to provide each SNF with a unique value-based incentive payment percentage.

We therefore proposed to change the rounding policy from that previously finalized for SNF VBP Program scoring methodology, and instead to award points to SNFs using the formulas that we adopted in last year's rule by rounding the results to the nearest ten-thousandth of a point. Using significant digits terminology, we proposed to use no more than five significant digits to the right of the decimal point when calculating SNF performance scores and subsequently calculating value-based incentive payments.

We view this policy change as necessary to ensure that the Program scores SNFs as precisely as possible and to ensure that value-based incentive payments reflect SNF performance scores as accurately as possible.

We sought public comments on this proposal. A discussion of these comments, along with our responses, appears below.

Comment: Some commenters supported our proposal to round SNF performance scores to the fifth significant digit, noting that the step is necessary to avoid ties and that it will have only minor financial impacts.

Response: We thank the commenters for their support.

Comment: Several commenters cautioned that we should not implement policy changes merely to ensure more differentiation among providers.

Response: We thank the commenters for their support. We agree with the commenters that we should not implement policy changes solely to ensure more differentiation, but we view this policy as necessary in order to ensure that SNF performance scores are accurate. We will also consider this caution as we adopt policies in future rulemaking.

Comment: One commenter opposed our proposal to round SNF performance scores to the nearest ten-thousandth of a point, stating that scoring in this manner is “too narrow.” The commenter recommended instead that we round scores to the nearest tenth of a point.

Response: We thank the commenter for this feedback, but we believe that rounding scores to the nearest tenth of a point would still result in numerous scoring ties due to the estimated 15,000 SNFs that will participate in the Program. We believe that the rounding policy we have proposed ensures that we have sufficient precision to calculate performance scores under the program.

Comment: One commenter suggested that if our proposed change to the rounding policy for SNF performance scores results in SNFs with nearly identical readmission rates receiving materially different VBP payment amounts, we should consider revising the methodology.

Response: We thank the commenter and agree. Our expectation is that the additional precision will not significantly affect SNFs' payment amounts when they have nearly identical SNF performance scores, but we will monitor this issue carefully.

After consideration of the public comments that we have received, we are finalizing that we will round the SNF performance scores to the fifth significant digit.

(2) Policies for Facilities With Zero Readmissions During the Performance Period

In our analyses of historical SNFRM data, we identified a unit imputation issue associated with certain SNFs' measured performance. Specifically, we found that a small number of facilities had zero readmissions during the applicable performance period. An observed readmission rate of zero is a desirable outcome; however, due to risk-adjustment and the statistical approach used to calculate the measure, outlier values are shifted towards the mean, particularly for smaller SNFs. As a result, observed readmission rates of zero result in risk-standardized readmission rates that are greater than zero. Analysis conducted by our measure development contractor revealed that it may be possible—although rare—for SNFs with zero readmissions to receive a negative value-based incentive payment adjustment. We are concerned that assigning a net negative value-based incentive payment to a SNF that achieved zero readmissions during the applicable performance period would not support the Program's goals.

We considered our policy options for SNFs that could be affected by this issue, including excluding SNFs with zero readmissions from the Program entirely to ensure that they are not unduly harmed by being assigned a non-zero RSRR by the measure's finalized methodology. However, because the Program's statute requires us to include all SNFs in the Program, we do not believe we have the authority to exclude any SNFs from the payment withhold and from value-based incentive payments. We also considered proposing to replace SNF performance scores for those SNFs in this situation with the median SNF performance score. But because we must pay SNFs ranked in the lowest 40 percent less than the amount they would otherwise be paid in the absence of the SNF VBP, we do not believe that assigning these SNFs the median performance rate on the applicable measure would necessarily protect them from receiving net negative value-based incentive payments.

We are considering different policy options to ensure that SNFs achieving zero readmissions among their patient populations during the performance period do not receive a negative payment adjustment. We intend to address this topic in future rulemaking, and we request public comments on what accommodations, if any, we should employ to ensure that SNFs meeting our quality goals are not penalized under the Program. We specifically sought comments on the form this potential accommodation should take. A discussion of these comments, along with our responses, appears below.

Comment: Some commenters expressed concerns about the risk adjustment methodology employed to calculate the measures, particularly for SNFs with zero readmissions during the applicable period. Commenters noted that the statistical approach employed by the measures means that SNFs with low volume or zero readmissions during the applicable period could receive a worse risk-standardized readmission rate, which could hide true differences in performance and may dampen SNFs' incentives to improve. Commenters suggested that we consider expanding the performance periods for SNFs with low volume to mitigate these effects. Other commenters suggested that we consider returning the full 2 percentage points withheld from SNFs' Medicare payments when those SNFs have zero readmissions during the applicable period, provide a rolling average readmission rate, or stratify readmission rates and value-based incentive payments by facility size.

Response: We intend to address this topic in future rulemaking, and will take these suggestions into account at that time.

Comment: One commenter believed that we should develop an exceptions policy for SNFs in special circumstances, and recommended that under this policy, we return affected SNFs' entire payment withhold and not assign public rankings or scores. The commenter recommended that we offer this exception to SNFs based on a small denominator size of fewer than 25 cases rather than zero readmissions. The commenter noted that a small denominator size would likely capture SNFs with zero readmissions and would ensure that low-volume SNFs do not stack at the top of the Program's ranking and harm non-zero denominator facilities' standing.

Response: We thank the commenter for this feedback and will take it into account in future rulemaking.

We thank the commenters for their feedback, and will take it into account in the future.

(3) Request for Comments on Extraordinary Circumstances Exception Policy

In other value-based purchasing programs, such as the Hospital VBP Program (see 78 FR 50704 through 50706), as well as several of our quality reporting programs, we have adopted Extraordinary Circumstances Exceptions policies intended to allow participating facilities to receive administrative relief from program requirements due to natural disasters or other circumstances beyond the facility's control that may affect the facility's ability to provide high-quality health care.

We are considering whether this type of policy would be appropriate for the SNF VBP Program. We intend to address this topic in future rulemaking. We therefore sought public comments on whether we should implement such a policy, and if so, the form the policy should take. If we propose such a policy in the future, our preference would be to align it with the Extraordinary Circumstances Exception policy adopted under our other quality programs. A summary of the public comments that we received, along with our responses, appears below.

Comment: Some commenters stated their belief that we should adopt an Extraordinary Circumstances Exception policy to provide administrative relief to SNFs suffering from circumstances beyond their control, and recommended that we align the policy with the Hospital VBP Program. Other commenters suggested that we consider adopting the same exception process as has been adopted under the SNF QRP.

Response: We thank the commenters for their suggestions, and will take it into consideration if we decide to propose an Extraordinary Circumstances Exception policy in future rulemaking.

f. SNF Value-Based Incentive Payments

(1) Exchange Function

We refer readers to the FY 2017 SNF PPS final rule (81 FR 52005 through 52006) for discussion of four possible exchange functions that we considered adopting to translate SNFs' performance scores into value-based incentive payments. We created new graphical representations of the four functions that we have considered in the past—linear, cube, cube root, and logistic—and presented those updated representations in the proposed rule (82 FR 21084). We noted that the actual exchange functions' forms and slopes will vary depending on the distributions of SNFs' performance scores from the FY 2019 performance period, and wished to emphasize that these representations are presented solely for the reader's clarity as we discussed our exchange function policy.

We have continued examining historical SNFRM data while considering our policy options for this program. We have attempted to assess how each of the four possible exchange functions that we set out in the FY 2017 SNF PPS final rule, as well as potential variations, would affect SNFs' incentive payments under the Program. We specifically considered the effects of the statutory constraints on the Program's value-based incentive payments and our belief that to create an effective incentive payment program, SNFs' value-based incentive payments must be widely distributed to reward higher performing SNFs through increased payment and to make reduced payments to lower performing SNFs. We also considered our desire to avoid unintended consequences of the Program's incentive payments, particularly since the Program is limited by statute to using a single measure at a time, and our view that an equitable distribution of value-based incentive payments would be most appropriate to ensure that all SNFs, including SNFs serving at-risk populations, could potentially qualify for incentive payments.

In our view, important factors when adopting an exchange function include the number of SNFs that receive more in value-based incentive payments than the number of SNFs for which a reduction is applied to their Medicare payments, as well as the incentive for SNFs to reduce hospital readmissions. We hold this view because we believe that the Program will be most effective at encouraging SNFs to improve the quality of care that they provide to Medicare beneficiaries if SNFs have the opportunity to earn incentives, rather than simply avoid penalties, through high performance on the applicable quality measure. We also believe that SNFs must have incentives to reduce hospital readmissions for their patients no matter where their performance lies in comparison to their peers.

Taking those considerations into account, we analyzed the four exchange functions on which we have previously sought comment—linear, cube, cube root, and logistic—as well as variations of those exchange functions. We scored SNFs using historical SNFRM data and modeled SNFs' value-based incentive payments using each of the functions in turn. We evaluated the distribution of value-based incentive payments that resulted from each function, as well as the number of SNFs with positive payment adjustments and the value-based incentive payment percentages that resulted from each function. We also evaluated the functions' results for the statutory requirements in section 1888(h)(5)(C)(ii) of the Act, including the requirements in subclause (I) that the percentage be based on the SNF performance score for each SNF, in subclause (II) that the application of all such percentages results in an appropriate distribution, and in items (aa), (bb), and (cc) of subclause (II), specifying that SNFs with the highest rankings receive the highest value-based incentive payment amounts, that SNFs with the lowest rankings receive the lowest value-based incentive payment amounts, and that the SNFs in the lowest 40 percent of the ranking receive a lower payment rate than would otherwise apply.

In our analyses of the four baseline functions, we found that the logistic function maximized the number of SNFs with positive payment adjustments among SNFs measured using the SNFRM. We also found that the logistic function best fulfills the requirement that the SNFs in the lowest 40 percent of the ranking receive a lower payment rate than would otherwise apply, resulted in an appropriate distribution of value-based incentive payment percentages, and fulfilled the other statutory requirements described in this final rule. Specifically, we noted that the logistic function provided a broad range of SNFs with net-positive value-based incentive payments, and while it did not provide the highest value-based incentive payment percentage to the top performers of all the functions, we viewed the number of SNFs with positive payment adjustments as a more important consideration than the highest value-based incentive payment percentages being awarded.

We also considered alignment of VBP payment methodologies across fee-for-service Medicare VBP programs, including the Hospital VBP program and Quality Payment Program (QPP). We recognize that aligning payment methodologies would help stakeholders that use VBP payment information across care settings better understand the SNF VBP payment methodology. Both the Hospital VBP program and QPP use some form of a linear exchange function for payment. Three key program aspects that facilitate the use of a linear exchange function are the programs' number of measures, measure weights, and correlation across program measures. These three aspects in tandem contribute to the approximately normal distribution of scores expected in the Hospital VBP program and QPP. No single measure is the key driver that might “tilt” scores to a non-normal distribution. Since both programs are required to be budget neutral, our modeling estimates that scores translate into an approximately equal number of providers with positive payment adjustments and providers receiving a net payment reduction.

In contrast, the SNF VBP payment adjustment is driven, in part, by two specific SNF VBP statutory requirements: The program's use of a single measure; and the requirement that the total amount of value-based incentive payments for all SNFs in a fiscal year be between 50 and 70 percent of the total amount of reductions to payments for that fiscal year, as estimated by the Secretary. Our analysis of the linear exchange function showed that more SNFs would receive a net payment reduction than a payment incentive because the total amount available for incentive payments in a fiscal year is limited to between 50 and 70 percent of the total amount of the reduction to SNF payments for that fiscal year. The linear exchange function also results in the provision of a net payment reduction to a higher percentage of SNFs that exceeded the 50th percentile of national performance, relative to the logistic payment function. We believe that these findings are unique to the SNF VBP program, relative to other fee-for-service Medicare programs, because of the limitation on the total amount that we can use for incentive payments, coupled with the use of a single measure and the corresponding scoring distribution.

In addition to the four baseline functions described further above, we considered adjusting the linear function to be able to make positive payment adjustments to a greater number of SNFs. Specifically, we tested an alternative where we reduced the baseline linear function by 20 percent, then redistributed the resulting funds to the middle 40 percent of SNFs. We found that the use of this linear function with adjustment would enable us to make a positive payment adjustment to a slightly greater number of SNFs than we would be able to make using the logistic function. However, we were concerned with the additional complexity involved in implementing this type of two-step adjustment to the linear exchange function.

Taking all of these considerations into account, we proposed to adopt a logistic function for the FY 2019 SNF VBP Program and subsequent years. Under this policy, we would:

1. Estimate Medicare spending on SNF services for the FY 2019 payment year;

2. Estimate the total amount of reductions to SNFs' adjusted Federal per diem rates for that year, as required by statute;

3. Calculate the amount realized under the payback percentage policy (discussed further below);

4. Order SNFs by their SNF performance scores; and

5. Assign a value-based incentive payment multiplier to each SNF that corresponds to a point on the logistic exchange function that corresponds to its SNF performance score.

As we discussed in the proposed rule (82 FR 21085), we would model the logistic exchange function in such a form that the estimated total amount of value-based incentive payments equals not more than 60 percent of the amounts withheld from SNFs' claims. While the function's specific form would also depend on the distribution of SNF performance scores during the performance period, the formula that we used to construct the logistic exchange function and that we proposed to use for FY 2019 program calculations is:

where x _i is the SNF's performance score.

We sought public comments on this proposal, and in particular, on whether a linear function with adjustment would alternatively be feasible for the SNF VBP Program, potentially beginning with FY 2019. A discussion of these comments, along with our responses, appears below.

Comment: Some commenters supported the logistic exchange function, agreeing that it best incentivizes SNFs to improve continuously and allows for the greatest number of SNFs to receive net-positive payments. The commenters also agreed that the linear function with adjustment could create confusion, and requested that we provide an example calculation of a provider's payment multiplier in the final rule.

Response: We thank the commenters for their support and feedback. In response to the commenters' request for an example, we can provide two hypothetical examples of SNFs' performance scores based on historical performance data and historical Medicare spending that would be subject to the Program. We would like to emphasize that the actual multipliers that will result from the calculation of the logistic exchange function for the FY 2019 Program year will depend on the distribution of SNF performance scores that result from the performance period as well as estimated Medicare spending subject to the Program for the FY 2019 payment year, and thus SNFs should not expect to receive the example multipliers below if their FY 2019 SNF performance scores approximate either of these examples.

A SNF with a baseline period SNFRM rate of 0.16980, which inverts to 0.83020, and a performance period SNFRM rate of 0.19989, which inverts to 0.80011, would, according to the formulas that we have adopted in previous regulations, receive 20.56057 points for achievement and 0 points for improvement since its measured performance declined. The higher of those two values is 20.56057, and that value would become the SNF's performance score. Based on the distribution of historical performance in the data sets that we analyzed, that SNF performance score translates into a value-based incentive payment multiplier of 0.150052 percent, which would be applied after the application of the 2% reduction required by section 1888(h)(6)(B).

Conversely, a SNF with a baseline period SNFRM rate of 0.18842, which inverts to 0.81158, and a performance period SNFRM rate of 0.17384, which inverts to 0.82616, would, according to the formulas that we have adopted in previous regulations, receive 70.23616 points for achievement and 4.78908 points for improvement. The higher of those two values is 70.23616, and that value would become the SNF's performance score. Based on the distribution of historical performance in the data sets that we analyzed, that SNF performance score translates into a value-based incentive payment multiplier of 2.64944 percent, which would be applied after the application of the 2 percent reduction required by section 1888(h)(6)(B) of the Act.

Comment: Two commenters requested additional details on the analyses that we conducted to reach the proposed policy, and also requested that we detail how the future transition to the SNFPPR would influence the distribution of incentive payments. One commenter suggested that we perform a “dry run” with the proposed methodology and provide confidential feedback reports to SNFs with the results.

Response: We thank the commenters for this feedback. We will consider providing a dry run or other additional information prior to the planned summer 2018 dissemination of Fiscal Year 2019 payment reports that will notify SNFs of the adjustments to their Medicare payments as required by section 1888(h)(7) of the Act. We also wish to inform the commenters that SNFs received confidential feedback reports with their calendar year 2015 baseline period readmission rates, as captured by the SNFRM, in early 2017. We continue to analyze the potential effects of the Program's transition to the SNFPPR, and we intend to provide additional details on the resulting distribution of value-based incentive payments in the future.

Comment: One commenter requested that we provide a scaling factor that we would use to ensure that payouts equate to 60 percent of the total amount withheld from SNFs' Medicare payments. The commenter also recommended that we not consider the cube exchange function, noting that it would result in extremely high payouts to top providers who may be outliers, and suggested that we provide the slope of each alternative function listed in the rule.

Response: We thank the commenter for the feedback on the exchange function form, and we agree with the commenter that the cube function results in an undesirable distribution of incentive payments to SNFs. As discussed further below, we are finalizing the logistic exchange function for the FY 2019 Program.

In response to the commenter's request that we provide the scaling factor that we would use to ensure that value-based incentive payments under the Program equal the 60 percent payback percentage that we proposed and are finalizing in this final rule, we note that the distribution of incentive payments provided under the Program depends entirely on the distribution of SNFs' performance on the applicable measure during the baseline and performance periods. We are unable to provide a scaling factor for the FY 2019 program year at this time because the performance period (CY 2017) has not concluded yet, though we may consider doing so after the performance period has concluded. We intend to provide additional detail on the distribution of SNF performance scores and the resulting value-based incentive payment percentages, potentially including the scaling factor, in the future.

After consideration of the public comments that we have received, we are finalizing the logistic exchange function as proposed.

(2) Payback Percentage

Section 1888(h)(6)(A) of the Act requires the Secretary to reduce the adjusted federal per diem rate determined under section 1888(e)(4)(G) of the Act otherwise applicable to a SNF for services furnished by that SNF during a fiscal year by the applicable percent (which, under section 1888(h)(6)(B) of the Act is 2 percent for FY 2019 and succeeding fiscal years) to fund the value-based incentive payments for that fiscal year. Section 1888(h)(5)(C)(ii)(III) of the Act further specifies that the total amount of value-based incentive payments under the Program for all SNFs in a fiscal year must be greater than or equal to 50 percent, but not greater than 70 percent, of the total amount of the reductions to payments for that fiscal year under the Program, as estimated by the Secretary. Thus, we must decide what percentage of the total amount of the reductions to payments for a fiscal year we will pay as value-based incentive payments to SNFs based on their performance under the Program for that fiscal year.

As with our exchange function policy described in this final rule, we view the important factors when specifying a payback percentage to be the number of SNFs that receive a positive payment adjustment, the marginal incentives for all SNFs to reduce hospital readmissions and make broad-based care quality improvements, and the Medicare Program's long-term sustainability through the additional estimated Medicare trust fund savings. We intend for the proposed payback percentage to appropriately balance these factors. We analyzed the distribution of value-based incentive payments using historical data, focusing on the full range of available payback percentages.

Taking these considerations into account, we proposed that the total amount of funds that would be available to pay as value-based incentive payments in a fiscal year would be 60 percent of the reductions to payments otherwise applicable to SNF Medicare payments for that fiscal year, as estimated by the Secretary. We believe that 60 percent is the most appropriate payback percentage to balance the considerations described in the proposed rule.

We noted that we intend to closely monitor the effects of the payback percentage policy on Medicare beneficiaries, on participating SNFs, and on their measured performance. We also stated that we intend to consider proposing to adjust the payback percentage in future rulemaking. In our consideration, we would include the Program's effects on readmission rates, potential unintended consequences of SNF care to beneficiaries included in the measure, and SNF profit margins. Since the SNF VBP Program is a new, single measure value-based purchasing program and will continue to evolve as we implement it—including, for example, changing from the SNFRM to the SNFPPR as required by statute—we stated that we intend to evaluate its effects carefully.

We noted also that the Medicare Payment Advisory Commission's research has shown that for-profit SNFs' average Medicare margins are significantly positive, though not-for-profit SNFs' average Medicare margins are substantially lower, and we requested comment on the extent to which that should be considered in our policy. We also recognized that there is some evidence that not-for-profit SNFs tend to perform better on measures of hospital readmissions than for-profit SNFs, and we requested comment on whether our proposed payback percentage appropriately balances Medicare's long-term sustainability with the need to provide strong incentives for quality improvement to top-performing but lower-margin SNFs.

We sought public comments on this proposal. A discussion of these comments, along with our responses, appears below.

Comment: Several commenters recommended that we finalize a 70 percent payback percentage, stating that the largest possible incentive pool will have a larger impact on changing practices and will provide a softer landing for participating providers. Commenters were also concerned that the actual payback percentage may be different than 60 percent if our forecast turns out to be erroneous, and suggested that we instead calculate confidence intervals around the payback percentage.

Other commenters stated that the greatest percentage of dollars should be made available to facilities that invest in their staffs and are therefore top performers, noting also that MedPAC analysis shows that top performers are not enjoying large margins on their Medicare business, and that a larger incentive pool would provide more incentive dollars to high-performing SNFs. Commenters also stated that the Medicare Trust Fund will benefit from reduced hospital spending resulting from lower readmission rates.

Some commenters recommended that we adopt a 70 percent payback percentage and that we use the other 30 percent of amounts withheld from SNFs' Medicare payments to fund quality improvement initiatives. One commenter cited the reduction to SNF PPS rates to fund physician payments, significant MDS changes that will drive staffing and training costs, and the possible revamping of the RUG methodology, as rationale for selecting the maximum possible payback percentage under the Program. The commenter stated that these changes mean that CMS should not make any additional funding reductions beyond those absolutely required.

Response: We thank the commenters for this feedback. Section 1888(h)(5)(C)(ii)(III) of the Act provides that the total amount of value-based incentive payments for all skilled nursing facilities in a fiscal year must be greater than or equal to 50 percent, but not greater than 70 percent of the total amount of the reductions to SNFs' Medicare payments for that fiscal year, as estimated by the Secretary. We are confident that our payback percentage can be implemented accurately, based on our experience estimating the total amount available for value-based incentive payments under the Hospital Value Based Purchasing program. We intend to utilize a similar methodology for the SNF VBP Program by using the most currently available historic SNF claims to estimate the pool of available funds, the finalized payback percentage and corresponding withhold percentage, and the finalized payment exchange function. It is important to note that the 50 to 70 percent range is based on national Medicare spending using the entire population of about 15,000 SNF claims data, and that large data set means that we are able to estimate the payment exchange function that applies the finalized withhold and payback percentage with a high degree of accuracy.

In response to comments that we finalize 70 percent as the payback percentage for the Program, we intended for the proposed payback percentage to balance several policy considerations, including the number of SNFs that receive a positive payment adjustment, the marginal incentives for SNFs to reduce hospital readmissions and make broad-based care quality improvements, and the long-term financial sustainability of the Medicare Program. We do not believe that finalizing a 70 percent payback percentage appropriately balances those factors, particularly the Medicare Program's long-term sustainability, because it results in significantly higher Medicare spending under the Program in a provider sector already experiencing significantly positive Medicare margins. We believe that the other policies we are finalizing in this final rule, including the logistic exchange function, ensure that we provide strong incentives for quality improvement to SNFs within the constraints imposed by the SNF VBP Program's statute.

We intend to carefully monitor the Program's effects on SNFs' care quality improvement efforts and providers' Medicare margins. We would also like to clarify that the savings realized from the Program (that is, the 30 to 50 percent of the amounts withheld from SNFs' claims) are not authorized to be distributed separately for quality improvement initiatives, and are instead retained in the Medicare Trust Fund and used for other Medicare Program purposes authorized by statute.

Comment: One commenter stated that it is unnecessary to adjust the payback percentage based on facility ownership type, stating that the data do not support differential treatment among SNFs.

Response: We thank the commenter for this feedback. However, we would like to clarify that we did not propose to adjust the payback percentage based on facility ownership type. We will monitor the Program's effects on SNFs carefully.

Comment: Two commenters requested that we provide additional information regarding the empirical modeling used to inform our proposed policies, including the proposed 60 percent payback percentage. The commenters stated that the explanations we provided in the proposed rule do not provide sufficient transparency into our decision-making.

Response: We believe that we released sufficient information in the proposed rule to give commenters enough information to submit meaningful comments on our selection of the 60 percent payback proposal, including the considerations that we took into account when developing our proposed policy (82 FR 21086) and the detailed analytical results that we presented in the proposed rule's regulatory impact analysis (82 FR 21094 through 21095). However, we are in the process of compiling additional empirical modeling information and intend to make that information available to the public on the CMS.gov Web site no later than November 2017.

Comment: One commenter stated that CMS should redistribute the full amount withheld from SNFs' claims in incentive payments rather than 50 to 70 percent. The commenter also stated that the requirement that the bottom 40 percent of SNFs not be eligible for incentive payments is unfair, and requested that we provide details on the funds not being redistributed to SNFs.

Response: We thank the commenter for this feedback. However, the requirements that the total amount available for value-based incentive payments in a fiscal year be greater than or equal to 50 percent, but not greater than 70 percent, as well as the requirement that the SNFs ranked in the lowest 40 percent receive a payment rate for services furnished during a fiscal year that less than the payment rate they would have received otherwise for that fiscal year, are statutory in origin. As a result, we do not believe we have the discretion to redistribute the full amount withheld from SNFs' claims as incentive payments or to pay SNFs in the bottom 40 percent the same or a higher rate than they would have otherwise received in the absence of the Program.

In response to the commenter's question about funds not being redistributed to SNFs (that is, the 30 to 50 percent of SNFs' Medicare payments remaining after the payment withhold is determined), as we stated above, those funds are not authorized to be distributed separately for quality improvement initiatives, and are instead retained in the Medicare Trust Fund and used for other Medicare Program purposes authorized by statute.

Comment: Commenter agreed in general with our view that the Program will be most effective if it offers incentive payments to SNFs rather than payment penalties.

Response: We believe that the policies we are finalizing in this final rule, including the payback percentage and the use of the logistical exchange function, will enable us to offer incentive payments to a broad number of SNFs while balancing that consideration with the Medicare Program's long-term sustainability.

After consideration of the public comments that we received, we are finalizing the payback percentage for the FY 2019 SNF VBP program as 60 percent of the total amount of the reduction to SNFs' Medicare payments for that fiscal year, as estimated by the Secretary. We will set the exchange function such that we remit 60 percent of the estimated total amount withheld from SNFs' Medicare payments as value-based incentive payments, though each individual SNF's value-based incentive payment percentage will vary according to its SNF performance score.

g. SNF VBP Reporting

(1) Confidential Feedback Reports

We refer readers to the FY 2017 SNF PPS final rule (81 FR 52006 through 52007) for discussion of our intention to use the QIES system CASPER files to fulfill the requirement in section 1888(g)(5) of the Act that we provide quarterly confidential feedback reports to SNFs on their performance on the Program's measures. We also responded in that final rule to public comments on the appropriateness of the QIES system.

We provided SNFs with a test report in September 2016, followed by data on SNFs' CY 2013 performance on the SNFRM in December 2016 and SNFs' CY 2014 performance on the SNFRM in March 2017. We then provided SNFs with their CY 2015 performance on the SNFRM in June 2017, along with a supplemental workbook providing patient-level data. We intend to continue providing SNFs with their performance data each quarter as required by the statute.

We sought feedback from SNFs on the contents of the quarterly reports and what additional elements, if any, we should consider including that would be useful for quality improvement efforts. We specifically sought comment on what patient-level data would be most helpful to SNFs if they were to request such data from us as part of their quality improvement efforts. A discussion of these comments, along with our responses, appears below.

Comment: Several commenters expressed their view that specific facility-level and patient-level data elements should be provided in quarterly confidential feedback reports. Other commenters expressed support for both the facility level and patient identifiers that we are providing. One commenter suggested that dual eligibility status for patients be provided in quarterly confidential feedback reports. Another commenter requested that we provide additional information in our quarterly confidential feedback reports, including national benchmarks used to calculate achievement and improvement scores, peer ranking information, and SNF-specific trend data and top causes of readmission. This commenter also requested that quarterly confidential feedback reports contain the SNF VBP Program measure calculated using 12 rolling months of data, and that we update such calculations quarterly. Lastly, one commenter requested that reports be provided more frequently than quarterly.

Response: We are currently providing many patient-level indicators to SNFs as part of the quarterly reports process, and since we began that reporting during the public comment period on the proposed rule, we believe some commenters may have erroneously believed that we did not intend to provide patient-level data. June 2017 quarterly confidential feedback reports and supplemental workbooks included the following patient-level data: Patient identifiers (Health Insurance Claim Number [HICN], Sex, Age); Index SNF information (admission/discharge dates, discharge status code); Prior proximal hospital information CMS Certification Number [CCN], admission/discharge dates, principal diagnosis); Readmission hospital information (CCN, admission/discharge dates, principal diagnosis); and SNFRM risk-adjustment factors. The following facility-level information is also included: Number of Eligible Stays, Number of Unplanned Readmissions, Observed Readmission Rate, Predicted Number of Readmissions, Expected Number of Readmissions, Standardized Risk Ratio (SRR), National Average Readmission Rate, RSRR. We will take the commenter's request to report patient's dual eligibility status under consideration for future reports.

We intend to publish performance standards for each program year in the SNF PPS final rule, and we intend to provide peer ranking information to SNFs as it becomes available. We believe that providing the SNF VBP program measure rate calculations using 12 rolling months of data updated quarterly would create confusion among providers regarding which of these rates would be used to calculate value-based incentive payments for a specific program year. We strive to provide information that is as user-friendly as possible and will take the commenter's request for SNF-specific trend data and top causes of readmission under consideration. Finally, while we appreciate the need for frequent updates, monthly reports containing this information are not logistically feasible at this time. However, we continue to look for ways in which we may provide this information more frequently in the future.

We thank the commenters for this feedback.

(2) Review and Corrections Process: Phase Two

In the FY 2017 SNF PPS final rule (81 FR 52007 through 52009), we adopted a two-phase review and corrections process for SNFs' quality measure data that will be made public under section 1888(g)(6) of the Act and SNF performance information that will be made public under section 1888(h)(9) of the Act. We explained that we would accept corrections to the quality measure data used to calculate the measure rates that is included in any SNF's quarterly confidential feedback report, and also that we would provide SNFs with an annual confidential feedback report containing the performance information that will be made public. We detailed the process for requesting Phase One corrections and finalized a policy whereby we would accept Phase One corrections to SNFs' quarterly reports through March 31 following the report's issuance via the CASPER system.

In the proposed rule (82 FR 21086 through 21087), we proposed additional specific requirements for the Phase Two review and correction process that we are finalizing in this final rule. Specifically, we proposed to limit Phase Two correction requests to the SNF's performance score and ranking because all SNFs would have already had the opportunity to correct their quality measure data through the Phase One corrections process.

We also proposed to provide these reports to SNFs at least 60 days prior to the FY involved. SNFs will not be allowed to request corrections to their value-based incentive payment adjustments. However, we stated that we will make confirming corrections to a SNF's value-based incentive payment adjustment if a SNF successfully requests a correction to its SNF performance score.

As with Phase One, we proposed that Phase Two correction requests must be submitted to the SNFVBPinquiries@cms.hhs.gov mailbox, and must contain the following information:

• SNF's CMS Certification Number (CCN);
• SNF Name;
• The correction requested and the SNF's basis for requesting the correction.

Specifically, the SNF must identify the error for which it is requesting correction, and explain the reason for requesting the correction. The SNF must also submit documentation or other evidence, if available, supporting the request. As noted above, corrections requested during Phase Two will be limited to SNFs' performance score and ranking. However, we noted that the SNFVBPinquiries@cms.hhs.gov mailbox cannot receive secured email messages. If any SNF believes it needs to submit patient-sensitive information as part of a correction request, we requested that the SNF contact us at the mailbox to arrange a secured transfer.

We further proposed that SNFs must make any correction requests no later than 30 days following the date of our posting of their annual SNF performance score report via the QIES system CASPER files. For example, if we post the reports on August 1, 2017, SNFs must review these reports and submit any correction requests by 11:59 p.m. Eastern Standard Time on August 31, 2017 (or the next business day, if the 30th day following the date of the posting is a weekend or federal holiday). We stated that we would not consider any requests for corrections to SNF performance scores or rankings that are received after this deadline.

We proposed to review all timely Phase Two correction requests that we receive and provide responses to SNFs that have requested corrections as soon as practicable. We also proposed to issue an updated SNF performance score report to any SNF that requests a correction with which we agree, and if necessary, to update any public postings on Nursing Home Compare and value-based incentive payment percentages, as applicable.

We sought public comments on this proposed Phase Two corrections process. A discussion of these comments, along with our responses, appears below.

Comment: Some commenters recommended that SNFs be provided access to the information used to calculate their SNFRM scores and estimate their payment adjustment factors based on the payment exchange function. Commenters stated that SNFs' may wish to replicate their SNF VBP performance scores as closely as possible, and requested that SNFs receive their predicted and expected readmission rates, national average readmission rates, and RSRRs for both the baseline and performance periods, as well as the cut points used to determine performance standards. Commenters explained that such information will help SNFs be more confident about their final payment adjustments as well as to understand what they need to do to improve their SNFRM scores and payment adjustments.

Response: We thank the commenters for this feedback. While it is correct that SNFs cannot calculate their own risk-standardized readmission rates because such a calculation would require national stay-level data, including risk-adjustment information, we believe that the additional patient-level and facility-level information that we are now providing to SNFs (as discussed further above) along with their quarterly reports will be useful to SNFs with their quality improvement efforts. We also provide SNFs with their predicted and expected readmission rates, national average readmission rates, and RSRRs in their quarterly confidential feedback reports and supplemental workbooks. We welcome commenters' continued feedback on the contents of the supplemental workbooks containing facility-level and patient-level data that accompany the quarterly confidential feedback reports.

Comment: One commenter requested that we provide Phase Two scoring reports to SNFs as soon as possible if we elect to change from calendar year to fiscal year performance periods to ensure that SNFs have sufficient time to review those reports and submit correction requests.

Response: We thank the commenter for this suggestion, and we will strive to provide SNF performance score reports to SNFs as quickly as possible. We note, however, that it is time consuming for us to complete the tasks necessary to ensure that the information contained in the performance score reports is accurate. At this time, we do not believe we can feasibly provide SNF performance score reports prior to the statutorily-required deadline described in section 1888(h)(7) of the Act that SNFs be notified of the adjustments to their Medicare payments as a result of the Program. We will consider future improvements if information technology and claims processing improvements allow for earlier dissemination of this information to SNFs.

Comment: One commenter supported our review and correction policies in general, but was unsure how a SNF could challenge its SNF performance score or ranking since SNFs do not receive patient-level data, and requested that we make such data available to SNFs. The commenter noted that additional information could be useful to SNFs, including their predicted readmission rate, their expected readmission rate, the national average, the SNF's baseline and performance period rates, the SNF's ranking, and the achievement and improvement thresholds.

Response: Our intention is to provide SNFs with the patient level data and associated data elements that the commenter suggests in the SNF performance score reports scheduled for delivery next year, though we note, as stated above, that we are now providing patient-level data in SNFs' quarterly confidential feedback reports. We welcome commenters' continued feedback on those data and any other elements that may be helpful to SNFs with their quality improvement efforts.

After consideration of the public comments that we received, we are finalizing the Phase Two review and corrections process, as proposed.

(3) SNF VBP Program Public Reporting

We refer readers to the FY 2017 SNF PPS final rule (81 FR 52009) for discussion of the statutory requirements governing the public reporting of SNFs' performance information under the SNF VBP Program. We also sought and responded to public comments on issues that we should take into account when posting performance information on Nursing Home Compare or a successor Web site.

We proposed to begin publishing SNF performance information under the SNF VBP Program on Nursing Home Compare not later than October 1, 2017. We stated that we would only publish performance information for which SNFs have had the opportunity to review and submit corrections. We sought comments on this proposal. A discussion of these comments, along with our responses, appears below.

Comment: One commenter supported posting SNF performance scores on Nursing Home Compare, but opposed posting quality measure performance scores, including achievement/improvement scores. The commenter stated that achievement and improvement scores are not required by statute to be publicly posted and could be confusing to the public. The commenter also noted that the Program's quality measures differ from those already posted on Nursing Home Compare, and stated that having multiple rehospitalization rates would not be ideal.

Response: We thank the commenter for this feedback. We note that section 1888(g)(6) of the Act directs the Secretary to make SNF-specific information available to the public, including information on measure-level performance, and we will consider the commenter's views as we develop our plans for public reporting of SNF VBP data in the future.

Comment: Commenter requested that we clarify our intentions for public reporting of SNF VBP information on Nursing Home Compare, wondering if this information will replace the current readmission rate information and definitions on the site or if SNF VBP information will be added to the site's current content. The commenter also expressed frustration that CMS is using multiple definitions of readmissions for different programs, and suggested that we align our efforts.

Response: We intend to publish SNF VBP performance information on Nursing Home Compare or a successor Web site as directed by the SNF VBP Program's statute. We are cognizant of the possibility for confusion, and we intend to align our efforts as much as possible across programs, including giving providers sufficient information to aid them in distinguishing between the readmission measures on Nursing Home Compare.

Comment: Commenter encouraged us to publish as much information as possible on Nursing Home Compare, including readmissions rates, achievement and improvement points, SNF performance scores, rankings, and payment adjustments. The commenter noted that many of these data points are available for the Hospital VBP and Readmissions Reduction Programs, and noted that the public should expect the same transparency for SNFs.

Response: We thank the commenter for this feedback and will take it into consideration as we continue developing our public reporting plans.

After consideration of the public comments that we have received, we are finalizing our public reporting policy as proposed.

(4) Ranking of SNFs' Performance

We refer readers to the FY 2017 SNF PPS final rule (81 FR 52009) for discussion of the statutory requirement that we rank SNFs based on their performance on the Program. In that rule, we discussed the statutory requirements to order SNF performance scores from low to high and publish those rankings on both the Nursing Home Compare and QualityNet Web sites, and to publish the ranking after August 1, 2018, when performance scores and value-based incentive payment adjustments will be made available to SNFs. We intend to publish the ranking for each program year once performance scores and value-based incentive payment adjustments are made available to SNFs.

Having considered those statutory requirements, we proposed to rank SNFs for the FY 2019 program year and to publish the ranking after August 1, 2018. We further proposed that the ranking include the following data elements:

• Rank,
• Provider ID,
• Facility name,
• Address,
• Baseline period (CY 2015) risk-standardized readmission rate,
• Performance period (CY 2017) risk-standardized readmission rate,
• Achievement score,
• Improvement score, and
• SNF performance score.

We believe that these data elements will provide consumers and other stakeholders with the necessary information to evaluate SNFs' performance under the program, including each component of the SNF performance score, including both achievement and improvement. We sought public comments on these proposals. We stated in the proposed rule that we would address rankings for future program years in subsequent rulemaking. A discussion of these comments, along with our responses, appears below.

Comment: One commenter stated its belief that we must publish the FY 2019 program ranking not later than August 1, 2018, rather than after August 1 as we described in the proposed rule. The commenter noted that publishing the ranking by that date will provide all stakeholders with sufficient time to review the ranking prior to the fiscal year.

Response: Section 1888(h)(9) of the Act does not provide a specific deadline for public reporting of SNF performance scores and the ranking for a given fiscal year. Our intention in stating that we would publish the ranking after August 1, 2018, was only to communicate that we would publish the ranking publicly after SNFs have been notified of their SNF performance scores, value-based incentive payment percentages, and ranking as required by section 1888(h)(7) of the Act, which must take place not later than 60 days prior to the fiscal year involved.

After consideration of the public comments, we are finalizing the SNF VBP Program's ranking policies as proposed.

4. Survey Team Composition

a. Background

To participate in the Medicare and Medicaid programs, long term care facilities, including skilled nursing facilities (SNFs) in Medicare and nursing facilities (NFs) in Medicaid, must be certified as meeting Federal participation requirements, which are specified in 42 CFR part 483. Section 1864(a) of the Act authorizes the Secretary to enter into agreements with state survey agencies to determine whether SNFs meet the federal participation requirements for Medicare and section 1902(a)(33)(B) of the Act provides for state survey agencies to perform the same survey tasks for NFs participating or seeking to participate in the Medicaid program. Surveys are performed directly by us and also under contract for certain surveys. The results of these surveys are used by us and the Medicaid state agency as the basis for a determination to enter into, deny, or terminate a provider agreement with the facility, or to impose an enforcement remedy or remedies on a facility, as appropriate, for failure to be in substantial compliance with federal participation requirements. To assess compliance with federal participation requirements, surveyors conduct onsite inspections (surveys) of facilities. In the survey process, surveyors gather evidence and directly observe the actual provision of care and services to residents and the effect or possible effects of that care, or lack thereof, to assess whether the care provided meets the assessed needs of individual residents.

Sections 1819(g) and 1919(g) of the Act, and corresponding regulations at 42 CFR part 488, subpart E, specify the requirements for the types and periodicity of surveys that are to be performed for each facility. Specifically, sections 1819(g)(2) and 1919(g)(2) of the Act reference standard, special, and extended surveys. Sections 1819(g)(2)(E) and 1919(g)(2)(E) of the Act specify that surveys under section 1819(g)(2) of the Act in general must consist of a multidisciplinary team of professionals, including a registered nurse. In addition, the statutory requirements governing the investigation of complaints and for monitoring on-site a SNF's or NF's compliance with participation requirements are found in sections 1819(g)(4) and 1919(g)(4) of the Act and § 488.332.

These sections specify that a specialized team, including an attorney, an auditor, and appropriate health care professionals may be maintained and utilized in the investigation of complaints for the purpose of identifying, surveying, gathering and preserving evidence, and carrying out appropriate enforcement actions against SNFs and NFs, respectively.

Consistent with the statutory provisions noted above, two separate regulations directly address survey team composition. Section 488.314, Survey Teams, reflects the statutory language under sections 1819(g)(2)(E)(i) and 1919(g)(2)(E)(i) of the Act, and states that “[s]urvey teams must be conducted by an interdisciplinary team of professions, which must include a registered nurse.” Section 488.332, Investigation of Complaints of Violations and Monitoring of Compliance, reflects the statutory language under sections 1819(g)(4) and 1919(g)(4) of the Act, and states that the state survey agency may use a specialized team, which may include an attorney, auditor, and appropriate health professionals, but not necessarily a registered nurse, to investigate complaints and conduct on-site monitoring. A survey conducted to monitor on-site a SNF's or NF's compliance with participation requirements, such as a revisit survey to determine whether a noncompliant facility has achieved substantial compliance, is also subject to the provisions of § 488.332, and not § 488.314.

Section 488.308(e) also addresses complaint investigations, but as currently written, it combines special surveys, which are authorized under sections 1819(g)(2)(A)(iii)(II) and 1919(g)(2)(A)(iii)(II) of the Act, with the requirements associated with the investigations of complaints, which are governed by sections 1819(g)(4) and 1919(g)(4) of the Act. In the statute, “special surveys” are referenced at sections 1819(g)(2)(A)(iii)(II) and 1919(g)(2)(A)(iii)(II) of the Act, while the investigation of complaints is referenced at sections 1819(g)(4) and 1919(g)(4) of the Act.

The regulations as currently written do not clearly indicate which survey team requirement applies to complaint surveys. The language at § 488.314 could be broadly interpreted to cover the survey team composition for all surveys, including those used to investigate a complaint. Such an interpretation, however, would ignore the provisions of § 488.332, which allow a state survey agency to utilize a specialized investigative team that does not necessarily include a registered nurse to survey a facility in connection with a complaint investigation. The placement of surveys to investigate a complaint together with special surveys under § 488.308(e) further places into question which survey team requirement applies to complaint surveys. However, CMS' State Operations Manual (SOM) (Internet Only Manual Pub. 100-07) notes that “Section 488.332 provides the Federal regulatory basis for the investigation of complaints about nursing homes,” thus indicating CMS' view that provisions related to survey team composition in § 488.332 apply to complaint surveys. See SOM, Ch. 5, Section 5300; see also SOM, Ch. 7, Sections 7203.5 and 7205.2(3); SOM, Appendix P, II.B.4A.

The lack of clarity as to which regulatory provision, that is, § 488.314 or § 488.332, applies to the survey team composition related to the investigation of complaints has been the cause of recent administrative litigation. We thus believe that regulatory changes are needed to clarify that only surveys conducted under sections 1819(g)(2) and 1919(g)(2) of the Act are subject to the requirement at § 488.314 that a survey team consist of an interdisciplinary team that must include a registered nurse. Complaint surveys and surveys related to on-site monitoring, including revisit surveys, are subject to the requirements of sections 1819(g)(4) and 1919(g)(4) of the Act and § 488.332, which allow the state survey agency to use a specialized investigative team that may include appropriate health care professionals but need not include a registered nurse.

b. Major Provisions

We proposed to make changes to §§ 488.30, 488.301, 488.308, and 488.314 to clarify the regulatory requirements for team composition for surveys conducted for investigating a complaint and to align regulatory provisions for investigation of complaints with the statutory requirements found in sections 1819 and 1919 of the Act.

(a) Proposed revision of the definition of “complaint survey” under § 488.30 to add a provision stating that the requirements of sections 1819(g)(4) and 1919(g)(4) of the Act and § 488.332 apply to complaint surveys.

(b) Proposed revision of the definition of “abbreviated standard survey” under § 488.301 to clarify that abbreviated standard surveys conducted to investigate a complaint or to conduct on-site monitoring to verify compliance with participation requirements are subject to the requirements of § 488.332.

(c) Proposed relocation of the requirements included in § 488.308(e)(2) and (3) related to surveys conducted to investigate a complaint from under the heading “Special Surveys” to a new paragraph (f), titled “Investigations of Complaints.”

(d) Proposed revision of the language at § 488.314(a)(1) to specify that the team composition requirements at § 488.314(a)(1) apply only to surveys under sections 1819(g)(2) and 1919(g)(2) of the Act.

Commenters submitted the following comments related to the proposed rule's discussion of the Survey Team Composition. A discussion of these comments, along with our responses, appears below.

Comment: We received one comment supporting our proposal and the commenter agreed with our clarification on the survey team composition. The commenter further stated that states should be able to determine the composition of the survey team based on the complaint received and the purpose of the revisit to determine compliance.

Response: We want to thank the commenter for their support of our clarifications to the survey team composition. We agree that the states should be able to determine which professional would be most appropriate based on the complaint received, such as a registered nurse for clinical concerns, a dietitian for dietary concerns, or a pharmacist for medication issues for example.

Comment: We received several comments recommending us to consider adding a Registered Nurse (RN) to all survey teams. Multiple commenters stated that an RN should be the individual to investigate any alleged incident. Another commenter stated that they believed statutory language is clear that a survey team must include a registered professional nurse, and that the citation of clinical violations should be observed and made by a registered professional nurse. One commenter recommended that we add a requirement for a psychosocial professional to be on each team in addition to a registered nurse. One commenter also recommended that in addition to having an RN on the survey team, the team should also include an additional professional based on the complaint type.

Response: We appreciate the feedback from the commenters regarding the suggestion to have an RN on all surveys or to add a psychosocial professional to the team, but the proposed change to the language regarding survey team composition is not to change the composition of survey teams, but to clarify the requirement that survey teams conducted by an interdisciplinary team of professionals, including a registered nurse applies only to surveys under sections 1819(g)(2) and 1919(g)(2) of the Act and does not apply to complaint surveys in which the appropriate professional would be used to conduct the investigation based on the type of allegation.

Comment: One commenter stated that they disagreed with our interpretation of its statutory authority. The commenter stated that they believed statutory requirement for a registered nurse on this team is clear and that the statute draws no distinction between a complaint survey and a standard survey. The commenter further stated that citations of clinical violations should be observed and confirmed or dismissed by a registered professional nurse based upon his or her clinical judgment.

Response: The preamble to the proposed rule states that the proposed change is to clarify the requirement that survey teams conducted by an multidisciplinary team of professionals, including a registered nurse, applies only to surveys described under sections 1819(g)(2) and 1919(g)(2) of the Act and does not apply to the investigation of complaints. The authority for complaint surveys arises under sections 1819(g)(4) and 1919(g)(4) of the Act, which authorizes the State survey agency to use a specialized team, which includes appropriate healthcare professionals that may or may not, if not required, include a registered nurse, for purposes of, among other things, “surveying” noncompliant facilities. As discussed in the preamble, we believe these clarifying changes are consistent with the statutory provisions of sections 1819(g)(2) and (g)(4) and 1919(g)(2) and (g)(4) of the Act, as well as our long standing interpretation of the statute, as expressed in the implementation of current regulations at §§ 488.314 and 488.332 and the State Operations Manual (“SOM”). We believe that if we were to require a registered nurse on all surveys including those that are meant to investigate complaint allegations, it would place an undue burden on the resources of state survey agencies and render the statutory language under sections 1819(g)(4) and 1919(g)(4) of the Act as meaningless. In addition, as previously mentioned, we believe that the statute enables us to determine which professional would be most appropriate to investigate complaint allegations based on the nature of the complaint allegation received.

Comment: We received one comment requesting a revision based on the decision at DAB No. CR4670 (2016) (H.H.S.), 2016 WL 499224, in which an Administrative Law Judge provided an interpretation of the survey composition provisions in the statute and current regulations.

Response: We appreciate the commenter's reference to this case, however the ALJ decision is currently being reviewed by the Departmental Appeals Board Appellate Division and therefore we cannot comment on this case at this time.

Based on the comments received, we are proceeding with the finalization of our proposal without any changes.

5. Correction of the Performance Period for the National Healthcare Safety Network (NHSN) Healthcare Personnel (HCP) Influenza Vaccination Immunization Reporting Measure in the End-Stage Renal Disease (ESRD) Quality Incentive Program (QIP) for Payment Year (PY) 2020

In the CY 2017 ESRD PPS final rule (81 FR 77834), we finalized that the performance period for the NHSN Healthcare Personnel Influenza Vaccination Reporting Measure for Payment Year (PY) 2020 would be from October 1, 2016, through March 31, 2017 (81 FR 77915). We proposed to revise that performance period so that it aligns with the schedule we previously set for this measure. Specifically, we previously finalized that for the PY 2018 ESRD QIP, the performance period for this measure would be from October, 1, 2015 through March 31, 2016, which is consistent with the length of the 2015-2016 influenza season (79 FR 66209), and that for the PY 2019 ESRD QIP, the performance period for this measure would be from October, 1, 2016 through March 31, 2017, which is consistent with the length of the 2016- 2017 influenza season (80 FR 69059 through 69060). Maintaining the performance period we finalized in the CY 2017 ESRD PPS final rule would result in scoring facilities on the same data twice, and would not be consistent with our intended schedule to collect data on the measure in successive influenza seasons. Therefore, we proposed to revise the performance period for the NHSN HCP Influenza Vaccination Reporting Measure for the PY 2020 ESRD QIP. Specifically, we proposed that for the PY 2020 ESRD QIP, the performance period for this measure would be October 1, 2017, through March 31, 2018, which is consistent with the length of the 2017-2018 influenza season.

We sought comments on this proposal. A discussion of these comments, along with our responses, appears below.

Comment: Commenters were generally supportive of our proposal to set the performance period as October 1, 2017 through March 31, 2018 because it is consistent with the length of the 2017-2018 influenza season, however they stated that to be truly consistent with the influenza season and the standard practice of administering the vaccine, the performance period for the measure should be aligned with the CDC's recommendations that vaccination occur as early as possible to protect against infection. They stated that without including the phrase “or when the vaccine becomes available,” the measure penalizes facilities that provide the vaccine as soon as it becomes available in August or September. One commenter also stated that not making this change could place patients at increased risk early in the influenza season.

Response: As stated in the CY 2015 ESRD PPS final rule (79 FR 66207) in response to a commenter who was concerned about whether vaccinations received before October 1 would qualify under this measure, “the performance period for the denominator (the number of healthcare personnel working in a facility) is from October 1 through March 31. However, the numerator measurement (vaccination status) includes vaccines obtained `as soon as the vaccine is available.' As a result, a Healthcare Personnel (HCP) working at the facility as of October 1 who was vaccinated in September would be considered vaccinated for the performance period under this measure” (79 FR 66207). As a result, facilities will not be penalized for providing the vaccine as soon as it becomes available and patients will not be placed at an increased risk at any point during the influenza season due to the vaccination status of HCPs working in the facility.

After carefully considering the comments received we are finalizing the Performance Period for the NHSN HCP Influenza Vaccination Reporting Measure for the ESRD QIP for Payment Year 2020 as proposed.

IV. Collection of Information Requirements

Under the Paperwork Reduction Act of 1995 (PRA) (44 U.S.C. 3501 et seq.), we are required to publish a 60-day notice in the Federal Register and solicit public comment before a collection of information requirement is submitted to the Office of Management and Budget (OMB) for review and approval.

To fairly evaluate whether an information collection should be approved by OMB, PRA section 3506(c)(2)(A) requires that we solicit comment on the following issues:

• The need for the information collection and its usefulness in carrying out the proper functions of our agency.
• The accuracy of our burden estimates.
• The quality, utility, and clarity of the information to be collected.
• Our effort to minimize the information collection burden on the affected public, including the use of automated collection techniques.

We solicited public comment in the FY 2018 SNF PPS proposed rule on each of the section 3506(c)(2)(A)-required issues for the following information collection requirements (ICRs).

A. Information Collection Requirements (ICRs)

1. ICRs Regarding the SNF VBP Program

As discussed in the FY 2016 SNF PPS final rule (80 FR 46473) and the FY 2017 SNF PPS final rule (81 FR 52049 through 52050), we have specified claims-based measures to fulfill the SNF VBP Program's requirements. As required by the SNF VBP Program's statute, we will score SNFs' performance on these measures in order to make value-based incentive payments to SNFs beginning in FY 2019.

In this final rule, we are finalizing additional policies for the SNF VBP Program, including performance standards and performance/baseline periods for the FY 2020 Program year, an exchange function for the FY 2019 Program year, and administrative requirements related to review and correction of performance information to be made public. None of these requirements result in any additional information collections or reporting burden associated with the Program.

Additionally, because claims-based measures are calculated based on claims figures that are already submitted to the Medicare program for payment purposes, there is no additional respondent burden associated with data collection or submission for either the SNFRM or SNFPPR measures. Thus, there is no additional reporting burden associated with the SNF VBP Program's measures finalized in this rule.

2. ICRs Regarding the Potentially Preventable 30-Day Post-Discharge Readmission Measure

This rule modifies the Potentially Preventable 30-Day Post-Discharge Readmission Measure by increasing the length of the measurement period and updating the confidential feedback and public reporting dates, as described in section III.D.2.h. Because this is a claims-based measure, no data collection beyond Medicare claims submitted by SNFs for the furnishing of SNF covered services are required for the calculation of this measure. We believe the SNF QRP burden estimate is unaffected by the modifications of this measure as the modifications have no impact on any of the claims-based reported data fields.

3. ICRs Exempt From the PRA

As discussed in this final rule, we are adopting five new measures beginning with the FY 2020 SNF QRP (see section III.D.2.g). The five new measures being finalized are: (1) Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury; (2) Application of the IRF Functional Outcome Measure: Change in Self-Care Score for Medical Rehabilitation Patients (NQF #2633); (3) Application of IRF Functional Outcome Measure: Change in Mobility Score for Medical Rehabilitation Patients (NQF #2634); (4) Application of IRF Functional Outcome Measure: Discharge Self-Care Score for Medical Rehabilitation Patients (NQF #2635); and (5) Application of IRF Functional Outcome Measure: Discharge Mobility Score for Medical Rehabilitation Patients (NQF #2636). The measures must be collected by SNFs and reported to CMS using the Resident Assessment Instrument, Minimum Data Set (MDS).

These measures will be calculated using data elements that are included in the MDS. The data elements are discrete questions and response codes that collect information on a SNF patient's health status, preferences, goals and general administrative information. To view the MDS, with the finalized data elements, we refer to the reader to https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/Skilled-Nursing-Facility-Quality-Reporting-Program/SNF-Quality-Reporting-Program-Measures-and-Technical-Information.html.

This rule also finalizes that SNFs would be required to report certain standardized resident assessment data beginning with the FY 2019 SNF QRP (see section III.D.2.j.). We are finalizing our definition of the term “standardized resident assessment data” as patient assessment questions and response options that are identical in all four PAC assessment instruments, and to which identical standards and definitions apply. The standardized resident assessment data are intended to be shared electronically among PAC providers and will otherwise enable the data to be comparable for various purposes, including the development of cross-setting quality measures and to inform payment models that take into account patient characteristics rather than setting.

Under section 1899B(m) of the Act, the Paperwork Reduction Act does not apply to the specific changes in the collections of information described in this final rule. These changes to the collections of information are being finalized under section 2(a) of the IMPACT Act, which added new section 1899B to the Act. That section requires SNFs to report standardized resident assessment data, data on quality measures, and data on resource use and other measures. All of this data must, under section 1899B(a)(1)(B) of the Act, be standardized and interoperable to allow for its exchange among PAC providers and other providers and the use by such providers to provide access to longitudinal information to facilitate coordinated care and improved Medicare beneficiary outcomes. Section 1899B(a)(1)(C) of the Act requires us to modify the MDS to allow for the submission of quality measure data and standardized resident assessment data to enable its comparison across SNFs and other providers. We are, however, setting out the burden as a courtesy to advise interested parties of the proposed actions' time and costs and for reference refer to section V.A of this final rule of the regulatory impact analysis (RIA). The requirement and burden will be submitted to OMB for review and approval when the modifications to the MDS have achieved standardization and are no longer exempt from the requirements under section 1899B(m) of the Act.

For the new measure “Changes in Skin Integrity Post-Acute Care: Pressure Ulcer/Injury” (NQF #2633) the items used to calculate this measure are already present on the MDS, so the adoption of this measure will not require SNFs to report any new data elements. In addition, we are removing some data elements related to pressure ulcers that have been identified as duplicative. Taking these final policies together, we estimate that there will be a 1.5 minute reduction in clinical staff time needed to report the pressure ulcer measure data. We are also removing 9 additional data elements from the MDS 3.0. The removal of these data elements from the skin integrity section of the MDS provide a reduction in burden with data reporting by SNFs and therefore serve as offsets to the SNF QRP. These removals are: Date of oldest Stage 2 pressure ulcer; three items pertaining to the dimensions of an unhealed pressure ulcer; the most severe tissue type for any pressure ulcer; and four data elements pertaining to healed pressure ulcers. We estimate that the data elements we are removing will reduce overall reporting burden from the assessments, constituting a reduction of an additional 7 minutes of clinical staff time per stay which provide a reduction in burden with data reporting by SNFs. Taken together, we are removing a total of 12 data elements from the skin integrity section of the MDS. Based on the data provided in Table 25 of this final rule, and estimating 2,886,336 discharges from 15,447 SNFs annually, we also estimate that the total cost of reporting these data will reduce overall reporting burden for the assessments from what was proposed constituting a total reduction of 8.5 minutes of clinical staff time per stay or $1,837 per SNF annually, or $28,377,493 for all SNFs annually. We believe that the MDS items will be completed by registered nurses (BLS Occupation Code: 29-1141) at $69.40/hr including overhead and fringe benefits.

For the four functional outcome measures (NQF: #2633, #2634, #2635, and #2636) that we are finalizing in this final rule, we note that although some of the data elements needed to calculate these measures are currently included on the MDS, other data elements need to be added to the MDS. As a result, we estimate that reporting these measures will require an additional 9 minutes of nursing and therapy staff time to report data on admission and 5.5 minutes of nursing and therapy time to report data on discharge, for a total of 14.5 additional minutes per stay. We estimate that the additional MDS items we are finalizing will be completed by Registered Nurses for approximately 7 percent of the time. Occupational Therapists (BLS Occupation Code: 29-1122) at $80.50/hr including overhead and fringe benefits for approximately 41 percent of the time, and Physical Therapists (BLS Occupation Code: 29-1123) at $83.86/hr including overhead and fringe benefits for approximately 52 percent of the time. Individual providers determine the staffing resources necessary. With 2,886,336 discharges from 15,447 SNFs annually, we estimate that the reporting of the four functional outcome measures would impose on SNFs an additional burden of 697,531 total hours (2,886,336 discharges × 14.5 min/60) or 45.16 hours per SNF (697,531 hr/15,447 SNFs). Of the 14.5 minutes per stay, 1 minute of that time is for a Registered Nurse, 3.5 minutes is for an Occupational Therapist, and 4.5 minutes is for a Physical Therapist for a total of 9 minutes are required for admission. For discharge, 2.5 minutes are for an Occupational Therapist, and 3 minutes for a Physical Therapist for a total of 5.5 minutes. For one stay we estimate a cost of $19.69 or, in aggregate, an annual cost of $56,829,551. Per SNF, we estimate an annual cost of $3,679. A summary of these estimates is provided in Table 25.

We are not finalizing our proposal to adopt 1 new standardized resident assessment data elements with respect to SNF admissions and 11 new standardized resident assessment data elements with respect to SNF discharges. This results in a reduction to the burden that we estimated in the proposed rule. We refer readers to the proposed rule (82 FR 21091 through 21092) for a discussion of our burden estimates for these proposals. Our updated estimate is provided in Table 25 (Revised Calculation of Burden), and results in a final estimated burden for the SNF QRP of $28,452,058.

We received the following public comments on our collection of information estimates.

Comment: A few commenters expressed concern about the administrative burden imposed by the SNF QRP, specifically referring to the volume and the pace of data collection that is required by the implementation of the SNF QRP.

Response: We appreciate the commenters' concerns regarding burden due to changes to the SNF QRP as a result of the fulfillment of the requirements of the IMPACT Act. We appreciate the importance of avoiding undue burden on providers and will continue to evaluate and avoid any unnecessary burden associated with the implementation of the SNF QRP. We will continue to work with stakeholders to explore ways to minimize and decrease burden as our mutual goal is to focus on improving patient care. Finally, in response to stakeholders' concerns regarding burden, we have decided not to finalize a number of the proposed standardized resident assessment data elements. This results in a reduction to the burden estimate that appeared in the proposed rule.

V. Economic Analyses

A. Regulatory Impact Analysis

1. Introduction

We have examined the impacts of this final rule as required by Executive Order 12866 on Regulatory Planning and Review (September 30, 1993), Executive Order 13563 on Improving Regulation and Regulatory Review (January 18, 2011), the Regulatory Flexibility Act (RFA, September 19, 1980, Pub. L. 96-354), section 1102(b) of the Act, section 202 of the Unfunded Mandates Reform Act of 1995 (UMRA, March 22, 1995; Pub. L. 104-4), Executive Order 13132 on Federalism (August 4, 1999), the Congressional Review Act (5 U.S.C. 804(2)), and Executive Order 13771 on Reducing Regulation and Controlling Regulatory Costs (January 30, 2017). Executive Orders 12866 and 13563 direct agencies to assess all costs and benefits of available regulatory alternatives and, if regulation is necessary, to select regulatory approaches that maximize net benefits (including potential economic, environmental, public health and safety effects, distributive impacts, and equity). Executive Order 13563 emphasizes the importance of quantifying both costs and benefits, of reducing costs, of harmonizing rules, and of promoting flexibility. This rule has been designated an economically significant rule, under section 3(f)(1) of Executive Order 12866. Accordingly, we have prepared a regulatory impact analysis (RIA) as further discussed below.

Executive Order 13771, titled Reducing Regulation and Controlling Regulatory Costs, was issued on January 30, 2017. This final rule is considered an EO 13771 regulatory action. Details on the estimated costs of this rule can be found in the preceding and subsequent analyses.

2. Statement of Need

This final rule updates the FY 2017 SNF prospective payment rates as required under section 1888(e)(4)(E) of the Act. It also responds to section 1888(e)(4)(H) of the Act, which requires the Secretary to provide for publication in the Federal Register before the August 1 that precedes the start of each FY, the unadjusted federal per diem rates, the case-mix classification system, and the factors to be applied in making the area wage adjustment. As these statutory provisions prescribe a detailed methodology for calculating and disseminating payment rates under the SNF PPS, we do not have the discretion to adopt an alternative approach on these issues.

3. Overall Impacts

This final rule sets forth updates of the SNF PPS rates contained in the SNF PPS final rule for FY 2017 (81 FR 51970). Based on the above, we estimate that the aggregate impact is an increase of $370 million in payments to SNFs in FY 2018, resulting from the SNF market basket update to the payment rates, as required by section 1888(e)(5)(B)(iii) of the Act. We would note that this estimate is different from the estimated impact of $390 million provided in the FY 2018 SNF PPS proposed rule (82 FR 21016, 21093), as we relied on an updated SNF baseline spending figure for the final rule which reflect baseline spending from the FY 2018 President's budget, as opposed to that used in the proposed rule which was based on the Mid-session review of the FY 2017 President's budget.

We would note that events may occur to limit the scope or accuracy of our impact analysis, as this analysis is future-oriented, and thus, very susceptible to forecasting errors due to events that may occur within the assessed impact time period.

In accordance with sections 1888(e)(4)(E) and 1888(e)(5) of the Act, if not for the enactment of section 411(a) of MACRA (as discussed in section III.B.2. of this final rule), we would update the FY 2017 payment rates by a factor equal to the market basket index percentage change adjusted by the MFP adjustment to determine the payment rates for FY 2018. As discussed previously, section 1888(e)(5)(B)(iii) of the Act establishes a special rule for FY 2018 requiring the market basket percentage used to update the federal SNF PPS rates to be equal to 1.0 percent. The impact to Medicare is included in the total column of Table 25. In updating the SNF PPS rates for FY 2018, we made a number of standard annual revisions and clarifications mentioned elsewhere in this final rule (for example, the update to the wage and market basket indexes used for adjusting the federal rates).

The annual update set forth in this final rule applies to SNF PPS payments in FY 2018. Accordingly, the analysis of the impact of the annual update that follows only describes the impact of this single year. Furthermore, in accordance with the requirements of the Act, we will publish a rule or notice for each subsequent FY that will provide for an update to the payment rates and include an associated impact analysis.

We estimate the impact for the SNF QRP based on 15,447 SNFs in FY 2016 which had a total of 2,886,336 Medicare covered discharges for Medicare fee for service beneficiaries. This would equate to 288,633 total added hours or 18.69 hours per SNF annually. We anticipate that the additional MDS items we finalized will be completed by Registered Nurses (RN), Occupational Therapists (OT), and/or Physical Therapists (PT), depending on the item. Individual providers determine the staffing resources necessary. We obtained mean hourly wages for these staff from the U.S. Bureau of Labor Statistics' May 2016 National Occupational Employment and Wage Estimates ( https://www.bls.gov/oes/current/oes_nat.htm ), and to account for overhead and fringe benefits, we have doubled the mean hourly wage.

Estimated impacts for the SNF QRP are based on analysis discussed in section III.D.2. of this final rule. For the 8.5 minute reduction in burden associated with the new pressure ulcer measure and the removal of duplicative pressure ulcer data elements and data elements no longer being used, and the additional 14.5 additional minutes of burden for the functional outcome measures, the overall cost associated with finalized changes to the SNF QRP is $28,452,058.

4. Detailed Economic Analysis

The FY 2018 SNF PPS payment impacts appear in Table 26. Using the most recently available data, in this case FY 2016, we apply the current FY 2017 wage index and labor-related share value to the number of payment days to simulate FY 2017 payments. Then, using the same FY 2016 data, we apply the FY 2018 wage index and labor-related share value to simulate FY 2018 payments. We tabulate the resulting payments according to the classifications in Table 26 (for example, facility type, geographic region, facility ownership), and compare the simulated FY 2017 payments to the simulated FY 2018 payments to determine the overall impact. The breakdown of the various categories of data in the table follows:

• The first column shows the breakdown of all SNFs by urban or rural status, hospital-based or freestanding status, census region, and ownership.
• The first row of figures describes the estimated effects of the various changes on all facilities. The next 6 rows show the effects on facilities split by hospital-based, freestanding, urban, and rural categories. The next 19 rows show the effects on facilities by urban versus rural status by census region. The last 3 rows show the effects on facilities by ownership (that is, government, profit, and non-profit status).
• The second column shows the number of facilities in the impact database.
• The third column shows the effect of the annual update to the wage index. This represents the effect of using the most recent wage data available. The total impact of this change is zero percent; however, there are distributional effects of the change.
• The fourth column shows the effect of all of the changes on the FY 2018 payments. The update of 1.0 percent is constant for all providers and, though not shown individually, is included in the total column. It is projected that aggregate payments will increase by 1.0 percent, assuming facilities do not change their care delivery and billing practices in response.

As illustrated in Table 26, the combined effects of all of the changes vary by specific types of providers and by location. For example, due to changes finalized in this rule, providers in the urban Pacific region could experience a 1.5 percent increase in FY 2018 total payments.

5. Estimated Impacts for the SNF QRP

We estimate the impact for the SNF QRP based on 15,447 SNFs in FY 2016 which had a total of 2,886,336 Medicare covered discharges for Medicare fee for service beneficiaries. This would equate to 288,633 total added hours or 18.69 hours per SNF annually. We anticipate that the additional MDS items we finalized will be completed by Registered Nurses (RN), Occupational Therapists (OT), and/or Physical Therapists (PT), depending on the item. Individual providers determine the staffing resources necessary. We obtained mean hourly wages for these staff from the U.S. Bureau of Labor Statistics' May 2016 National Occupational Employment and Wage Estimates ( https://www.bls.gov/oes/current/oes_nat.htm ), and to account for overhead and fringe benefits, we have doubled the mean hourly wage.

Estimated impacts for the SNF QRP are based on analysis discussed in section III.D.2. of this final rule. For the 8.5 minute reduction in burden associated with the new pressure ulcer measure and the removal of duplicative pressure ulcer data elements and data elements no longer being used, and the additional 14.5 additional minutes of burden for the functional outcome measures, the overall cost associated with finalized changes to the SNF QRP is $28,452,058.

6. Estimated Impacts for the SNF VBP Program

Estimated impacts of the FY 2019 SNF VBP Program are based on historical data that appear in Table 28. We modeled SNFs' performance under the Program using SNFRM data from CY 2013 as the baseline period and CY 2015 as the performance period. Additionally, we modeled a logistic exchange function with a payback percentage of 60 percent, as discussed further in the preamble to this final rule.

As illustrated in Table 28, the effects of the SNF VBP Program vary by specific types of providers and by location. For example, we estimate that rural SNFs perform better on the SNFRM, on average, compared to urban SNFs. Similarly, we estimate that non-profit SNFs perform better on the SNFRM compared to for-profit SNFs, and that government-owned SNFs perform better still. We also estimate that smaller SNFs (measured by bed size) tend to perform better, on average, compared to larger SNFs. (We note that the risk-standardized readmission rates presented below are not inverted; that is, lower rates represent better performance).

These differences in performance on the SNFRM result in differences in value-based incentive payment percentages computed by the Program. For example, we estimate that, at the proposed 60 percent payback percentage, SNFs in urban areas would receive a 1.161 percent incentive multiplier, on average, in FY 2019, while SNFs in rural areas would receive a slightly higher incentive multiplier of 1.227 percent, on average. Additionally, SNFs in the smallest 25 percent as measured by bed size would receive an incentive multiplier of 1.203 percent, on average, while SNFs in the 2nd quartile as measured by bed size would receive an incentive multiplier of 1.166 percent, on average. We note that the multipliers that we have listed in Table 27 are applied to SNFs' adjusted Federal per diem rates after application of the 2 percent reduction to those rates required by statute.

7. Regulatory Review Costs

If regulations impose administrative costs on private entities, such as the time needed to read and interpret this final rule, we should estimate the cost associated with regulatory review. Due to the uncertainty involved with accurately quantifying the number of entities that will review the rule, we assume that the total number of unique commenters on the published proposed rule will be the number of reviewers of this final rule. We acknowledge that this assumption may understate or overstate the costs of reviewing this final rule. It is possible that not all commenters reviewed the proposed rule in detail, and it is also possible that some reviewers chose not to comment on the proposed rule. For these reasons we thought that the number of comments received on the proposed rule would be a fair estimate of the number of reviewers of this final rule.

We also recognize that different types of entities are in many cases affected by mutually exclusive sections of this final rule, and therefore for the purposes of our estimate we assume that each reviewer reads approximately 50 percent of the rule.

Using the wage information from the BLS for medical and health service managers (Code 11-9111), we estimate that the cost of reviewing this rule is $105.16 per hour, including overhead and fringe benefits ( https://www.bls.gov/oes/current/oes_nat.htm ) Assuming an average reading speed, we estimate that it would take approximately 4 hours for the staff to review half of this final rule. For each SNF that reviews the rule, the estimated cost is $421 (4 hours × $105.16). Therefore, we estimate that the total cost of reviewing this regulation is $103,987 ($421 × 247 reviewers).

8. Alternatives Considered

As described in this section, we estimate that the aggregate impact for FY 2018 under the SNF PPS is an increase of $370 million in payments to SNFs, resulting from the SNF market basket update to the payment rates, as required by section 1888(e)(5)(B)(iii) of the Act.

Section 1888(e) of the Act establishes the SNF PPS for the payment of Medicare SNF services for cost reporting periods beginning on or after July 1, 1998. This section of the statute prescribes a detailed formula for calculating base payment rates under the SNF PPS, and does not provide for the use of any alternative methodology. It specifies that the base year cost data to be used for computing the SNF PPS payment rates must be from FY 1995 (October 1, 1994, through September 30, 1995). In accordance with the statute, we also incorporated a number of elements into the SNF PPS (for example, case-mix classification methodology, a market basket index, a wage index, and the urban and rural distinction used in the development or adjustment of the federal rates). Further, section 1888(e)(4)(H) of the Act specifically requires us to disseminate the payment rates for each new FY through the Federal Register, and to do so before the August 1 that precedes the start of the new FY; accordingly, we are not pursuing alternatives for this process.

9. Accounting Statement

As required by OMB Circular A-4 (available online at https://obamawhitehouse.archives.gov/omb/circulars_a004_a-4/ ) in Table 29, we have prepared an accounting statement showing the classification of the expenditures associated with the provisions of this final rule for FY 2018. Table 29 provides our best estimate of the possible changes in Medicare payments under the SNF PPS as a result of the policies in this final rule, based on the data for 15,468 SNFs in our database and the cost for the SNF QRP of implementing the IMPACT Act.

10. Conclusion

This final rule sets forth updates of the SNF PPS rates contained in the SNF PPS final rule for FY 2017 (81 FR 51970). Based on the above, we estimate the overall estimated payments for SNFs in FY 2018 are projected to increase by $370 million, or 1.0 percent, compared with those in FY 2017. We estimate that in FY 2018 under RUG-IV, SNFs in urban and rural areas will experience, on average, a 1.1 percent increase and 0.4 percent increase, respectively, in estimated payments compared with FY 2017. Providers in the rural New England region will experience the largest estimated increase in payments of approximately 2.5 percent. Providers in the urban Outlying region will experience the largest estimated decrease in payments of 1.0 percent.

Additionally, § 488.314 regarding survey team composition implements section 1819(g)(4) of the Act and provides that States may maintain and utilize a specialized team that need not include a registered nurse for the investigation of complaints. Section 1919 of the Act contains the same statutory language as applicable to nursing facilities (NFs). Part 488 was originally established under the authority of sections 1819 and 1919 of the Act, which were added by the Omnibus Budget Reconciliation Act of 1987 (OBRA 87, Pub. L. 100-203, enacted on December 22, 1987) and further amendments to OBRA 87 by subsequent 1988, 1989, and 1990 legislation.

Sections 4204(b) and 4214(d) of OBRA 87 pertain to SNFs and NFs, respectively, and provide for a waiver of PRA requirements for the regulations that implement the OBRA 87 requirements. The provisions of OBRA 87 that exempt agency actions to collect information from states or facilities relevant to survey and enforcement activities from the PRA are not time-limited.

B. Regulatory Flexibility Act Analysis

The RFA requires agencies to analyze options for regulatory relief of small entities, if a rule has a significant impact on a substantial number of small entities. For purposes of the RFA, small entities include small businesses, non-profit organizations, and small governmental jurisdictions. Most SNFs and most other providers and suppliers are small entities, either by reason of their non-profit status or by having revenues of $27.5 million or less in any 1 year. We utilized the revenues of individual SNF providers (from recent Medicare Cost Reports) to classify a small business, and not the revenue of a larger firm with which they may be affiliated. As a result, we estimate approximately 97 percent of SNFs are considered small businesses according to the Small Business Administration's latest size standards (NAICS 623110), with total revenues of $27.5 million or less in any 1 year. (For details, see the Small Business Administration's Web site at https://www.sba.gov/contracting/getting-started-contractor/make-sure-you-meet-sba-size-standards ). In addition, approximately 23 percent of SNFs classified as small entities are non-profit organizations. Finally, individuals and states are not included in the definition of a small entity.

This final rule sets forth updates of the SNF PPS rates contained in the SNF PPS final rule for FY 2017 (81 FR 51970). Based on the above, we estimate that the aggregate impact for FY 2018 is an increase of $370 million in payments to SNFs, resulting from the SNF market basket update to the payment rates. While it is projected in Table 26 that most providers will experience a net increase in payments, we note that some individual providers within the same region or group may experience different impacts on payments than others due to the distributional impact of the FY 2018 wage indexes and the degree of Medicare utilization.

Guidance issued by the Department of Health and Human Services on the proper assessment of the impact on small entities in rulemakings, utilizes a cost or revenue impact of 3 to 5 percent as a significance threshold under the RFA. In their March 2017 Report to Congress (available at http://medpac.gov/docs/default-source/reports/mar17_medpac_ch8.pdf ), MedPAC states that Medicare covers approximately 11 percent of total patient days in freestanding facilities and 21 percent of facility revenue (March 2017 MedPAC Report to Congress, 202). As a result, for most facilities, when all payers are included in the revenue stream, the overall impact on total revenues should be substantially less than those impacts presented in Table 26. As indicated in Table 25, the effect on facilities is projected to be an aggregate positive impact of 1.0 percent for FY 2018. As the overall impact on the industry as a whole, and thus on small entities specifically, is less than the 3 to 5 percent threshold discussed previously, the Secretary has determined that this final rule will not have a significant impact on a substantial number of small entities for FY 2018.

In addition, section 1102(b) of the Act requires us to prepare a regulatory impact analysis if a rule may have a significant impact on the operations of a substantial number of small rural hospitals. This analysis must conform to the provisions of section 604 of the RFA. For purposes of section 1102(b) of the Act, we define a small rural hospital as a hospital that is located outside of an MSA and has fewer than 100 beds. This final rule affects small rural hospitals that (1) furnish SNF services under a swing-bed agreement or (2) have a hospital-based SNF.

We anticipate that the impact on small rural hospitals will be similar to the impact on SNF providers overall. Moreover, as noted in previous SNF PPS final rules (most recently, the one for FY 2017 (81 FR 51970)), the category of small rural hospitals is included within the analysis of the impact of this final rule on small entities in general. As indicated in Table 25, the effect on facilities for FY 2018 is projected to be an aggregate positive impact of 1.0 percent. As the overall impact on the industry as a whole is less than the 3 to 5 percent threshold discussed above, the Secretary has determined that this final rule does not have a significant impact on a substantial number of small rural hospitals for FY 2018.

C. Unfunded Mandates Reform Act Analysis

Section 202 of the Unfunded Mandates Reform Act of 1995 also requires that agencies assess anticipated costs and benefits before issuing any rule whose mandates require spending in any 1 year of $100 million in 1995 dollars, updated annually for inflation. In 2017, that threshold is approximately $148 million. This final rule will impose no mandates on state, local, or tribal governments or on the private sector.

D. Federalism Analysis

Executive Order 13132 establishes certain requirements that an agency must meet when it issues a final rule that imposes substantial direct requirement costs on state and local governments, preempts state law, or otherwise has federalism implications. This final rule has no substantial direct effect on state and local governments, preempt state law, or otherwise have federalism implications.

E. Congressional Review Act

This regulation is subject to the Congressional Review Act provisions of the Small Business Regulatory Enforcement Fairness Act of 1996 (5 U.S.C. 801 et seq.) and has been transmitted to the Congress and the Comptroller General for review.

In accordance with the provisions of Executive Order 12866, this final rule was reviewed by the Office of Management and Budget.

List of Subjects

42 CFR Part 409

• Health facilities
• Medicare

42 CFR Part 411

• Diseases
• Medicare
• Reporting and recordkeeping requirements

42 CFR Part 413

• Health facilities
• Diseases
• Medicare
• Reporting and recordkeeping requirements

42 CFR part 424

• Emergency medical services
• Health facilities
• Health professions
• Medicare
• Reporting and recordkeeping requirements

42 CFR Part 488

• Administrative practice and procedure
• Health facilities
• Medicare
• Reporting and recordkeeping requirements

For the reasons set forth in the preamble, the Centers for Medicare & Medicaid Services amends 42 CFR chapter IV as set forth below:

PART 409—HOSPITAL INSURANCE BENEFITS

1. The authority citation for part 409 continues to read as follows:

Authority: Secs. 1102 and 1871 of the Social Security Act (42 U.S.C. 1302 and 1395hh).

2. Section 409.30 is amended by revising the introductory text to read as follows:

PART 411—EXCLUSIONS FROM MEDICARE AND LIMITATIONS ON MEDICARE PAYMENT

3. The authority citation for part 411 continues to read as follows:

Authority: Secs. 1102, 1860D-1 through 1860D-42, 1871, and 1877 of the Social Security Act (42 U.S.C. 1302, 1395w-101 through 1395w-152, 1395hh, and 1395nn).

4. Section 411.15 is amended by revising paragraph (p)(3)(iii) to read as follows:

PART 413—PRINCIPLES OF REASONABLE COST REIMBURSEMENT; PAYMENT FOR END-STAGE RENAL DISEASE SERVICES; PROSPECTIVELY DETERMINED PAYMENT RATES FOR SKILLED NURSING FACILITIES; PAYMENT FOR ACUTE KIDNEY INJURY DIALYSIS

5. The authority citation for part 413 continues to read as follows:

Authority: 42 U.S.C. 1302; 42 U.S.C. 1395d(d); 42 U.S.C. 1395f(b); 42 U.S.C. 1395g; 42 U.S.C. 1395l(a), (i), and (n); 42 U.S.C. 1395x(v); 42 U.S.C. 1395hh; 42 U.S.C. 1395rr; 42 U.S.C. 1395tt; 42 U.S.C. 1395ww; sec. 124 of Public Law 106-113, 113 Stat. 1501A-332; sec. 3201 of Public Law 112-96, 126 Stat. 156; sec. 632 of Public Law 112-240, 126 Stat. 2354; sec. 217 of Public Law 113-93, 129 Stat. 1040; sec. 204 of Public Law 113-295, 128 Stat. 4010; and sec. 808 of Public Law 114-27, 129 Stat. 362.

6. The heading for part 413 is revised to read as set forth above.

7. Section 413.333 is amended by revising the definition of “Resident classification system” to read as follows:

8. Section 413.337 is amended by adding paragraph (d)(4) to read as follows:

9. Section 413.338 is added to read as follows:

10. Section 413.345 is revised to read as follows:

11. Section 413.360 is added to subpart J to read as follows:

PART 424—CONDITIONS FOR MEDICARE PAYMENT

12. The authority citation for part 424 continues to read as follows:

Authority: Secs. 1102 and 1871 of the Social Security Act (42 U.S.C. 1302 and 1395hh).

13. In § 424.20—

a. Amend paragraph (a)(1)(ii) by removing the phrase “to one of the Resource Utilization Groups designated” and adding in its place the phrase “one of the case-mix classifiers that CMS designates”; and

b. Amend paragraph (e)(2)(ii)(B)( 2) by removing the reference “§ 483.40(e)” and adding in its place the reference “§ 483.30(e)”.

PART 488—SURVEY, CERTIFICATION, AND ENFORCEMENT PROCEDURES

14. The authority citation for part 488 continues to read as follows:

Authority: Secs. 1102, 1128l, 1864, 1865, 1871 and 1875 of the Social Security Act, unless otherwise noted (42 U.S.C 1302, 1320a-7j, 1395aa, 1395bb, 1395hh) and 1395ll.

15. Section 488.30(a) is amended by revising the definition of “Complaint surveys” to read as follows:

16. Section 488.301 is amended by revising the definition of “Abbreviated standard survey” to read as follows:

17. In § 488.308—

a. Redesignate paragraphs (e)(2) and (3) as paragraphs (f)(1) and (2);

b. Reserve paragraph (e)(2);

c. Add a paragraph heading for new paragraph (f); and

d. Revise newly redesignated paragraph (f)(1) introductory text.

The addition and revision read as follows:

18. Section 488.314 is amended by revising paragraph (a)(1) to read as follows: