Medicare Program; Changes to the Hospital Inpatient Prospective Payment Systems and Fiscal Year 2007 Rates; Fiscal Year 2007 Occupational Mix Adjustment to Wage Index; Health Care Infrastructure Improvement Program; Selection Criteria of Loan Program for Qualifying Hospitals Engaged in Cancer-Related Health Care and Forgiveness of Indebtedness; and Exclusion of Vendor Purchases Made Under the Competitive Acquisition Program (CAP) for Outpatient Drugs and Biologicals Under Part B for the Purpose of Calculating the Average Sales Price (ASP)

AGENCY:

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

ACTION:

Final rules and interim final rule with comment period.

SUMMARY:

We are revising the Medicare hospital inpatient prospective payment systems (IPPS) for operating and capital-related costs to implement changes arising from our continuing experience with these systems, and to implement a number of changes made by the Deficit Reduction Act of 2005 (Pub. L. 109-171). In addition, in the Addendum to this final rule, we describe the changes to the amounts and factors used to determine the rates for Medicare hospital inpatient services for operating costs and capital-related costs. We also are setting forth rate-of-increase limits as well as policy changes for hospitals and hospital units excluded from the IPPS that are paid in full or in part on a reasonable cost basis subject to these limits. These changes are applicable to discharges occurring on or after October 1, 2006.

In this final rule, we discuss public comments we received on our proposals to refine the diagnosis-related group (DRG) system under the IPPS to better recognize severity of illness among patients—to use a hospital-specific relative value (HSRV) cost center weighting methodology to adjust DRG relative weights; and to implement consolidated severity-adjusted DRGs or alternative severity adjustment methods.

Among the other policy changes that we are making are those changes related to: limited revisions of the reclassification of cases to DRGs; the long-term care (LTC)-DRGs and relative weights; the wage data, including the occupational mix data, used to compute the wage index; applications for new technologies and medical services add-on payments; payments to hospitals for the direct and indirect costs of graduate medical education; submission of hospital quality data; payments to sole community hospitals and Medicare-dependent, small rural hospitals; and provisions governing emergency services under the Emergency Medical Treatment and Labor Act of 1986 (EMTALA).

We are responding to requested public comments on a number of other issues that include performance-based hospital payments for services and health information technology, as well as how to improve health data transparency for consumers.

In addition, we are responding to public comments received on a proposed rule issued in the Federal Register on May 17, 2006 that proposed to revise the methodology for calculating the occupational mix adjustment to the wage index for the FY 2007 hospital inpatient prospective payment system by applying an adjustment to 100 percent of the wage index using new 2006 occupational mix survey data collected from hospitals.

We are finalizing two policy documents published in the Federal Register relating to the implementation of the Health Care Infrastructure Improvement Program, a hospital loan program for cancer research, established under the Medicare Prescription Drug, Improvement, and Modernization Act of 2003.

This final rule also revises the definition of the term “unit” to specify the exclusion of units of drugs sold to approved Medicare Competitive Acquisition Program (CAP) vendors for use under the CAP from average sales price (ASP) calculations for a period of up to 3 years, at which time we will reevaluate our policy.

DATES:

Effective Dates: The provisions of these final rules are effective on October 1, 2006, with the exception of the provisions in § 412.8, § 414.802, and the procedures for withdrawing or terminating reclassifications established in section III.H.4. of the preamble. The provisions of § 412.8, § 414.802, and the procedures for withdrawing or terminating reclassifications established in section II.H.4. of the preamble are effective August 18, 2006. This rule is a major rule as defined in 5 U.S.C. 804(2). Pursuant to 5 U.S.C. 801(a)(1)(A), we are submitting a report to the Congress on this rule on August 1, 2006.

Comment Date: We will consider comments on the exclusion of CAP drugs from the ASP calculation (§ 414.802) as discussed in section XII. of the preamble of this final rule, if we receive them at one of the addresses provided below, no later than 5 p.m. on October 2, 2006.

ADDRESSES:

In commenting, on section XII. of this rule, please refer to file code CMS-1325-IFC4.

Because of staff and resource limitations, we cannot accept comments by facsimile (FAX) transmission.

You may submit comments in one of four ways (no duplicates, please):

1. Electronically. You may submit electronic comments on specific issues in this regulation to http://www.cms.hhs.gov/eRulemaking. Click on the link “Submit electronic comments on CMS regulations with an open comment period.” (Attachments should be in Microsoft Word, WordPerfect, or Excel; however, we prefer Microsoft Word.)

2. By regular mail. You may mail written comments (one original and two copies) to the following address ONLY: Centers for Medicare & Medicaid Services, Department of Health and Human Services, Attention: CMS-1325-IFC4, P.O. Box 8011, Baltimore, MD 21244-1850.

Please allow sufficient time for mailed comments to be received before the close of the comment period.

3. By express or overnight mail. You may send written comments (one original and two copies) to the following address ONLY: Centers for Medicare & Medicaid Services, Department of Health and Human Services, Attention: CMS-1325-IFC4, Mail Stop C4-26-05, 7500 Security Boulevard, Baltimore, MD 21244-1850.

4. By hand or courier. If you prefer, you may deliver (by hand or courier) your written comments (one original and two copies) before the close of the comment period to one of the following addresses. If you intend to deliver your comments to the Baltimore address, please call telephone number (410) 786-7195 in advance to schedule your arrival with one of our staff members. Room 445-G, Hubert H. Humphrey Building, 200 Independence Avenue, SW., Washington, DC 20201; or 7500 Security Boulevard, Baltimore, MD 21244-1850.

(Because access to the interior of the HHH Building is not readily available to persons without Federal Government identification, commenters are encouraged to leave their comments in the CMS drop slots located in the main lobby of the building. A stamp-in clock is available for persons wishing to retain a proof of filing by stamping in and retaining an extra copy of the comments being filed.)

Comments mailed to the addresses indicated as appropriate for hand or courier delivery may be delayed and received after the comment period.

For information on viewing public comments, see the beginning of the SUPPLEMENTARY INFORMATION section.

FOR FURTHER INFORMATION CONTACT:

Marc Hartstein, (410) 786-4548, Operating Prospective Payment, Diagnosis-Related Groups (DRGs), Wage Index, Occupational Mix Adjustment, New Medical Services and Technology Add-On Payments, Hospital Geographic Reclassifications, Sole Community Hospital, Disproportionate Share Hospital, and Medicare-Dependent, Small Rural Hospital Issues.

Tzvi Hefter, (410) 786-4487, Capital Prospective Payment, Excluded Hospitals, Graduate Medical Education, Critical Access Hospitals, Long-Term Care (LTC)-DRGs, and Terms of Hospital Loans under Health Care Infrastructure Improvement Program Issues.

Siddhartha Mazumdar, (410) 786-6673, Rural Community Hospital Demonstration Issues.

Sheila Blackstock, (410) 786-3502, Quality Data for Annual Payment Update Issues.

Thomas Valuck, (410) 786-7479, Hospital Value-Based Purchasing Issues.

Frederick Grabau, (410) 786-0206, Services in Foreign Hospitals Issues.

Brian Reitz, (410) 786-5001, Obsolete Paper Claims Forms Issues.

Melinda Jones, (410) 786-7069, Loan Forgiveness Criteria for Health Care Infrastructure Improvement Program.

Corinne Axelrod, (410) 786-5620, Competitive Acquisition Program (CAP) for Part B Drugs Issues.

Angela Mason, (410) 786-7452, Payment for Covered Outpatient Drugs and Biologicals Issues.

Submitting Comments: We welcome comments from the public on all issues set forth in this rule to assist us in fully considering issues and developing policies. You can assist us by referencing the file code CMS-1325-IFC4 and the specific “issue identifier” that precedes the section on which you choose to comment.

Inspection of Public Comments: All comments received before the close of the comment period are available for viewing by the public, including any personally identifiable or confidential business information that is included in a comment. We post all comments received before the close of the comment period on a public Web site as soon as possible after they are received: http://www.cms.hhs.gov/eRulemaking. Clink on the link “Electronic Comments on CMS Regulations” on that Web site to view public comments.

Comments received timely will also be available for public inspection as they are received, generally beginning approximately 3 weeks after publication of a document, at the headquarters of the Centers for Medicare & Medicaid Services, 7500 Security Boulevard, Baltimore, Maryland 21244, Monday through Friday of each week from 8:30 a.m. to 4 p.m. To schedule an appointment to view public comments, phone 1-800-743-3951.

Electronic Access

This Federal Register document is also available from the Federal Register online database through GPO Access, a service of the U.S. Government Printing Office. Free public access is available on a Wide Area Information Server (WAIS) through the Internet and via asynchronous dial-in. Internet users can access the database by using the World Wide Web; the Superintendent of Documents' home page address is http://www.gpoaccess.gov/,, by using local WAIS client software, or by telnet to swais.access.gpo.gov, then login as guest (no password required). Dial-in users should use communications software and modem to call (202) 512-1661; type swais, then login as guest (no password required).

Acronyms

AHA American Hospital Association

AHIMA American Health Information Management Association

AHRO Agency for Health Care Research and Quality

AMI Acute myocardial infarction

AOA American Osteopathic Association

APR DRG All Patient Refined Diagnosis-Related Group System

ASC Ambulatory surgical center

ASP Average sales price

AWP Average wholesale price

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

BBRA Medicare, Medicaid, and SCHIP [State Children's Health Insurance Program] Balanced Budget Refinement Act of 1999, Pub. L. 106-113

BIPA Medicare, Medicaid, and SCHIP [State Children's Health Insurance Program] Benefits Improvement and Protection Act of 2000, Pub. L. 106-554

BLS Bureau of Labor Statistics

AH Critical access hospital

AP Competitive Acquisition Program

CART CMS Abstraction & Reporting Tool

CBSAs Core-based statistical areas

CC Complication or comorbidity

CDAC Clinical Data Abstraction Center

CIPI Capital input price index

CPI Consumer price index

CMI Case-mix index

CMS Centers for Medicare & Medicaid Services

CMSA Consolidated Metropolitan Statistical Area

COBRA Consolidated Omnibus Reconciliation Act of 1985, Pub. L. 99-272

CPI Consumer price index

CRNA Certified registered nurse anesthetist

CY Calendar year

DRA Deficit Reduction Act of 2005, Pub. L. 109-171

DRG Diagnosis-related group

DSH Disproportionate share hospital

ECI Employment cost index

EMR Electronic medical record

EMTALA Emergency Medical Treatment and Labor Act of 1986, Pub. L. 99-272

FDA Food and Drug Administration

FFY Federal fiscal year

FIPS Federal information processing standards

FQHC Federally qualified health center

FTE Full-time equivalent

FY Fiscal year

GAAP Generally Accepted Accounting Principles

GAF Geographic Adjustment Factor

GME Graduate medical education

HCAHPS Hospital Consumer Assessment of Healthcare Providers and Systems

HCFA Health Care Financing Administration

HCRIS Hospital Cost Report Information System

HHA Home health agency

HHS Department of Health and Human Services

HIC Health insurance card

HIPAA Health Insurance Portability and Accountability Act of 1996, Pub. L. 104-191

HIPC Health Information Policy Council

HIS Health information system

HIT Health information technology

HMO Health maintenance organization

HSA Health savings account

HSCRC Maryland Health Services Cost Review Commission

HSRV Hospital-specific relative value

HSRVcc Hospital-specific relative value cost center

HQA Hospital Quality Alliance

HQI Hospital Quality Initiative

HwH Hospital-within-a-hospital

ICD-9-CM International Classification of Diseases, Ninth Revision, Clinical Modification

ICD-10-PCS International Classification of Diseases, Tenth Edition, Procedure Coding System

ICU Intensive care unit

IHS Indian Health Service

IME Indirect medical education

IOM Institute of Medicine

IPF Inpatient psychiatric facility

IPPS Acute care hospital inpatient prospective payment system

IRF Inpatient rehabilitation facility

IRP Initial residency period

JCAHO Joint Commission on Accreditation of Healthcare Organizations

LAMCs Large area metropolitan counties

LTC-DRG Long-term care diagnosis-related group

LTCH Long-term care hospital

MCE Medicare Code Editor

MCO Managed care organization

MCV Major cardiovascular condition

MDC Major diagnostic category

MDH Medicare-dependent, small rural hospital

MedPAC Medicare Payment Advisory Commission

MedPAR Medicare Provider Analysis and Review File

MEI Medicare Economic Index

MGCRB Medicare Geographic Classification Review Board

MMA Medicare Prescription Drug, Improvement, and Modernization Act of 2003, Pub. L. 108-173

MRHFP Medicare Rural Hospital Flexibility Program

MSA Metropolitan Statistical Area

NAICS North American Industrial Classification System

NCD National coverage determination

NCHS National Center for Health Statistics

NCQA National Committee for Quality Assurance

NCVHS National Committee on Vital and Health Statistics

NECMA New England County Metropolitan Areas

NICU Neonatal intensive care unit

NQF National Quality Forum

NTIS National Technical Information Service

NVHRI National Voluntary Hospital Reporting Initiative

OES Occupational employment statistics

OIG Office of the Inspector General

OMB Executive Office of Management and Budget

O.R. Operating room

OSCAR Online Survey Certification and Reporting (System)

PRM Provider Reimbursement Manual

PPI Producer price index

PMSAs Primary metropolitan statistical areas

PPS Prospective payment system

PRA Per resident amount

ProPAC Prospective Payment Assessment Commission

PRRB Provider Reimbursement Review Board

PS&R Provider Statistical and Reimbursement (System)

QIG Quality Improvement Group, CMS

QIO Quality Improvement Organization

RHC Rural health clinic

RHQDAPU Reporting hospital quality data for annual payment update

RNHCI Religious Nonmedical Health Care Institution

RRC Rural referral center

RUCAs Rural-urban commuting area codes

RY Rate year

SAF Standard Analytic File

SCH Sole community hospital

SFY State fiscal year

SIC Standard Industrial Classification

SNF Skilled nursing facility

SOCs Standard occupational classifications

SOM State Operations Manual

SSA Social Security Administration

SSI Supplemental Security Income

TAG Technical Advisory Group

TEFRA Tax Equity and Fiscal Responsibility Act of 1982, Pub. L. 97-248

UHDDS Uniform hospital discharge data set

Table of Contents

I. Background

A. Summary

1. Acute Care Hospital Inpatient Prospective Payment System (IPPS)

2. Hospitals and Hospital Units Excluded from the IPPS

a. Inpatient Rehabilitation Facilities (IRFs)

b. Long-Term Care Hospitals (LTCHs)

c. Inpatient Psychiatric Facilities (IPFs)

3. Critical Access Hospitals (CAHs)

4. Payments for Graduate Medical Education (GME)

B. Provisions of the Deficit Reduction Act of 2005 (DRA)

C. Summary of the Provisions of the FY 2007 IPPS and FY 2007 Occupational Mix Adjustment to the Wage Index Proposed Rules

1. DRG Reclassifications and Recalibrations of Relative Weights

2. Changes to the Hospital Wage Index

3. Other Decisions and Changes to the IPPS for Operating Costs, GME Costs, and Promoting Hospitals’ Effective Use of Health Information Technology

4. Changes to the PPS for Capital-Related Costs

5. Changes for Hospitals and Hospital Units Excluded from the IPPS

6. Payments for Services Furnished Outside the United States

7. Payment for Blood Clotting Factor Administered to Inpatients with Hemophilia

8. Limitation on Payments to Skilled Nursing Facilities for Bad Debt

9. Determining Prospective Payment Operating and Capital Rates and Rate-of-Increase Limits

10. Impact Analysis

11. Recommendation of Update Factors for Operating Cost Rates of Payment for Inpatient Hospital Services

12. Discussion of Medicare Payment Advisory Commission Recommendations

13. Appendix C and Appendix D

D. Public Comments Received in Response to the FY 2007 IPPS and FY 2007 Occupational Mix Adjustment to the Wage Index Proposed Rules

E. Interim Final Rule on Selection Criteria of Loan Program for Qualifying Hospitals Engaged in Cancer-Related Health Care

F. Proposed Rule on Forgiveness of Indebtedness under the Health Care Infrastructure Improvement Program

G. Interim Final Rule on the Exclusion of Vendor Purchases Made Under the Competitive Acquisition Program for Part B Outpatient Drugs and Biologicals for the Purpose of Calculating the Average Sales Price

II. Changes to DRG Classifications and Relative Weights

A. Background

B. DRG Reclassifications

1. General

2. Yearly Review for Making DRG Changes

C. Revisions to the DRG System Used Under the IPPS

1. MedPAC Recommendations

2. Refinement of the Relative Weight Calculation

3. Refinement of DRGs Based on Severity of Illness

a. Comparison of the CMS DRG System and the APR DRG System

b. CS DRGs for Use in the IPPS

c. Changes to CMI from a New DRG System

4. Effect of CS DRGs on the Outlier Threshold

5. Impact of Refinement of DRG System on Payments

6. Conclusions

7. Severity Refinement to CMS DRGs

a. MDC 1 (Diseases and Disorders of the Nervous System)

b. MDC 4 (Diseases and Disorders of the Respiratory System): Respiratory System Diagnosis with Ventilator Support

c. MDC 6 (Diseases and Disorders of the Digestive System)

d. MDC 11 (Diseases and Disorders of the Kidney and Urinary Tract): Major Bladder Procedures

e. MDC 16 (Diseases and Disorders of the Blood and Blood Forming Organs and Immunological Disorders): Major Hematological and Immunological Diagnoses

f. MDC 18 (Infectious and Parasitic Diseases (Systemic or Unspecified Sites)): O.R. Procedure for Patients with Infectious and Parasitic Diseases

g. Severe Sepsis

D. Changes to Specific DRG Classifications

1. Pre-MDCs

a. Heart Transplant or Implant of Heart Assist System: Addition of Procedure to DRG 103

b. Pancreas Transplants

2. MDC 1 (Diseases and Disorders of the Nervous System)

a. Implantation of Intracranial Neurostimulator System for Deep Brain Stimulation (DBS)

b. Carotid Artery Stents

3. MDC 5 (Diseases and Disorders of the Circulatory System)

a. Insertion of Epicardial Leads for Defibrillator Devices

b. Application of Major Cardiovascular Diagnoses (MCVs) List to Defibrillator DRGs

4. MDC 8 (Diseases and Disorders of the Musculoskeletal System and Connective Tissue)

a. Hip and Knee Replacements

b. Spinal Fusion

c. CHARITE^TM Spinal Disc Replacement Device

5. MDC 18 (Infectious and Parasitic Diseases (Systemic or Unspecified Sites)): Severe Sepsis

6. Medicare Code Editor (MCE) Changes

a. Edit: Newborn Diagnoses

b. Edit: Diagnoses for Pediatric—Age 0-17 Years Old

c. Edit: Maternity Diagnoses—Age 12 through 55

d. Edit: Diagnoses Allowed for Females Only

e. Edit: Diagnoses Allowed for Males Only

f. Edit: Procedures Allowed for Females Only

g. Edit: Manifestations Not Allowed as Principal Diagnosis

h. Edit: Nonspecific Principal Diagnosis

i. Edit: Unacceptable Principal Diagnosis

j. Edit: Nonspecific O.R. Procedures

k. Edit: Noncovered Procedures

l. Edit: Bilateral Procedure

7. Surgical Hierarchies

8. Refinement of Complications and Comorbidities (CC) List

a. Background

b. Comprehensive Review of the CC List

c. CC Exclusions List for FY 2007

9. Review of Procedure Codes in DRGs 468, 476, and 477

a. Moving Procedure Codes from DRG 468 or DRG 477 to MDCs

b. Reassignment of Procedures among DRGs 468, 476, and 477

c. Adding Diagnosis or Procedure Codes to MDCs

10. Changes to the ICD-9-CM Coding System

11. Other Issues

a. Chronic Kidney Disease

b. Bronchial Valve

c. Female Reproductive System Reconstruction Procedures

d. Devices That are Replaced Without Cost or Where Credit for a Replaced Device is Furnished to the Hospital

E. Recalibration of DRG Weights

F. LTC-DRG Reclassifications and Relative Weights for LTCHs for FY 2007

1. Background

2. Changes in the LTC-DRG Classifications

a. Background

b. Patient Classifications into DRGs

3. Development of the FY 2007 LTC-DRG Relative Weights

a. General Overview of Development of the LTC-DRG Relative Weights

b. Data

c. Hospital-Specific Relative Value Methodology

d. Low-Volume LTC-DRGs

4. Steps for Determining the FY 2007 LTC-DRG Relative Weights

5. Summary of Public Comments and Departmental Responses

G. Add-On Payments for New Services and Technologies

1. Background

2. Public Input Before Publication of a Notice of Proposed Rulemaking on Add-On Payments

3. FY 2007 Status of Technologies Approved for FY 2006 Add-On Payments

a. Kinetra® Implantable Neurostimulator (Kinetra®) for Deep Brain Stimulation

b. Endovascular Graft Repair of the Thoracic Aorta

c. Restore® Rechargeable Implantable Neurostimulator

4. FY 2007 Applications for New Technology Add-On Payments

a. C-Port® Distal Anastomosis System

b. NovoSeven® for Intracerebral Hemorrhage

c. X STOP Interspinous Process Decompression System

5. Interim and Final Cost Threshold Tables Due to Changes to Wage Index and Budget Neutrality Factors

III. Changes to the Hospital Wage Index

A. Background

B. Core-Based Statistical Areas for the Hospital Wage Index

C. Occupational Mix Adjustment to the FY 2007 Wage Index

1. Development of Data for the FY 2007 Occupational Mix Adjustment

2. Timeline for the Collection, Review, and Correction of the Occupational Mix Data

3. Calculation of the Occupational Mix Adjustment

D. Worksheet S-3 Wage Data for the FY 2007 Wage Index Update

E. Verification of Worksheet S-3 Wage Data

F. Computation of the FY 2007 Unadjusted Wage Index

G. Implementation of the FY 2007 Occupational Mix Adjustment to the Wage Index

H. Revisions to the Wage Index Based on Hospital Redesignations

1. General

2. Effects of Reclassification/Redesignation

3. FY 2007 MGCRB Reclassifications

4. Procedures for Hospitals Applying for Reclassification Effective in FY 2008 and Reinstating Reclassifications in FY 2008

5. FY 2007 Redesignations Under Section 1886(d)(8)(B) of the Act

6. Reclassifications Under Section 508 of Pub. L. 108-173

7. Wage Indices for Reclassified Hospitals and Reclassification Budget Neutrality Factor

I. FY 2007 Wage Index Adjustment Based on Commuting Patterns of Hospital Employees

J. Process for Requests for Wage Index Data Corrections

K. Labor-Related Share for the Wage Index for FY 2007

L. Proxy for the Hospital Market Basket

IV. Other Decisions and Changes to the IPPS for Operating Costs and GME Costs

A. Reporting of Hospital Quality Data for Annual Hospital Payment Update

1. Background

2. New Procedures for Hospital Reporting of Quality Data

a. Two Percentage Point Reduction

b. New Procedures

c. Expanded Quality Measures

d. HCAHPS® Survey

e. Data Submission

f. RHQDAPU Program Withdrawal and Chart Validation Requirements

g. Data Validation and Attestation

h. Public Display and Reconsideration Procedures

i. Conclusion

3. Electronic Medical Records

B. Value-Based Purchasing

1. Introduction

2. Premier Hospital Quality Incentive Demonstration

3. RHQDAPU Program

a. Section 501(b) of Pub. L. 108-173 (MMA)

b. Section 5001(a) of Pub. L. 109-171 (DRA)

4. Plan for Implementing Hospital Value-Based Purchasing Beginning with FY 2009

a. Measure Development and Refinement

b. Data Infrastructure

c. Incentive Methodology

d. Public Reporting

5. Considerations Related to Certain Conditions, Including Hospital-Acquired Infections

6. Promoting Effective Use of Health Information Technology

C. Sole Community Hospitals (SCHs) and Medicare-Dependent, Small Rural Hospitals (MDHs)

1. Background

2. Volume Decrease Adjustment for SCHs and MDHs

a. HAS/Monitrend Data

b. HAS/Monitrend Data Book Replacement Alternative

3. Mandatory Reporting Requirements for Any Changes in the Circumstances Under Which a Hospital Was Designated as an SCH or MDH

4. Payment Changes for MDHs under the DRA of 2005

a. Background

b. Regulation Changes

5. Technical Change

D. Rural Referral Centers

1. Case-Mix Index

2. Discharges

E. Indirect Medical Education (IME) Adjustment

1. Background

2. IME Adjustment Factor for FY 2007

3. Technical Change to Revise Cross-Reference

F. Payment Adjustment for Disproportionate Share Hospitals (DSHs)

1. Background

2. Technical Corrections

3. Reinstatement of Inadvertently Deleted Provisions on DSH Payment Adjustment Factors

4. Enhanced DSH Adjustment for MDHs

G. Geographic Reclassifications

1. Background

2. Reclassifications under Section 508 of Pub. L. 108-173

3. Multicampus Hospitals

4. Urban Group Hospital Reclassifications

5. Effect of Change of Ownership on Urban County Group Reclassifications

6. Requested Reclassification for Hospitals Located in a Single Hospital MSA Surrounded by Rural Counties

7. Special Adjustment for the Hospital Group Reclassification Denied on the Basis of Incomplete CSA Listing

H. Payment for Direct Graduate Medical Education

1. Background

2. Determination of Weighted Average Per Resident Amounts (PRAs) for Merged Teaching Hospitals

3. Determination of Per Resident Amounts (PRAs) for New Teaching Hospitals

4. Requirements for Counting and Appropriate Documentation of FTE Residents: Clarification

5. Resident Time Spent in Nonpatient Care Activities as Part of Approved Residency Programs

6. Medicare GME Affiliated Groups: Technical Changes to Regulations

I. Payment for the Costs of Nursing and Allied Health Education Activities: Clarification

J. Hospital Emergency Services under EMTALA

1. Background

2. Role of the EMTALA Technical Advisory Group (TAG)

3. Definition of “Labor”

4. Application of EMTALA Requirements to Hospitals Without Dedicated Emergency Departments

5. Clarification of Reference to “Referral Centers”

K. Other Technical Changes

1. Cross-Reference Correction in Regulations on Limitations on Beneficiary Charges

2. Cross-Reference Corrections in Regulations on Payment Denials Based on Admissions and Quality Reviews

3. Cross-Reference Correction in Regulations on Outlier Payments

4. Removing References to Two Paper Claims Forms

L. Rural Community Hospital Demonstration Program

M. Health Care Information Transparency Initiative

V. Changes to the PPS for Capital-Related Costs

A. Background

B. Treatment of Certain Urban Hospitals Reclassified as Rural Hospitals Under § 412.103

C. Other Technical Corrections Relating to the Capital PPS Geographic Adjustment Factors

VI. Changes for Hospitals and Hospital Units Excluded from the IPPS

A. Payments to Excluded Hospitals and Hospital Units

1. Payments to Existing Excluded and New Hospitals and Hospital Units

2. Separate PPS for IRFs

3. Separate PPS for LTCHs

4. Separate PPS for IPFs

5. Grandfathering of Hospitals-Within-Hospitals (HwHs) and Satellite Facilities

6. Changes to the Methodology for Determining LTCH Cost-to-Charge Ratios (CCRs) and the Reconciliation of High-Cost and Short-Stay Outlier Payments under the LTCH PPS

a. Background

b. High-Cost Outliers

c. Short-Stay Outliers

d. CCR Ceiling

e. Statewide Average CCRs

f. Data Used to Determine a CCR

g. Reconciliation of Outlier Payments Upon Cost report Settlement

7. Technical Corrections Relating to LTCHs

8. Cross-Reference Correction in Authority Citations for 42 CFR 412 and 413

9. Report of Adjustment (Exceptions) Payments

B. Critical Access Hospitals (CAHs)

1. Background

2. Sunset of Designation of CAHs as Necessary Providers: Technical Correction

VII. Payment for Services Furnished Outside the United States

A. Background

B. Proposed Clarification of Regulations

VIII. Payment for Blood Clotting Factor Administered to Inpatients with Hemophilia

IX. Limitation on Payments to Skilled Nursing Facilities for Bad Debt

A. Background

B. Changes Made by Section 5004 of Pub. L. 109-171

C. Proposed Regulation Changes

X. MedPAC Recommendations

XI. Health Care Infrastructure Improvement Program: Selection Criteria for Loan Program for Qualifying Hospitals Engaged in Cancer-Related Health Care and Forgiveness of Indebtedness

A. Background

B. Issuance of an Interim Final Rule with Comment Period and a Proposed Regulation

C. Provisions of the Interim Final Rule With Comment Period

1. Loan Qualifying Criteria

2. Selection Criteria

3. Terms of the Loan

4. Public Comments Received on the Interim Final Rule With Comment Period

5. Provisions of this Final Rule

D. Proposed Rule on Forgiveness of Indebtedness

1. Conditions for Loan Forgiveness

2. Plan Criteria for Meeting the Conditions for Loan Forgiveness

3. Public Comments Received on the Proposed Rule and Our Responses

4. Provisions of the Final Rule

E. Statutory Requirements for Issuance of Regulations

XII. Exclusion of Vendor Purchases Made Under the Competitive Acquisition Program (CAP) for Outpatient Drugs and Biologicals Under Part B for the Purpose of Calculating the Average Sales Price (ASP)

A. Background

1. Average Sales Price (ASP)

2. Competitive Acquisition Program (CAP)

3. Regulatory History

B. Regulation Change

XIII. Other Required Information

A. Requests for Data from the Public

B. Collection of Information Requirements

C. Waiver of Proposed Rulemaking and Delay in the Effective Date

D. Response to Comments

Regulation Text

Addendum—Schedule of Tentative Standardized Amounts, Tentative Update Factors and Rate-of-Increase Percentages Effective With Cost Reporting Periods Beginning On or After October 1, 2006

I. Summary and Background

II. Changes to Prospective Payment Rates for Hospital Inpatient Operating Costs

A. Calculation of the Tentative Adjusted Standardized Amount

1. Standardization of Base-Year Costs or Target Amounts

2. Computing the Tentative Average Standardized Amount

3. Updating the Tentative Average Standardized Amount

4. Other Adjustments to the Average Standardized Amount

a. Recalibration of DRG Weights and Updated Wage Index—Budget Neutrality Adjustment

b. Reclassified Hospitals—Tentative Budget Neutrality Adjustment

c. Outliers

d. Tentative Rural Community Hospital Demonstration Program Adjustment (Section 410A of Pub. L. 108-173)

5. Tentative FY 2007 Standardized Amount

B. Tentative Adjustments for Area Wage Levels and Cost-of-Living

1. Tentative Adjustment for Area Wage Levels

2. Final Adjustment for Cost-of-Living in Alaska and Hawaii

C. DRG Relative Weights

D. Calculation of the Prospective Payment Rates

1. Federal Rate

2. Hospital-Specific Rate (Applicable Only to SCHs and MDHs)

a. Calculation of Hospital-Specific Rate

b. Updating the FY 1982, FY 1987, FY 1996, and FY 2002 Hospital-Specific Rates for FY 2007

3. General Formula for Calculation of Prospective Payment Rates for Hospitals Located in Puerto Rico Beginning On or After October 1, 2006, and Before October 1, 2007

a. Puerto Rico Rate

b. National Rate

III. Changes to Payment Rates for Acute Care Hospital Inpatient Capital-Related Costs for FY 2007

A. Determination of Federal Hospital Inpatient Capital-Related Prospective Payment Rate Update

1. Projected Capital Standard Federal Rate Update

a. Description of the Update Framework

b. Comparison of CMS and MedPAC Update Recommendation

2. Outlier Payment Adjustment Factor

3. Budget Neutrality Adjustment Factor for Changes in DRG Classifications and Weights and the GAF

4. Exceptions Payment Adjustment Factor

5. Capital Standard Federal Rate for FY 2007

6. Special Capital Rate for Puerto Rico Hospitals

B. Calculation of the Inpatient Capital-Related Prospective Payments for FY 2007

C. Capital Input Price Index

1. Background

2. Forecast of the CIPI for FY 2007

IV. Payment Rates for Excluded Hospitals and Hospital Units: Rate-of-Increase Percentages

A. Payments to Existing Excluded Hospitals and Units

B. New Excluded Hospitals and Units

V. Payment for Blood Clotting Factor Administered to Inpatients with Hemophilia

Tables

The following tables are included as part of this final rule:

Table 1A—National Adjusted Operating Standardized Amounts, Labor/Nonlabor (69.7 Percent Labor Share/30.3 Percent Nonlabor Share If Wage Index Is Greater Than 1) (Tentative)

Table 1B—National Adjusted Operating Standardized Amounts, Labor/Nonlabor (62 Percent Labor Share/38 Percent Nonlabor Share If Wage Index Is Less Than or Equal to 1) (Tentative)

Table 1C—Adjusted Operating Standardized Amounts for Puerto Rico, Labor/Nonlabor (Tentative)

Table 1D—Capital Standard Federal Payment Rate (Tentative)

Table 4J—Out-Migration Wage Adjustment—FY 2007 (Tentative)

Table 5—List of Diagnosis-Related Groups (DRGs), Relative Weighting Factors, and Geometric and Arithmetic Mean Length of Stay (LOS) (Tentative)

Table 6A—New Diagnosis Codes

Table 6B—New Procedure Codes

Table 6C—Invalid Diagnosis Codes

Table 6D—Invalid Procedure Codes

Table 6E—Revised Diagnosis Code Titles

Table 6F—Revised Procedure Code Titles

Table 6G—Additions to the CC Exclusions List

Table 6H—Deletions from the CC Exclusions List

Table 7A—Medicare Prospective Payment System Selected Percentile Lengths of Stay: FY 2005 MedPAR Update March 2006 GROUPER V23.0

Table 7B—Medicare Prospective Payment System Selected Percentile Lengths of Stay: FY 2005 MedPAR Update March 2006 GROUPER V24.0

Table 8A—Statewide Average Operating Cost-to-Charge Ratios—July 2006

Table 8B—Statewide Average Capital Cost-to-Charge Ratios—July 2006

Table 8C— Statewide Average Total Cost-to-Charge Ratios for LTCHs—July 2006

Table 9A—Hospital Reclassifications and Redesignations by Individual Hospital and CBSA for FY 2007 (Tentative)

Table 9B—Hospital Reclassifications and Redesignation by Individual Hospital Under Section 508 of Pub. L. 108-173 for FY 2007 (Tentative)

Table 9C—Hospitals Redesignated as Rural under Section 1886(d)(8)(E) of the Act for FY 2007 (Tentative)

Table 10—Geometric Mean Plus the Lesser of .75 of the National Adjusted Operating Standardized Payment Amount (Increased to Reflect the Difference Between Costs and Charges) or .75 of One Standard Deviation of Mean Charges by Diagnosis-Related Group (DRG)—July 2006 (Tentative)

Table 11—FY 2007 LTC-DRGs, Relative Weights, Geometric Average Length of Stay, and 5/6ths of the Geometric Average Length of Stay

Appendix A—Regulatory Impact Analysis

I. Overall Impact

II. Objectives

III. Limitations on Our Analysis

IV. Hospitals Included In and Excluded From the IPPS

V. Effects on Excluded Hospitals and Hospital Units

VI. Quantitative Effects of the Policy Changes Under the IPPS for Operating Costs

A. Basis and Methodology of Estimates

B. Analysis of Table I

C. Effects on the Hospitals that Failed the Quality Data Submission Process (Column 2)

D. Effects of the DRA Provision Related to MDHs (Column 3)

E. Effects of the Changes to the DRG Reclassifications and Relative Cost-Based Weights (Column 4)

F. Effects of Wage Index Changes (Column 5)

G. Combined Effects of DRG and Wage Index Changes, Including Budget Neutrality Adjustment (Column 6)

H. Effects of the 3-Year Provision Allowing Urban Hospitals that Were Converted to Rural as a Result of the FY 2005 Labor Market Area Changes to Maintain the Wage Index of the Urban Labor Market Area in Which They Were Formerly Located (Column 7)

I. Effects of MGCRB Reclassifications (Column 8)

J. Effects of the Wage Index Adjustment for Out-Migration (Column 9)

K. Effects of All Changes (Column 10)

L. Effects of Policy on Payment Adjustments for Low-Volume Hospitals

M. Impact Analysis of Table II

VII. Effects of Other Policy Changes

A. Effects of LTC-DRG Reclassifications and Relative Weights for LTCHs

B. Effects of New Technology Add-On Payments

C. Effects of Requirements for Hospital Reporting of Quality Data for Annual Hospital Payment Update

D. Effects of Other Policy Changes Affecting Sole Community Hospitals (SCHs) and Medicare-Dependent, Small Rural Hospitals (MDHs)

E. Effects of Policy on Payment for Direct Costs of Graduate Medical Education

1. Determination of Weighted Average GME PRAs for Merged Teaching Hospitals

2. Determination of PRAs for New Teaching Hospitals

3. Requirements for Counting and Appropriate Documentation of FTE Residents

4. Resident Time Spent in Nonpatient Care Activities as Part of an Approved Residency Program

F. Effects of Policy Changes Relating to Emergency Services under EMTALA

G. Effects of Policy on Rural Community Hospital Demonstration Program

H. Effects of Policy on Hospitals-within-Hospitals and Satellite Facilities

I. Effects of Policy Changes to the Methodology for Determining LTCH CCRs and the Reconciliation of LTCH PPS Outlier Payments

J. Effects of Policy on Payment for Services Furnished Outside the United States

K. Effects of Final Policy on Limitation on Payments to SNFs

L. Effects of Policy on CAP for Outpatient Drugs and Biologicals under Part B for the Purpose of Calculating the ASP

VIII. Impact of Changes in the Capital PPS

A. General Considerations

B. Results

IX. Impact of Changes Relating to the Loan Program for Capital Cost under the Health Care Infrastructure Improvement Program

A. Effects on Hospitals

B. Effects on the Medicare and Medicaid Programs

X. Alternatives Considered

XI. Overall Conclusion

XII. Accounting Statement

XIII. Executive Order 12866

Appendix B—Recommendation of Update Factors for Operating Cost Rates of Payment for Inpatient Hospital Services

I. Background

II. Secretary's Final Recommendation for Updating the Prospective Payment System Standardized Amounts

III. Secretary's Final Recommendation for Updating the Rate-of-Increase Limits for Excluded Hospitals and Hospital Units

IV. Secretary's Recommendation for Updating the Capital Prospective Payment Amounts

I. Background

A. Summary

1. Acute Care Hospital Inpatient Prospective Payment System (IPPS)

Section 1886(d) of the Social Security Act (the Act) sets forth a system of payment for the operating costs of acute care hospital inpatient stays under Medicare Part A (Hospital Insurance) based on prospectively set rates. Section 1886(g) of the Act requires the Secretary to pay for the capital-related costs of hospital inpatient stays under a prospective payment system (PPS). Under these PPSs, Medicare payment for hospital inpatient operating and capital-related costs is made at predetermined, specific rates for each hospital discharge. Discharges are classified according to a list of diagnosis-related groups (DRGs).

The base payment rate is comprised of a standardized amount that is divided into a labor-related share and a nonlabor-related share. The labor-related share is adjusted by the wage index applicable to the area where the hospital is located; and if the hospital is located in Alaska or Hawaii, the nonlabor-related share is adjusted by a cost-of-living adjustment factor. This base payment rate is multiplied by the DRG relative weight.

If the hospital treats a high percentage of low-income patients, it receives a percentage add-on payment applied to the DRG-adjusted base payment rate. This add-on payment, known as the disproportionate share hospital (DSH) adjustment, provides for a percentage increase in Medicare payments to hospitals that qualify under either of two statutory formulas designed to identify hospitals that serve a disproportionate share of low-income patients. For qualifying hospitals, the amount of this adjustment may vary based on the outcome of the statutory calculations.

If the hospital is an approved teaching hospital, it receives a percentage add-on payment for each case paid under the IPPS, known as the indirect medical education (IME) adjustment. This percentage varies, depending on the ratio of residents to beds.

Additional payments may be made for cases that involve new technologies or medical services that have been approved for special add-on payments. To qualify, a new technology or medical service must demonstrate that it is a substantial clinical improvement over technologies or services otherwise available, and that, absent an add-on payment, it would be inadequately paid under the regular DRG payment.

The costs incurred by the hospital for a case are evaluated to determine whether the hospital is eligible for an additional payment as an outlier case. This additional payment is designed to protect the hospital from large financial losses due to unusually expensive cases. Any outlier payment due is added to the DRG-adjusted base payment rate, plus any DSH, IME, and new technology or medical service add-on adjustments.

Although payments to most hospitals under the IPPS are made on the basis of the standardized amounts, some categories of hospitals are paid the higher of a hospital-specific rate based on their costs in a base year (the higher of FY 1982, FY 1987, FY 1996, or FY 2002) or the IPPS rate based on the standardized amount. For example, sole community hospitals (SCHs) are the sole source of care in their areas, and Medicare-dependent, small rural hospitals (MDHs) are a major source of care for Medicare beneficiaries in their areas. Both of these categories of hospitals are afforded special payment protection in order to maintain access to services for beneficiaries. (Through FY 2007, an MDH receives the IPPS rate plus 50 percent of the difference between the IPPS rate and its hospital-specific rate if the hospital-specific rate is higher than the IPPS rate. In addition, an MDH may not use FY 1996 as its base year for the hospital-specific rate. As discussed below, for discharges occurring on or after October 1, 2007, but before October 1, 2011, an MDH will receive the IPPS rate plus 75 percent of the difference between the IPPS rate and its hospital-specific rate, if the hospital-specific rate is higher than the IPPS rate.)

Section 1886(g) of the Act requires the Secretary to pay for the capital-related costs of inpatient hospital services “in accordance with a prospective payment system established by the Secretary.” The basic methodology for determining capital prospective payments is set forth in our regulations at 42 CFR 412.308 and 412.312. Under the capital PPS, payments are adjusted by the same DRG for the case as they are under the operating IPPS. Capital PPS payments are also adjusted for IME and DSH, similar to the adjustments made under the operating IPPS. In addition, hospitals may receive outlier payments for those cases that have unusually high costs.

The existing regulations governing payments to hospitals under the IPPS are located in 42 CFR Part 412, Subparts A through M.

2. Hospitals and Hospital Units Excluded From the IPPS

Under section 1886(d)(1)(B) of the Act, as amended, certain specialty hospitals and hospital units are excluded from the IPPS. These hospitals and units are: inpatient rehabilitation hospitals and units (commonly referred to as inpatient rehabilitation facilities (IRFs); long-term care hospitals (LTCHs); inpatient psychiatric hospitals and units (commonly referred to as inpatient psychiatric facilities (IPFs); children's hospitals; and cancer hospitals. Religious nonmedical health care institutions (RNHCIs) are also excluded from the IPPS. Various sections of the Balanced Budget Act of 1997 (Pub. L. 105-33), the Medicare, Medicaid and SCHIP [State Children's Health Insurance Program] Balanced Budget Refinement Act of 1999 (Pub. L. 106-113), and the Medicare, Medicaid, and SCHIP Benefits Improvement and Protection Act of 2000 (Pub. L. 106-554) provide for the implementation of PPSs for IRFs, LTCHs, and IPFs, as discussed below. Children's hospitals, cancer hospitals, and RNHCIs continue to be paid solely under a reasonable cost-based system.

The existing regulations governing payments to excluded hospitals and hospital units are located in 42 CFR Parts 412 and 413.

a. Inpatient Rehabilitation Facilities (IRFs)

Under section 1886(j) of the Act, IRFs have been transitioned from payment based on a blend of reasonable cost reimbursement and the adjusted IRF Federal prospective payment rate for cost reporting periods beginning on or after January 1, 2002, through September 30, 2002, to payment at 100 percent of the Federal rate effective for cost reporting periods beginning on or after October 1, 2002. IRFs subject to the blend were also permitted to elect payment based on 100 percent of the Federal rate. The existing regulations governing payments under the IRF PPS are located in 42 CFR Part 412, Subpart P.

b. Long-Term Care Hospitals (LTCHs)

Under the authority of sections 123(a) and (c) of Pub. L. 106-113 and section 307(b)(1) of Pub. L. 106-554, LTCHs that do not meet the definition of “new” under § 412.23(e)(4) are being transitioned from being paid for inpatient hospital services based on a blend of reasonable cost-based reimbursement under section 1886(b) of the Act to 100 percent of the Federal rate during a 5-year period with cost reporting periods beginning on or after October 1, 2002. Those LTCHs that do not meet the definition of “new” may elect to be paid based on 100 percent of the Federal prospective payment rate instead of a blended payment in any year during the 5-year transition. For cost reporting periods beginning on or after October 1, 2006, LTCHs will be paid 100 percent of the Federal rate. The existing regulations governing payment under the LTCH PPS are located in 42 CFR Part 412, Subpart O.

c. Inpatient Psychiatric Facilities (IPFs)

Under the authority of sections 124(a) and (c) of Pub. L. 106-113, IPFs are paid under the IPF PPS. Under the IPF PPS, some IPFs are transitioning from being paid for inpatient hospital services based on a blend of reasonable cost-based payment to a Federal per diem payment rate, effective for cost reporting periods beginning on or after January 1, 2005 (November 15, 2004 IPF PPS final rule (69 FR 66922) and May 9, 2006 IPF PPS final rule (71 FR 27040)). For cost reporting periods beginning on or after January 1, 2008, all IPFs will be paid 100 percent of the Federal per diem payment amount. The existing regulations governing payment under the IPF PPS are located in 42 CFR 412, Subpart N.

3. Critical Access Hospitals (CAHs)

Under sections 1814, 1820, and 1834(g) of the Act, payments are made to critical access hospitals (CAHs) (that is, rural hospitals or facilities that meet certain statutory requirements) for inpatient and outpatient services based on 101 percent of reasonable cost. Reasonable cost is determined under the provisions of section 1861(v)(1)(A) of the Act and existing regulations under 42 CFR Parts 413 and 415.

4. Payments for Graduate Medical Education (GME)

Under section 1886(a)(4) of the Act, costs of approved educational activities are excluded from the operating costs of inpatient hospital services. Hospitals with approved graduate medical education (GME) programs are paid for the direct costs of GME in accordance with section 1886(h) of the Act; the amount of payment for direct GME costs for a cost reporting period is based on the hospital's number of residents in that period and the hospital's costs per resident in a base year. The existing regulations governing payments to the various types of hospitals are located in 42 CFR Part 413.

B. Provisions of the Deficit Reduction Act of 2005 (DRA)

On February 8, 2006, the Deficit Reduction Act of 2005 (DRA), Pub. L. 109-171, was enacted. Pub. L. 109-171 made a number of changes to the Act relating to prospective payments to hospitals and other providers for inpatient services. This final rule implements amendments made by the following sections of Pub. L. 109-171:

• Section 5001(a), which, effective for FY 2007 and subsequent years, allows for expansion of the requirements for hospital quality data reporting.
• Section 5003, which makes several changes to the MDH program. It extends special payment provisions, requires MDHs to use FY 2002 as their base year for determining whether use of their hospital-specific rate enhances payment (but permits them to continue to use either their 1982 or 1987 hospital-specific rate if using either of those rates results in higher payments), and removes the application of the 12-percent cap on the DSH payment adjustment factor for MDHs.
• Section 5004, which reduces certain allowable SNF bad debt payments by 30 percent. Payments for the bad debts of full-benefit, dual eligible individuals are not reduced.

In this final rule, we also discuss the provisions of section 5001(b) of Pub. L. 109-171, which require us to develop a plan to implement, beginning with FY 2009, a value-based purchasing plan for section 1886(d) hospitals and summarize the public comments received in response to our invitation for public comments. This discussion also includes the provisions of section 5001(c) of Pub. L. 109-171, which requires a quality adjustment in DRG payments for certain hospital-acquired conditions, effective for FY 2008.

C. Summary of the Provisions of the FY 2007 IPPS and FY 2007 Occupational Mix Adjustment to the Wage Index Proposed Rules

In the FY 2007 IPPS proposed rule, we set forth proposed changes to the Medicare IPPS for operating costs and for capital-related costs in FY 2007. We also set forth proposed changes relating to payments for GME costs, payments to certain hospitals and units that continue to be excluded from the IPPS and paid on a reasonable cost basis, and payments for SCHs and MDHs. The changes were proposed to be effective for discharges occurring on or after October 1, 2006, unless otherwise noted.

After publication of the FY 2007 IPPS proposed rule, the United States Court of Appeals for the Second Circuit issued a decision in the Bellevue case that caused us to modify our proposals on the implementation of the occupational mix adjustment. As a result, we published a second proposed rule in the May 17, 2006 Federal Register that superseded the occupational mix proposals that had been made in the FY 2007 IPPS proposed rule (published April 25, 2006). The following is a summary of the major changes that we proposed to make and the issues that we addressed in the FY 2007 IPPS and FY 2007 Occupational Mix Adjustment to the Wage Index proposed rules:

1. DRG Reclassifications and Recalibrations of Relative Weights

As required by section 1886(d)(4)(C) of the Act, we proposed limited annual revisions to the DRG classifications structure. In this section, we responded to several recommendations made by MedPAC intended to improve the DRG system. We also proposed to use, for FY 2007, hospital-specific relative values (HSRVs) for 10 cost centers to compute DRG relative weights. In addition, we proposed to use consolidated severity-adjusted DRGs or alternative severity adjustment methods in FY 2008 (if not earlier).

We presented our reevaluation of certain FY 2006 applicants for add-on payments for high-cost new medical services and technologies, and our analysis of FY 2007 applicants (including public input, as directed by Pub. L. 108-173, obtained in a town hall meeting).

We proposed the annual update of the long-term care diagnosis-related group (LTC-DRG) classifications and relative weights for use under the LTCH PPS for FY 2007.

2. Changes to the Hospital Wage Index

We proposed revisions to the wage index and the annual update of the wage data. Specific issues addressed include the following:

• The FY 2007 wage index update, using wage data from cost reporting periods that began during FY 2003.
• The FY 2007 occupational mix adjustment to the wage index (discussed inthe May 17, 2006 proposed rule).
• The revisions to the wage index based on hospital redesignations and reclassifications.
• The adjustment to the wage index for FY 2007 based on commuting patterns of hospital employees who reside in a county and work in a different area with a higher wage index.
• The timetable for reviewing and verifying the wage data that will be in effect for the proposed FY 2007 wage index.
• The special timetable that will apply in FY 2007 in order to allow us to make presumptive reclassification withdrawal or termination decisions on behalf of affected hospitals which will then become final unless reversed or modified by the affected hospitals in accordance with CMS procedural rules.
• The labor-related share for the FY 2007 wage index, including the labor-related share for Puerto Rico.

3. Other Decisions and Changes to the IPPS for Operating Costs, GME Costs, and Promoting Hospitals' Effective Use of Health Information Technology

In the proposed rule, we discussed a number of provisions of the regulations in 42 CFR Parts 412 and 413 and related proposed changes, including the following:

• The reporting of hospital quality data as a condition for receiving the full annual payment update increase.
• Changes in payments to SCHs and MDHs.
• Updated national and regional case-mix values and discharges for purposes of determining rural referral center status.
• The statutorily-required IME adjustment factor for FY 2007.
• Changes relating to hospitals' geographic classifications, including reclassifications under section 508 of Pub. L. 108-173, multicampus hospitals, urban group hospital reclassification and the effect of change in ownership on urban county group reclassifications.
• Changes and clarifications relating to GME that address determining the per resident amounts (PRAs) for merged hospitals and new teaching hospitals, counting and appropriate documentation of FTE residents, and counting of resident time spent in nonpatient care activities as part of approved residency programs.
• Changes relating to payment for costs of nursing and allied health education programs.
• Changes relating to requirements for emergency services for hospitals under EMTALA.
• Discussion of the third year of implementation of the Rural Community Hospital Demonstration Program.

We also invited comments on promoting hospitals' effective use of health information technology.

4. Changes to the PPS for Capital-Related Costs

In the proposed rule, we discussed the payment policy requirements for capital-related costs and capital payments to hospitals and proposed several technical corrections to the regulations.

5. Changes for Hospitals and Hospital Units Excluded From the IPPS

In the proposed rule, we discussed payments made to excluded hospitals and hospital units, proposed policy changes regarding decreases in square footage or decreases in the number of beds of the “grandfathering” HwHs and satellite facilities, and proposed changes to the methodology for determining LTCH CCRs and the reconciliation of high-cost and short-stay outlier payments under the LTCH PPS. In addition, we proposed a technical change relating to the designation of CAHs as necessary providers.

6. Payments for Services Furnished Outside the United States

In the proposed rule, we set forth proposed changes to clarify what is considered “outside the United States” for Medicare payment purposes.

7. Payment for Blood Clotting Factor Administered to Inpatients With Hemophilia

In the proposed rule, we discussed the proposed changes in payment for blood clotting factor administered to Medicare beneficiaries with hemophilia for FY 2007.

8. Limitation on Payments to Skilled Nursing Facilities for Bad Debt

In the proposed rule, we proposed to implement section 5004 of Pub. L. 109-171 relating to reduction in payments to SNFs for bad debt.

9. Determining Prospective Payment Operating and Capital Rates and Rate-of-Increase Limits

In the Addendum to the proposed rule, we set forth proposed changes to the amounts and factors for determining the FY 2007 prospective payment rates for operating costs and capital-related costs. We also proposed to establish the threshold amounts for outlier cases. In addition, we addressed the proposed update factors for determining the rate-of-increase limits for cost reporting periods beginning in FY 2007 for hospitals and hospital units excluded from the PPS.

10. Impact Analysis

In Appendix A of the proposed rule, we set forth an analysis of the impact that the proposed changes would have on affected hospitals.

11. Recommendation of Update Factors for Operating Cost Rates of Payment for Inpatient Hospital Services

In Appendix B of the proposed rule, as required by sections 1886(e)(4) and (e)(5) of the Act, we provided our recommendations of the appropriate percentage changes for FY 2007 for the following:

• A single average standardized amount for all areas for hospital inpatient services paid under the IPPS for operating costs (and hospital-specific rates applicable to SCHs and MDHs).
• Target rate-of-increase limits to the allowable operating costs of hospital inpatient services furnished by hospitals and hospital units excluded from the IPPS.

12. Discussion of Medicare Payment Advisory Commission Recommendations

Under section 1805(b) of the Act, MedPAC is required to submit a report to the Congress, no later than March 1 of each year, in which MedPAC reviews and makes recommendations on Medicare payment policies. MedPAC's March 2006 recommendation concerning hospital inpatient payment policies addressed the update factor for inpatient hospital operating costs and capital-related costs under the IPPS and for hospitals and distinct part hospital units excluded from the IPPS. This recommendation was addressed in Appendix B of the proposed rule. For further information relating specifically to the MedPAC reports or to obtain a copy of the reports, contact MedPAC at (202) 220-3700 or visit MedPAC's Web site at: www.medpac.gov.

13. Appendix C and Appendix D

In Appendix C of the proposed rule, we listed the combinations of the consolidated severity-adjusted DRGs that we proposed to implement on FY 2008 (if not earlier), as discussed in section II.C. of the preamble of the proposed rule. In Appendix D of the proposed rule, we provided a crosswalk of the proposed consolidated severity-adjusted DRG system to the respective All Patient Related Diagnosis-Related Group (APR DRG) system.

D. Public Comments Received in Response to the FY 2007 IPPS and FY 2007 Occupational Mix Adjustment to the Wage Index Proposed Rules

We received over 2,300 timely items of correspondence containing multiple comments on the FY 2007 IPPS proposed rule. We also received over 100 timely items of correspondence on the FY 2007 Occupational Mix Adjustment to the Wage Index proposed rule. Summaries of the public comments and our responses to those comments are set forth under the appropriate heading.

E. Interim Final Rule on Selection Criteria of Loan Program for Qualifying Hospitals Engaged in Cancer-Related Health Care

On September 30, 2005, we published in the Federal Register (70 FR 57368) an interim final rule with comment period (CMS-1287-IFC) that set forth the criteria for implementing a loan program for qualifying hospitals engaged in research in the causes, prevention, and treatment of cancer, as specified in section 1016 of the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (Pub. L. 108-173). Specifically, this interim final rule established a loan application process by which qualifying hospitals, including specified entities, may apply for a loan for the capital costs of health care infrastructure improvement projects. The interim final rule was effective on November 29, 2005.

We received seven timely items of correspondence on the interim final rule. In section XI. of the preamble to this final rule, we are finalizing this interim final rule with comment period. In that section, we discuss the provisions of the program, the public comments received, our responses to those comments, and the final policy.

F. Proposed Rule on Forgiveness of Indebtedness under the Health Care Infrastructure Improvement Program

On September 30, 2005, we published in the Federal Register (70 FR 57376) a proposed rule (CMS-1320-P) to establish the loan forgiveness criteria for qualifying hospitals who receive loans under the Health Care Infrastructure Improvement Program that was established under section 1016 of Pub. L. 108-173.

We received one timely item of correspondence on this proposed rule. We address the provisions of the proposed rule, a summary of the public comments received and our responses, and the provisions of the final rule in section XI. of the preamble of this final rule.

G. Interim Final Rule on the Exclusion of Vendor Purchases Made Under the Competitive Acquisition Program for Part B Outpatient Drugs and Biologicals for the Purpose of Calculating the Average Sales Price

In November 21, 2005 Federal Register (70 FR 70748), we published an interim final rule with comment period (CMS-1325-IFC3) to clarify and solicit comments on the relationship between drugs supplied under the CAP for Part B Drugs and Biologicals and the calculation of the ASP.

We did not receive any timely items of correspondence on this interim final rule with comment period. We summarize the provisions of the July 6, 2005 and the November 21, 2005 interim final rules and the current interim final provisions in section XII. of the preamble of this final rule.

II. Changes to DRG Classifications and Relative Weights

A. Background

Section 1886(d) of the Act specifies that the Secretary shall establish a classification system (referred to as DRGs) for inpatient discharges and adjust payments under the IPPS based on appropriate weighting factors assigned to each DRG. Therefore, under the IPPS, we pay for inpatient hospital services on a rate per discharge basis that varies according to the DRG to which a beneficiary's stay is assigned. The formula used to calculate payment for a specific case multiplies an individual hospital's payment rate per case by the weight of the DRG to which the case is assigned. Each DRG weight represents the average resources required to care for cases in that particular DRG, relative to the average resources used to treat cases in all DRGs.

Congress recognized that it would be necessary to recalculate the DRG relative weights periodically to account for changes in resource consumption. Accordingly, section 1886(d)(4)(C) of the Act requires that the Secretary adjust the DRG classifications and relative weights at least annually. These adjustments are made to reflect changes in treatment patterns, technology, and any other factors that may change the relative use of hospital resources.

B. DRG Reclassifications

1. General

As discussed in section II.D. of the preamble to the FY 2007 IPPS proposed rule (71 FR 24030), for FY 2007, we are making only limited changes to the current DRG classifications that will be applicable to discharges occurring on or after October 1, 2006. We are limiting our changes because, as discussed in detail in section II.C. of the preamble to the proposed rule and to this final rule, we are focusing our efforts on addressing the recommendations made last year by MedPAC to refine the entire CMS DRG system by taking into account severity of illness and applying hospital-specific relative value (HSRV) weights to DRGs.

Currently, cases are classified into CMS DRGs for payment under the IPPS based on the principal diagnosis, up to eight additional diagnoses, and up to six procedures performed during the stay. In a small number of DRGs, classification is also based on the age, sex, and discharge status of the patient. The diagnosis and procedure information is reported by the hospital using codes from the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM).

The process of forming the DRGs was begun by dividing all possible principal diagnoses into mutually exclusive principal diagnosis areas, referred to as Major Diagnostic Categories (MDCs). The MDCs were formed by physician panels as the first step toward ensuring that the DRGs would be clinically coherent. The diagnoses in each MDC correspond to a single organ system or etiology and, in general, are associated with a particular medical specialty. Thus, in order to maintain the requirement of clinical coherence, no final DRG could contain patients in different MDCs. Most MDCs are based on a particular organ system of the body. For example, MDC 6 is Diseases and Disorders of the Digestive System. This approach is used because clinical care is generally organized in accordance with the organ system affected. However, some MDCs are not constructed on this basis because they involve multiple organ systems (for example, MDC 22 (Burns)). For FY 2006, cases are assigned to one of 526 DRGs in 25 MDCs. The table below lists the 25 MDCs.

In general, cases are assigned to an MDC based on the patient's principal diagnosis before assignment to a DRG. However, for FY 2006, there are nine DRGs to which cases are directly assigned on the basis of ICD-9-CM procedure codes. These DRGs are for heart transplant or implant of heart assist systems, liver and/or intestinal transplants, bone marrow transplants, lung transplants, simultaneous pancreas/kidney transplants, pancreas transplants, and for tracheostomies. Cases are assigned to these DRGs before they are classified to an MDC. The table below lists the nine current pre-MDCs.

Once the MDCs were defined, each MDC was evaluated to identify those additional patient characteristics that would have a consistent effect on the consumption of hospital resources. Because the presence of a surgical procedure that required the use of the operating room would have a significant effect on the type of hospital resources used by a patient, most MDCs were initially divided into surgical DRGs and medical DRGs. Surgical DRGs are based on a hierarchy that orders operating room (O.R.) procedures or groups of O.R. procedures by resource intensity. Medical DRGs generally are differentiated on the basis of diagnosis and age (0 to 17 years of age or greater than 17 years of age). Some surgical and medical DRGs are further differentiated based on the presence or absence of a complication or a comorbidity (CC).

Generally, nonsurgical procedures and minor surgical procedures that are not usually performed in an operating room are not treated as O.R. procedures. However, there are a few non-O.R. procedures that do affect DRG assignment for certain principal diagnoses. An example is extracorporeal shock wave lithotripsy for patients with a principal diagnosis of urinary stones.

Once the medical and surgical classes for an MDC were formed, each class of diagnoses was evaluated to determine if complications, comorbidities, or the patient's age would consistently affect the consumption of hospital resources. Physician panels classified each diagnosis code based on whether the diagnosis, when present as a secondary condition, would be considered a substantial CC. A substantial CC was defined as a condition which, because of its presence with a specific principal diagnosis, would cause an increase in the length of stay by at least one day in at least 75 percent of the patients. Each medical and surgical class within an MDC was tested to determine if the presence of any substantial CC would consistently affect the consumption of hospital resources.

A patient's diagnosis, procedure, discharge status, and demographic information is fed into the Medicare claims processing systems and subjected to a series of automated screens called the Medicare Code Editor (MCE). The MCE screens are designed to identify cases that require further review before classification into a DRG.

After patient information is screened through the MCE and any further development of the claim is conducted, the cases are classified into the appropriate DRG by the Medicare GROUPER software program. The GROUPER program was developed as a means of classifying each case into a DRG on the basis of the diagnosis and procedure codes and, for a limited number of DRGs, demographic information (that is, sex, age, and discharge status).

After cases are screened through the MCE and assigned to a DRG by the GROUPER, the PRICER software calculates a base DRG payment. The PRICER calculates the payment for each case covered by the IPPS based on the DRG relative weight and additional factors associated with each hospital, such as IME and DSH adjustments. These additional factors increase the payment amount to hospitals above the base DRG payment.

The records for all Medicare hospital inpatient discharges are maintained in the Medicare Provider Analysis and Review (MedPAR) file. The data in this file are used to evaluate possible DRG classification changes and to recalibrate the DRG weights. However, in the July 30, 1999 IPPS final rule (64 FR 41500), we discussed a process for considering non-MedPAR data in the recalibration process. In order for us to consider using particular non-MedPAR data, we must have sufficient time to evaluate and test the data. The time necessary to do so depends upon the nature and quality of the non-MedPAR data submitted. Generally, however, a significant sample of the non-MedPAR data should be submitted by mid-October for consideration in conjunction with the next year's proposed rule. This allows us time to test the data and make a preliminary assessment as to the feasibility of using the data. Subsequently, a complete database should be submitted by early December for consideration in conjunction with the next year's proposed rule.

In the FY 2007 IPPS proposed rule, we proposed limited changes to the DRG classification system for FY 2007 for the FY 2007 GROUPER, Version 24.0 and to the methodology used to recalibrate the DRG weights. The changes we proposed, the public comments we received concerning the proposed changes, the final DRG changes, and the methodology used to calculate the DRG weights are set forth below. The changes we are implementing in this final rule will be reflected in the FY 2007 GROUPER, Version 24.0, and are effective for discharges occurring on or after October 1, 2006. Unless otherwise noted in this final rule, our DRG analysis is based on data from the March 2006 update of the FY 2005 MedPAR file, which contains hospital bills received through March 31, 2006, for discharges occurring in FY 2005.

2. Yearly Review for Making DRG Changes

Many of the changes to the DRG classifications are the result of specific issues brought to our attention by interested parties. We encourage individuals with concerns about DRG classifications to bring those concerns to our attention in a timely manner so they can be carefully considered for possible inclusion in the annual proposed rule Therefore, similar to the timetable for interested parties to submit non-MedPAR data for consideration in the DRG recalibration process, concerns about DRG classification issues should be brought to our attention no later than early December in order to be considered and possibly included in the next annual proposed rule updating the IPPS.

The actual process of forming the DRGs was, and continues to be, highly iterative, involving a combination of statistical results from test data combined with clinical judgment. For purposes of this final rule, in deciding whether to create a separate DRG, we consider whether the resource consumption and clinical characteristics of the patients with a given set of conditions are significantly different than the remaining patients in the existing DRG. We evaluate patient care costs using average charges and lengths of stay as proxies for costs and rely on the judgment of our medical officers to decide whether patients are clinically distinct or similar to other patients in the DRG. In evaluating resource costs, we consider both the absolute and percentage differences in average charges between the cases we are selecting for review and the remainder of cases in the DRG. We also consider variation in charges within these groups; that is, whether observed average differences are consistent across patients or attributable to cases that are extreme in terms of charges or length of stay, or both. Further, we also consider the number of patients who will have a given set of characteristics and generally prefer not to create a new DRG unless it will include a substantial number of cases.

C. Revisions to the DRG System Used Under the IPPS

1. MedPAC Recommendations

In the FY 2006 IPPS final rule, we discussed a number of recommendations made by MedPAC regarding revisions to the DRG system used under the IPPS (70 FR 47473 through 47482).

In Recommendation 1-3 in the 2005 Report to Congress on Physician-Owned Specialty Hospitals, MedPAC recommended that CMS refine the current DRGs to more fully capture differences in severity of illness among patients, including:

• Base the DRG relative weights on the estimated cost of providing care.
• Base the weights on the national average of the hospital-specific relative values (HSRVs) for each DRG (using hospital-specific costs to derive the HSRVs).
• Adjust the DRG relative weights to account for differences in the prevalence of high-cost outlier cases.
• Implement the case-mix measurement and outlier policies over a transitional period.

As we noted in the FY 2006 IPPS final rule, we had insufficient time to complete a thorough evaluation of these recommendations for full implementation in FY 2006. However, we did adopt severity-weighted cardiac DRGs in FY 2006 to address public comments on this issue and the specific concerns of MedPAC regarding cardiac surgery DRGs. We also indicated that we planned to further consider all of MedPAC's recommendations and thoroughly analyze options and their impacts on the various types of hospitals in the FY 2007 IPPS proposed rule. Following the publication of the FY 2006 IPPS final rule, we contracted with 3M Health Information Systems to assist us in performing this analysis.

Beginning with MedPAC's relative weight recommendations, we analyzed MedPAC's recommendations to move to a cost-based HSRV weighting methodology. In performing this portion of the analysis, we studied hospital cost report data, departmental cost-to-charge ratios (CCRs), MedPAR claims data, and HSRV weighting methodology. Our intention in undertaking this portion of the analysis was to find an administratively feasible approach to improving the accuracy of the DRG weights. As we described in the proposed rule, we believe some changes can be made to MedPAC's methodology for determining the relative weights that will make it more feasible to replicate on an annual basis but will result in similar impacts.

In conjunction with analyzing MedPAC's relative weight recommendations, we looked at refining the current DRG system to better recognize severity of illness. Starting with the APR DRG GROUPER used by MedPAC in its analysis, we studied Medicare claims data. Based on this analysis, we developed a CS DRG GROUPER that we believe could be a better alternative for recognizing severity of illness among the Medicare population. We note that MedPAC's recommendations with regard to revising the DRGs to better recognize severity of illness may have implications for the outlier threshold, the measurement of real case-mix versus apparent case-mix, and the IME and the DSH adjustments. We discuss these implications in more detail in the following sections.

As we present below, we believe that the recommendations made by MedPAC, or some variants of them, have significant promise to improve the accuracy of the payment rates in the IPPS. We agree with MedPAC about exploring possible refinements to our payment methodology even in the absence of concerns about the proliferation of specialty hospitals. In the FY 2006 final rule, we indicated that until we had completed further analysis of the options and their effects, we could not predict the extent to which changing to APR DRGs would provide payment equity between specialty and general hospitals. In fact, we cautioned that any system that groups cases will always present some opportunities for providers to specialize in cases they believe to have higher margins. We believe that improving payment accuracy should reduce these opportunities and potentially reduce the incentives that Medicare payments may provide for the further development of specialty hospitals.

We considered MedPAC's recommendation to adjust the relative weights to account for differences in the prevalence of outlier cases. However, we placed most of our attention and resources on the recommendations related to refinement of the current DRGs to more fully capture differences in severity of illness among patients, as we do not have the statutory authority to make the specific changes to our outlier policy that MedPAC recommended. While we have not made MedPAC's recommendation regarding outliers a central focus of our analysis, we do intend to examine this issue in more detail in the future. In sections II.C.2. through C.6. of the FY 2007 proposed rule, we discussed a number of issues related to the MedPAC recommendations. We also presented our analysis and specific proposals for FY 2007 and FY 2008 including their estimated impacts. In this final rule, we present the public comments received on the proposed rule, our responses to those comments, our final decisions for FY 2007 and our intended actions for FY 2008.

2. Refinement of the Relative Weight Calculation

MedPAC made two recommendations with respect to the DRG relative weight calculation. First, MedPAC recommended that CMS base the DRG relative weights on the estimated cost of providing care. Second, MedPAC recommended that CMS base the weights on the national average of hospitals' relative values in each DRG. Because both of these recommendations address the relative weight calculation, we are addressing them together. The work we have done to address these recommendations was discussed in detail in the proposed rule (71 FR 24006-24011).

MedPAC recommended that CMS replace its charge-based relative weight methodology with cost-based weights, as it believed that the charge-based relative weight methodology that CMS has utilized since 1985 has introduced bias into the weights due to differential markups for ancillary services among the DRGs. In analyzing claims data, it is evident to us that some hospital types (for example, teaching hospitals) are systematically more expensive overall than the average hospital and certain case types are more commonly treated at these more expensive facilities. Higher average charges for cases that are treated at more expensive hospitals may result in higher weights for these types of cases. MedPAC suggested a hospital-specific relative value (HSRV) methodology which MedPAC believed would reduce the effect of cost differences among hospitals that may be present in the national relative weights due to differences in case-mix adjusted costs.

Under the HSRV methodology recommended by MedPAC, charges are standardized for each provider by converting its charges for each case to hospital-specific relative charge values and then adjusting those values for the hospital's case-mix. The first step in this process involves dividing the charge for each case at the hospital by the average charge for all cases at the hospital in which the case was treated. The hospital-specific relative charge value, by definition, averages 1.0 for each hospital. The resulting ratio is then multiplied by the hospital's case-mix index (CMI). In this way, each hospital's relative charge value is adjusted by its case-mix to an average that reflects the complexity of the cases it treats relative to the complexity of the cases treated by all other hospitals. We discuss this issue in further detail below.

Our analysis of departmental-level CCRs from the Medicare cost report data has shown that charges for routine days, intensive care days, and various ancillary services are not marked up by a consistent amount. For example, the markup amounts for cardiology services are higher than average. Because charges are the current basis for the DRG relative weights, the practice of differential markups can lead to bias in the DRG weights because various DRGs use, on average, more or less of particular ancillary services. MedPAC believes that the bias in the national DRG relative weights that may arise as a result of differential markups across various cost centers can be removed by moving from charge-based to cost-based weights. Based on the analysis we have conducted, we agree that it is appropriate to adjust the DRG relative weights to account for the differences in charge markups across cost centers.

In the proposed rule, we indicated several concerns about the methodology used by MedPAC. MedPAC's methodology to reduce hospital charges to cost is administratively burdensome, not only to develop, but also to maintain. First, MedPAC developed CCRs for individual hospitals at the most detailed department level. Specifically, in calculating costs as the basis for the relative weights, MedPAC applied hospital-specific CCRs from each provider's cost report to the line item charges on the claims that the hospital submitted during the same time period. This methodology required matching cost report data to claims data, and because cost report data take longer to compile and file, the method necessitates using older claims data to set relative weights. The most recent complete set of Medicare cost reports available to us is from FY 2003. Thus, if we were to model the exact approach used by MedPAC and use claims data for a matching year, we would be using claims data from FY 2003 instead of using FY 2005 claims data, as we would if we were to continue with our current methodology. In addition, MedPAC's hospital-specific approach required detailed cost center distinctions for each hospital that are difficult to define, map, and apply. This approach also required the use of the Standard Analytic File (SAF) because MedPAR data that we currently use to set DRG weights did not have the necessary level of detail. Using the SAF increases processing time and adds further complexity to the process of setting the relative weights.

Second, because MedPAC applied these CCRs at the individual claim level, missing or invalid data resulted in MedPAC deleting a large number of claims (approximately 10 percent) from the relative weight calculation. Lastly, MedPAC acknowledged that its method was too difficult to replicate on an annual basis and suggested that the weights be recalculated once every 5 years with other adjustments based on charges during the intervening years.

As we explained in the FY 2007 IPPS proposed rule, we developed an alternative to MedPAC's approach that we believe achieves similar results in a more administratively feasible manner. This method involves developing hospital-specific charge relative weights at the cost center level and then scaling the weights to costs using the national cost center charge ratios developed from the cost report data. After studying Medicare cost report data, we established 10 cost center categories based upon broad hospital accounting definitions. In our cost center categories, there are 8 ancillary cost groups in addition to routine day costs and intensive care day costs, and each category represents at least 5 percent of the charges in the claims data. The specific cost report lines that contribute to each category and the corresponding charge lines from the MedPAR claims data are itemized in Table A below.

In the proposed rule, we stated that this alternative approach, which we labeled as the HSRV cost center (HSRVcc) methodology, has several advantages. First, the use of national average rather than hospital-specific CCRs avoids the complexity encountered with cost center CCRs at the hospital level and allows us to retain more data for use in the relative weight calculation. In addition, the methodology eliminates the need to match claims to the time period of the CCRs, resulting in the ability to use more timely claims data. Furthermore, the alternative approach makes it more feasible to update the relative weights annually using a single methodology. We do not have to replicate the methodology once every 5 years and make adjustments based on changes in charges in the intervening years. The HSRVcc methodology is described in detail in the proposed rule (71 FR 24008 through 24011).

Comment: Several commenters supported CMS' effort to restructure the DRG relative weights based on cost. They stated that using charges as a proxy for hospital costs in determining resource utilization under the current system is inappropriate and encouraged CMS to implement a cost-based system consistent with the agency's original intent without delay.

Response: We appreciate the commenters' support of our proposal to implement a cost-based weighting methodology. We believe that adopting cost-based weights will result in significant improvements to Medicare's IPPS payments. MedPAC concluded after an extensive analysis of Medicare hospital inpatient claims and cost data that the IPPS payment rates are badly distorted, resulting in Medicare paying too much for some types of patients and too little for others. As indicated below, we are making some modifications to our proposals in response to the public comments. However, we are adopting a system of cost-based weights for FY 2007 to address the concerns raised by MedPAC. As a result, all hospitals, including specialty hospitals, will be paid more appropriately. In addition, based on our analysis, we concur with MedPAC that the current DRG system needs to be changed to better account for severity of illness among patients. This issue is discussed in more detail in the next section of this final rule.

Comment: A majority of commenters supported CMS' efforts to improve the accuracy of the DRG weights, and better reflect variations in patients' severity of illness. However, many commenters viewed the HSRVcc proposal as flawed from both a methodological and policy perspective, and believed the proposal to implement cost-based weights should be delayed for at least a year. They believed that CMS needs to further consider a number of issues raised in the public comments before such sweeping changes are implemented. In addition, the commenters indicated that CMS needs to provide hospitals with more lead-time before implementing changes so they can budget accurately. They urged CMS to use the current standardized charge-based approach in FY 2007 until these issues can be addressed. At a minimum, they believed CMS should address what were characterized as methodological flaws and publish revised relative weights along with hospital impacts for public comment prior to implementation.

Response: We appreciate the commenters' concerns with regard to a rapid and full implementation of the changes we proposed to the relative weight methodology. However, based on our analysis and study of the MedPAC recommendations that we presented in our proposed rule, it has come to our attention that differential markups between routine and ancillary cost centers have introduced significant bias into the relative weights. In order to reduce the bias in weights and make more appropriate payments under the IPPS, we believe it is necessary to initiate the transition to a cost-based relative weight methodology in FY 2007. However, we have considered the commenters' requests to further review the HSRV methodology. Therefore, in this final rule, we are not adopting our proposal to standardize charges using the HSRV methodology. However, we are adopting our proposal to reduce charges to estimated costs prior to setting DRG weights. We will undertake further analysis of the HSRV methodology during the next year. Based on this analysis, we will consider proposing further changes to adopt the HSRV methodology for FY 2008.

Comment: Many commenters disagreed with CMS' assertion that the more administratively feasible HSRVcc approach achieves similar results to the MedPAC methodology. While they supported CMS' efforts to ensure the DRG weights are updated annually to reflect the most recent trends in inpatient care, they expressed concern with the specifics of the HSRVcc methodology.

First, they noted that CMS stated in the proposed rule that organ acquisition costs were eliminated from hospital charges before the HSRVcc weights were calculated. However, it had come to their attention that organ acquisition charges were actually included in the calculation of DRG weights under the proposed methodology. They stated that organ acquisition is reimbursed by Medicare on a cost basis and should not be included in the weight calculation. Furthermore, the commenters asserted that the inclusion of organ acquisition charges improperly overstated the transplant DRG HSRVcc weights. Commenters recommended that CMS remove the organ acquisition charges from the computation of the DRG weights if the HSRVcc methodology is to be adopted.

Second, commenters believe CMS made questionable methodological decisions when calculating the national CCRs. Under the proposed methodology, CMS calculated hospital-weighted rather than charge-weighted CCRs for each of the 10 cost centers used to scale the charge-based weights. Because the averages are unweighted, the commenters stated that the CCRs do not account for the differential contribution of each hospital to total charges. The commenters asserted that, mathematically, the only correct way to get from total hospital charges to total hospital costs is to use a charge-weighted average of hospital CCRs. Failure to use charge-weighted averages overestimates routine and ICU costs and underestimates ancillary costs, which ultimately exaggerates the shift in payments, according to the commenters. Therefore, commenters believed CMS should recalculate the mean national CCRs using a charge-weighted method.

Third, commenters believed CMS applied questionable trimming criteria in computing the cost center CCRs. They stated that trimming the cost center CCRs at 1.96 standard deviations (rather than 3 standard deviations) from the geometric mean inappropriately excluded over 200 large hospitals that account for 25 percent of routine accommodation charges. They noted that the CCRs for these hospitals appear to be predominantly correct. In addition, the commenters noted that CMS applied the CCRs to the charge data for hospitals that were excluded from the national average CCR calculation. Thus, the commenters argued there is a significant mismatch between the hospital data that was included in the CCR and HSRVcc calculations. These commenters recommended that CMS exclude hospital data from the CCRs if it is more than 3 standard deviations (rather than 1.96) from the mean CCR. Many commenters characterized these methodological decisions as errors and indicated that their combined impact is significant. If CMS is to use the HSRVcc methodology, the commenters indicated that these issues should be addressed.

A few commenters stated that we made incorrect assumptions that may have resulted in new distortions to the relative weights. Specifically, the commenters stated that we were incorrect in applying the same CCR across all hospitals for a given cost center and applying the same percent mix of services by cost center to all DRGs. The commenters recommended that we first convert charges to costs for each hospital and DRG, and then compute hospital-specific relative values. They stated that the reversal of the calculations in the HSRVcc methodology accommodates cost center mix and charge markup differences across hospitals and across DRGs.

Many commenters argued that the hospital-specific relative value methodology is unnecessary and compresses the DRG weights. Commenters cited past research indicating that HSRV has a disproportionate impact on certain types of hospitals and types of care, and reduces the range of DRG weights between the lowest and highest weight DRGs.¹ Commenters noted that the HSRV methodology “produces more compressed DRG weights” than the existing standardization methodology and that “the greater compression of the HSRV weights is counter balanced by the fact that more high-weighted cases qualify as [high cost] outlier cases.” A few commenters expressed concern that adopting MedPAC's recommendation to exclude high-cost outliers in addition to statistical outliers from the computation of the DRG weights so that the weights reflect the average cost only of inlier cases would compound the DRG weight compression caused by the HSRV methodology because high-cost outlier cases occur most frequently in high-weighted DRGs. The commenters indicated that the finding raises the concern of patient access to care for services in higher cost DRGs.

Commenters also believed that the HSRV methodology fails to take into account legitimate variation in costs that occur between hospitals. Therefore, any hospital-level variation in cost that is not explained by the IPPS case mix index is simply ignored, according to the commenters. To the extent that certain services are provided most frequently in hospitals with higher than average cost, the commenters believed that the HSRV methodology will result in inappropriately lower DRG weights for these services.

Therefore, commenters strongly recommended that the HSRV methodology be eliminated in favor of the cost-based weighting methodology adopted under the OPPS. They stated that the main difference between these two approaches is the treatment of cost variation that is not otherwise explained with IPPS payment factors. In the standardization approach employed by OPPS, any variation in hospital costs that is not explained by CMS payment factors affects the calibration of DRG weights. They stated that the HSRV approach proposed by CMS, by contrast, ignores any hospital level variation in charges that is not explained by the case mix index. Many commenters added that CMS could propose to remove other sources of cost variation beyond its current practice of standardizing for wage index, DSH, and IME. They believed a factor-specific approach to standardization would lead to more precise and valid adjustments than those recognized under the HSRV methodology, which eliminates all sources of charge variation irrespective of whether there are legitimate differences among hospitals in costs that are not taken into account in the payment system.

Response: In preparing the FY 2007 relative weights, the costs of organ acquisition were inadvertently included in the relative weight for the calculation of “other services.” The costs of organ acquisition are paid by Medicare on a cost basis and should not be included in setting the IPPS relative weights. These costs have been excluded from the IPPS relative weights calculated for this final rule.

In response to the concerns expressed regarding the CCR calculation, we proposed to establish the geometric mean CCRs using a hospital-weighted methodology because we believed that it served as an acceptable measure of central tendency. In addition, we proposed to trim the CCRs on the basis of 1.96 standard deviations since we were using national averages and thought a more stringent statistical trim would be appropriate. In response to comments, however, we have reconsidered our approach and have implemented the 3 standard deviation statistical trim supported by commenters. Further, we are also adopting the charge-weighted method of calculating CCRs, as we now believe it may be more appropriate to apply CCRs based on aggregate costs and charges among hospitals to the charges that are aggregated by DRG and used to set the relative weights.

Although commenters asserted that the HSRV methodology exacerbates the effect of charge compression on the relative weights, we have not had sufficient time between the close of the comment period and the publication of this final rule to analyze this assertion. Therefore, in response to comments (and as stated above), we are postponing the implementation of the HSRV methodology until we can study this comment further. Instead, as suggested by many commenters, we are using an approach to calculating the IPPS relative weights that is more similar to the approach used in the OPPS. That is, rather than using a hospital-specific relative weighting methodology, we are standardizing charges to remove relevant payment factor adjustments and then adjusting those charges to costs using national cost center CCRs. As we stated in the proposed rule, it is not administratively feasible to adjust charges to cost using hospital-specific cost to charge ratios. Therefore, while we are standardizing charges for the IPPS cost-based weights using a similar process to the OPPS, we are still utilizing national average CCRs to determine cost. Specifically, we are standardizing the charges for each DRG by cost center to remove differences in wage index, indirect medical education and disproportionate share adjustments and are then reducing the standardized charges to cost using the national average CCRs. The relative weights we are adopting in this final rule are calculated based on the average total cost for a DRG in relation to the national average total cost.

Comment: Many commenters expressed concern that CMS collapsed the full set of at least 37 cost centers into only 10. They believed this approach eliminates detail that is available on the cost report. The commenters requested that CMS elaborate on the process it went through to derive the 10 cost centers used to calculate the HSRVcc weights. Some commenters stated CMS should use all 37 cost centers that are used in calculating the OPPS relative weights for the IPPS. Other commenters suggested that CMS expand the number of cost centers used in the calculation. MedPAC found that the CCRs within the proposed 10 cost centers varied significantly in some areas and recommended that CMS expand the number to 13 by distinguishing anesthesia and labor and delivery from the operating room cost center and distinguishing inhalation therapy from the therapy services cost center. Several commenters supported MedPAC's recommendation. Further, MedPAC recommended that the CCRs be based on Medicare-specific costs and charges rather than on the costs and charges for the entire facility. Some commenters advocated that a separate cost center be added for implantable devices. They believed this additional cost center would better identify the mark-up for high cost technological devices than using the average for all supplies and equipment.

Several commenters encouraged CMS to specifically incorporate nursing costs into the weighting methodology. They stated that nursing care represents approximately 30 percent of all hospital expenditures and nearly half of all direct care costs and have been essentially ignored in the payment formula. Specifically, these commenters urged CMS to create a unique Nursing Cost Center that identifies the inpatient direct and indirect costs for registered nurses, licensed practical nurses, and unlicensed assistive personnel. They defined direct nursing costs as those associated with licensed and assistive nursing personnel assigned to care for an individual patient. Indirect nursing costs are all other salary and benefits related to licensed and assistive nursing personnel not directly assigned to care for individual patients. They suggested that the routine and intensive care cost centers in the proposed HSRVcc methodology be replaced with a nursing cost center and a separate facility cost center to identify the non-nursing cost component of care. They urged CMS to set aside funds to study and implement the above recommendation using methodologically sound research and demonstration projects.

Response: As we stated in the proposed rule, we established 10 cost center categories based upon broad hospital accounting definitions. These 10 cost center categories consist of 8 ancillary cost groups, a routine days cost group, and an intensive care days cost group. These cost centers were selected because each category represents at least 5 percent of the charges in the claims data.

We thoroughly reviewed the comments advocating that we expand the number of cost centers used in the calculation. We currently use the MedPAR data set for charge detail. The MedPAR file does not provide enough granularity in the charge detail to support 37 different cost centers. In addition, in the proposed methodology, we eliminated claims for providers that did not have costs greater than zero for at least 8 of the 10 cost centers. At least 96 percent of the providers in the MedPAR file had charges for at least 8 of the 10 cost centers. We believe that if we were to expand to the full set of 37 cost centers outlined in the cost report, we would eliminate a greater number of claims in the calculation of the DRG relative weights.

While we do not believe expanding to 37 cost centers is feasible, we agree with MedPAC that we may have consolidated a few revenue centers that have significantly different CCRs. Upon further examination of the data, in this final rule, we are expanding the number of cost centers from 10 to 13 by creating separate cost centers for anesthesia, labor and delivery, and inhalation therapy. We also agree with MedPAC that it would be more appropriate to set the CCRs based on Medicare-specific charges and costs rather than on the costs and charges for the entire facility. Therefore, in this final rule, we are modifying our CCR calculations to incorporate Medicare-specific charge data from Worksheet D Part 4 in addition to the cost and charge data from Worksheet C Part I that we used in the proposed rule.

Other commenters suggested that we also create separate cost centers for implantable devices and nursing. As noted in the comments, the MedPAR file does not contain the necessary detail to identify a separate cost center for implantable devices or nursing. In addition, we did not have enough time to evaluate whether it would be reasonable to utilize a nursing cost center in the methodology in the future. However, we anticipate undertaking further analysis of the relative weight methodology over the next year in conjunction with the research we are doing on charge compression to determine if additional cost centers are necessary.

Comment: Commenters, referring to Table A, “Charge Line Items from MedPAR Included in Cost Center Charge Group,” noted that MedPAR charge descriptions do not match the Form CMS-2552-96 Cost Center description(s) for several cost centers. For example:

(a) MedPAR lists (18) Lithotripsy Charges where the cost reporting form lists Radioisotopes;

(b) MedPAR lists (6) Other Services where the cost reporting form lists Whole Blood and Packed Red Blood Cells;

(c) MedPAR lists (19) Cardiology Charges as including line 54 of the cost report, which is Electroencephalography;

(d) MedPAR lists (16) Blood Administration Charges where the cost reporting form lists ASC (Non-Distinct Part);

(e) MedPAR lists (24) Outpatient Services Charges where the cost reporting form lists Emergency;

(f) MedPAR lists (25) Emergency Room Charges where the cost reporting form lists Ambulance Services;

(g) MedPAR lists (26) Ambulance Charges where the cost reporting form lists Renal Dialysis;

(h) MedPAR lists (29) ESRD Revenue Setting Charges where the cost reporting form lists Clinic;

(i) MedPAR lists (30) Clinic Visit Charges where the cost reporting form lists Other Outpatient Services, Other Ancillary, Home Program Dialysis and Ambulance Services;

(j) Ambulance Services appear to be included twice, once in (30) Clinic Visit Charges and once in (25) Emergency Room Charges;

(k) Lithotripsy is included in Radiology Services;

(l) Line 62 “Observation Beds” is not reflected separately in Table A; and

(m) Line 68 “Other reimbursement” of the cost report is not listed in Table A.

In addition, commenters were unclear as to whether CMS accounted for subscripted lines in the cost report when calculating CCRs. The commenters noted that subscripted lines did not appear in Table A. Commenters believed this inconsistency in reporting may lead to distorted DRG weights. Therefore, commenters recommended that CMS examine this issue thoroughly before implementing cost-based weights. Several commenters requested that CMS publish a crosswalk of the revenue codes that are used for each MedPAR charge data group and require intermediaries to review cost report data to ensure that providers have reported data consistent with the mapping to the MedPAR data.

Response: We wish to clarify to the commenters that the charge description titles shown in the MedPAR charge description column in Table A were not meant to also be interpreted as the title for each of the cost report line items. That is, we were simply using Table A to illustrate the MedPAR charge groups and the cost report line numbers that were used to create the 10 proposed cost centers. To alleviate this confusion, we are revising Table A to show both the MedPAR charge titles and the titles of the cost report line items. In response to comments (j) and (l), we note that the cost report line item number 65 for ambulance was inadvertently listed twice in the proposed rule; line item 62, observation beds, was used in establishing the CCR for the other services category. Line 65 for ambulance was only used once in the actual other services CCR calculation. Line item 62 should have appeared in the “other services” cost center grouping printed in Table A in the proposed rule. We have corrected this error in the final version of Table A. In addition, in regards to comment (k) above, we have moved the lithotripsy charges from MedPAR to the “other services” cost center grouping and we have also revised the CCR for “other services” to include the cost report line item 43 for radioisotopes, which was formerly included in the radiology CCR.

In response to the commenters’ question regarding the inclusion of subscripted lines, when we calculated the CCRs for the proposed rule and subsequently for this final rule, we relied on a HCRIS data set that contains rolled-up cost report fields such that line items which are subscripted contain the total value for the line item and any subscripted lines below. Therefore, most subscripted lines were included in the proposed rule CCRs and continue to be included in the final rule CCR calculations. However, some subscripted line items are not rolled up and continue to have their own field on the HCRIS data set that we used to calculate the CCRs. Therefore, we are now including the cost report line item 6201 for observation beds, the cost report line item 6350 for Rural Health clinics and the cost report line item 6360 for Federally Qualified Health clinics in the other services CCR. Cost report line items 6350 and 6360 are only reported by provider-based Rural Health clinics and Federally Qualified Health clinics and are necessary in order to identify all incurred costs applicable to furnishing an observation bed prior to a decision to admit a patient to the hospital. Further, we are now including the cost report line item 68 for other reimbursement in the other services CCR, and we are including professional services charges from MedPAR in the other services charge grouping. In response to the commenters’ requests that we show the revenue codes that comprise the MedPAR charges, we have also inserted an additional column in Table A that lists the revenue codes MedPAR groups into each charge field that we are using in the final 13 cost centers. The final version of Table A appears below:

Comment: Many commenters warned that the redistribution of payments from the surgical to the medical DRGs under the proposed methodology may create unintended consequences. Several of these commenters stated that this redistribution poses a threat to patients' access to the latest medical advances and highest quality care. They feared that hospitals will invest less in new medical technologies because Medicare would not pay sufficiently for the DRGs that use them. Another commenter stated that the increased reimbursement for psychiatric DRGs may create an incentive for IPFs to decertify and become inpatient units.

Response: We appreciate the commenters' concern that payment redistribution may create the potential for unintended consequences. However, we wish to emphasize that the redistribution of payments among DRGs is necessary to improve payment accuracy and eliminate the distortions in the current IPPS payment rates. Under the methodology in this final rule, we will increase payment for relatively underpaid cases and reduce payment for relatively overpaid cases.

We are adopting a methodology that will realign payments with costs to pay more appropriately for services rendered by hospitals. Therefore, we do not believe altering the DRG relative weighting methodology will affect patients' access to quality medical care. Patients should have continued and uninterrupted access to new, innovative technologies.

We have analyzed the impact of the increased reimbursement for psychiatric DRGs in response to the commenter's concern that increased reimbursement may provide incentives for IPFs to decertify their units and be paid under the IPPS. Because of the differences in payment between the IPPS and the IPF PPS, we do not believe that the DRG relative weights we are adopting in this final rule will provide increased incentive for IPFs to decertify units. Whereas under the IPF PPS, hospitals receive a daily base rate and adjustments to account for certain patient and facility characteristics, hospitals paid under the IPPS are paid a specified amount based on the DRG for the same cases, regardless of the length of the hospital stay. Our analysis suggests that even though the average payment per day (total payment divided by average length of stay) for the psychiatric DRGs in the IPPS proposed rule may be higher than under the IPF PPS, the total average payment per episode of care remains lower (product of the average IPF payment per day and the average length of stay). Thus, because payments per episode of care remain lower under the IPPS than under the IPF PPS, we are not concerned that IPFs will decertify to get paid using the IPPS. In addition, as indicated above, we are making some modifications to our methodology in response to the public comments. Based on these changes, the increase in the relative weights for the psychiatric DRGs presented in this final rule will not be as significant as those contained in the proposed rule.

Comment: Commenters expressed concern that because hospitals often allocate charges on the cost reports differently than charges on the claims, the cost-center level CCRs are calculated based on a different set of charges than the charges on the claims to which the CCRs are later applied. Commenters expressed concern that Medicare cost report data are not detailed enough or consistently reported accurately to determine costs accurately at a DRG level since such data lack specific cost data on individual items and services. They reiterated that the Medicare cost reports, which serve as the primary source of data under the proposed system, were not designed to be used in a prospective payment system and have not been used to establish hospital rates for inpatient services for some time. They noted several limitations in using the cost reports to derive estimated costs utilized in the DRG relative weight calculations that should be carefully examined and addressed before moving forward with the proposed system of hospital-specific cost weights.

First, the commenters believed that CMS should address cost report accuracy. The commenters stated that because the cost reports have only been used for payment in limited circumstances (DSH, IME, outlier policy), hospitals have had little incentive to report accurately and completely for the services provided to Medicare beneficiaries. In addition, they claimed the cost reports do not contain the level of detail necessary to accurately determine costs at the DRG level. Instead, the cost report provides payments, costs, and some reimbursement totals by department or cost center. The commenters also advised that CMS perform additional auditing of the cost reports to ensure accuracy. The commenters were concerned that if CMS implements a cost-based weighting methodology, the DRG weights will be based on largely un-audited cost reports since approximately 15 percent of hospital cost reports are audited each year. They noted that MedPAC estimated that a full-scale audit could require 1,000 to 2,000 hours from a fiscal intermediary, as well as additional time and resources from the hospital. In addition, a few commenters stated that CMS should only use final settled cost report data, not as-submitted data, in calculating DRG weights.

Second, some commenters contended that CMS should evaluate the overall timeliness of cost report data. They stated that cost report data used to recalibrate the DRG weights are outdated and significantly older than the charge-based data currently used to determine DRG weights under the IPPS. Under the proposed methodology, CMS used hospital claims data from FY 2005 and hospital cost reports from FY 2003. The commenters were concerned that because a lag between the cost report year and the payment year exists, the proposed methodology would rely on older data that does not reflect the costs of many newer technologies. The commenters supported an approach that uses more recent claims and cost report data and also urge CMS to explore options for using alternative data sources that include current information on the costs of inpatient care.

Third, the commenters stated that CMS should examine the comparability of cost reports due to variability in how hospitals allocate costs. Commenters explained that a cost allocation methodology must be used to estimate the cost of individual items and services from the aggregate costs reported for each cost-center on the cost-report. They stated that the proposed methodology assumes that all hospitals consistently allocate costs to the same cost centers. However, hospitals may have inconsistent cost accounting practices or use different cost allocation methods (for example, utilization or square-footage) according to the commenters. The commenters suggested these factors and the compression of charges both within and across cost-centers, limits the usefulness of cost report data to accurately estimate costs. According to the commenters, each hospital uses its own method to allocate costs among cost centers, often resulting in cost assignments that do not reflect the departments to which charges are assigned in the MedPAR data. For example, some commenters indicated that they included cardiac catheterization in lines other than 53 and 54 that group to the cardiac cost center. In addition, several commenters noted that hospitals report medical supply costs inconsistently. While some report them in the supply cost center, others report the medical supply cost in the cost center for the procedure in which the device was used (that is, medical supplies specific to the Emergency Room are included in line 61 of the cost report). The commenters suggested that more specific cost report instructions may be necessary to ensure that hospitals report the information correctly and consistently. Some commenters believed that cost report data were not intended or designed to be used to develop accurate payment rates and suggested developing a proxy to more accurately allocate costs at the DRG level, such as collecting data from hospitals that utilize “sophisticated cost accounting tools that provide more accurate allocation of costs.”

Some commenters also recommended that CMS convene an expert panel to explore ways to address the current limitations of the cost report. They stated that this effort should identify methods to better use or improve hospital cost reports for use in setting the inpatient and outpatient relative weights. The expert panel should aim to identify changes to the cost report that reduce the net information burden on hospitals, while improving overall payment accuracy. The panel should report its recommendations by April 2007 to enable CMS enough time to consider the recommendations in setting the relative weights for FY 2008. Other commenters advocated that CMS initiate a national project to correct any misalignments between cost and charges in cost reports and on the MedPAR claims. Other commenters suggested that CMS postpone the adoption of the proposed HSRVcc methodology until such time that providers improve the accuracy of the source data used in the determination of the DRG weights.

Response: With respect to the commenters' recommendation regarding the reporting of costs and charges for services, CMS requires hospitals to report their costs and charges through the cost report with sufficient specificity to support CMS' use of cost report data for monitoring and payment. Within generally accepted principles of cost accounting, CMS allows providers flexibility to accommodate the unique attributes of each institution's accounting systems. For example, providers must match the generally intended meaning of the line-item cost centers, both standard and non-standard, to the unique configuration of department and service categories used by each hospital's accounting system. Also, while the cost report provides a recommended basis of allocation for the general service cost centers, a provider is permitted, within specified guidelines, to use an alternative basis for a general service cost if it can support to its intermediary that the alternative is more accurate than the recommended basis. This approach creates internal consistency between a hospital's accounting system and the cost report but cannot guarantee the precise comparability of costs and charges for individual cost centers across institutions.

However, we believe that achieving greater uniformity by, for example, specifying the exact components of individual cost centers, would be very burdensome for hospitals and auditors. Hospitals would need to tailor their internal accounting systems to reflect a national definition of a cost center. It is not clear that the marginal improvement in precision created here is worth the additional administrative burden. The current hospital practice of matching costs to the generally intended meaning of a cost center ensures that most services in the cost center will be comparable across providers, even if the precise composition of a cost center among hospitals differs. Further, every hospital provides a different mix of services. Even if CMS specified the components of each cost center, costs and charges on the cost report would continue to reflect each hospital's mix of services. At the same time, internal consistency is very important to the IPPS. Costs are estimated on claims by matching CCRs for a given hospital to their own claims data through a cost center-to-revenue code crosswalk.

Despite the concerns raised in the comments, we believe that costs and charges are reported through the cost report with sufficient specificity to support CMS' use of cost report data to develop cost-based weights. The information we obtained from the cost report on the differing level of charge markups occurring between routine and ancillary hospital departments supports MedPAC's conclusions that the most profitable DRGs that are leading to the development of specialty hospitals are those that require a lot of ancillary services with high markups and low CCRs. To the extent that charge markups vary significantly between the various routine and ancillary hospital departments, we believe that there is a need to adjust charges to cost prior to setting the relative weights. We will continue to rely on the cost report to establish the CCRs that we are finalizing to use to adjust the DRG charges to costs.

However, we continue to be interested in receiving suggestions on ways that hospitals can uniformly and consistently report charges and costs related to all cost centers that also acknowledge the ubiquitous tradeoff between greater precision in developing CCRs and administrative burden coupled with reduced flexibility in hospital accounting practices. Another issue to consider is the potential changes to the relative weights from undertaking efforts of this magnitude that will be costly for both CMS, its fiscal intermediaries and costly and burdensome to hospitals. Although we are not modifying the cost report or our cost report instructions at this time, we would be open to making improvements in the future.

Comment: Several commenters applauded CMS' efforts to find “an administratively feasible approach to improving the accuracy of the DRG weights.” However, they expressed serious concerns about whether the proposed approach achieves that goal. Many commenters asserted that CMS proposes to move to a new cost-based methodology without offering any evidence that the proposed method actually improves payment accuracy.

A few commenters submitted analyses that suggest that the impact of the proposed HSRVcc methodology is substantially different than the MedPAC recommendations, and may even decrease payment accuracy relative to the charge-based weights. A few commenters specifically noted that cardiac procedures are more adversely impacted by the HSRVcc methodology. The proposed methodology reduces relative weights for the three major implantable cardioverter defibrillator (ICD) DRGs (515, 535, and 536) by 25 percent or more. While these proposed reductions imply that the weights based on the existing charge-based methodology overstate the costs of ICD procedures and therefore overpay them, the commenters presented analyses suggesting that these cases are actually underpaid. One such analysis by MedPAC, in its report on physician-owned specialty hospitals, found ICD procedures to have “lower marginal” profitability or “possibly a loss” for hospitals, based on calculation of payment-to-cost ratios and surveys of specialty hospitals. They also indicated that CMS, in approving cardiac resynchronization therapy defibrillators (CRT-D) for new technology add-on payments, found the device to be inadequately paid and granted the add-on payments to defray the costs of the therapy. Given that payment rates under the charge-based weights appear to be inadequate in many of the cardiovascular DRGs, the commenters believed the severe reductions resulting from the proposed HSRVcc methodology appear to be unjustified and provide ample reason to believe that the proposed methodology does not accomplish the goal of improving payment accuracy.

These commenters emphasized that while measuring improved payment accuracy is difficult, the large degree to which the weights fluctuate given the methodological changes alone indicates the need for further analysis and study. The commenters believed CMS should publish reliable indicators that demonstrate how the goal of payment accuracy is achieved. One commenter requested that CMS produce and publish estimates of payment-to-cost ratios and the relative profitability by DRG to determine the effectiveness of different weight-setting and patient classification methodologies in improving overall payment accuracy. The commenter emphasized that such estimates must be adjusted to account for the cost of providing services that include high-technology devices that are understated in the cost reports. Another commenter recommended that CMS construct a process to test the sensitivity of weights to various methodological assumptions and publicly share the results, including: a comparison of the CMS weights to MedPAC's HSRV cost approach; a comparison of CMS weights to an approach using standardized costs (as opposed to HSRV); comparison of CMS weights to weights calculated by estimating costs at the claim level using the 10 cost center approach; evaluation of other alternative methodologies for estimating costs; and an evaluation of the stability of weights over time.

Response: We appreciate the commenters concerns regarding the HSRVcc relative weight setting methodology we proposed and the large change in the relative weights that result from the application of this methodology. As we stated in the FY 2006 IPPS final rule, given the potential for significant redistribution in payments, the MedPAC recommendations should be studied extensively before any broad fundamental changes are made to the current system. In the proposed rule, we provided the results of such an extensive analysis and concluded that changes can be made to the relative weight methodology and the DRG system to improve payment accuracy. Although we agree that adopting a methodology that results in large changes in payment should not be adopted without careful study, we do not believe that the mere presence of such significant impacts invalidates the methodology. On the contrary, we believe large payment impacts may suggest there is a significant degree of distortion present in the current payment system. In our view, we believe that the changes to the IPPS should be evaluated based on whether they represent an improvement to the current system. MedPAC has studied the IPPS extensively and found that improvement can be found in payment accuracy from adopting its recommendations that are similar to those we proposed.

While we acknowledge the need for further study and evaluation of the HSRVcc methodology, we continue to believe that the differential markups among departmental CCRs have introduced distortion into the charge-based relative weights. We note that MedPAC found that “the current payment system encourages community hospitals to allocate capital to profitable services such as cardiology and stimulates the formation of specialty hospitals that often focus on providing profitable services and tend to care for low-severity patients.” The information we obtained from the cost reports on the differing level of charge markups occurring between routine and ancillary hospital departments supports MedPAC's conclusions that the most profitable DRGs that are leading to the development of cardiac specialty hospitals are those that require a lot of ancillary services with high markups and low CCRs. We note that the proposed rule showed that these hospitals are almost exclusively affected by changes to the relative weight methodology providing further evidence of bias and distortion in the relative weights by setting them using hospital charges. To the extent that charge markups vary significantly between the various routine and ancillary hospital departments, we believe that there is a need to adjust charges to cost prior to setting the relative weights. Although it suggested refinements to CMS' proposal (all of which we have adopted in this final rule), we note that MedPAC found that the CMS proposals made great strides toward achieving the goal of improvements in payment accuracy. Therefore, as discussed in section II. C., we are using the national average CCRs to adjust the cost center charges for each DRG to cost prior to setting the relative weights. While we acknowledge that no payment methodology can be perfect because DRG-specific costs cannot be determined, we believe the cost-based methodology we are finalizing in this rule represents a significant improvement over the current charge-based methodology for all of the reasons we specified above. Under the cost-based methodology in this final rule, we will increase payment for relatively underpaid cases and reduce payment for relatively overpaid cases. We believe this reform is badly needed to reduce the bias in the weights and make more appropriate payments for both medical and surgical DRGs.

In order to mitigate the impact of the changes in the relative weights, we are implementing the new cost-based weight methodology in a 3-year transition, where the weights in the first year will be set based on 33 percent of the cost-based weight and 67 percent of the charge based weight. We will continue to study the HSRVcc methodology, the potential effects of charge compression and ways in which we can better account for severity of illness within the DRG system in the coming year.

With respect to the changes in the new patient classification system, the proposed rule noted that we modeled the CS DRGs and observed a 12-percent increase in the explanatory power (or R-quare statistic) of the DRG system to explain hospital charges. That is, we found more uniformity among hospital total charges within the CS DRGs than we did with Medicare's current DRG system (71 FR 24027). Thus, we believe that there is clear evidence that improvements can be made to the current DRG system that will reduce heterogeneity among patients within a given DRG. While this statistic indicates that the current CMS DRG system can be refined to improve payment accuracy, we agree that it does not necessarily mean we should adopt the system we proposed. For a variety of reasons explained further below, we believe that a number of factors must be considered in deciding how to revise the DRG system to better recognize severity of illness.

Comment: One commenter asserted that CMS published incorrect and deficient information about the HSRVcc methodology, its impact on hospitals, and the underlying data utilized in developing the proposed rule. Specifically, the commenter believed the HSRVcc methodology was flawed and therefore stated that the published impacts were inaccurate. The commenter believed that we failed to comply with the Federal Data Quality Act, and OMB, HHS, and CMS Guidelines which address the quality of the data used for policy development, in particular, meeting standards of utility, objectivity, integrity, and transparency and reproducibility. Because the commenter believed that we have violated these data quality standards, the public was deprived of the opportunity to submit meaningful comments, as required by the Administrative Procedure Act (APA). The commenter urged CMS to take the appropriate steps that would result in the withdrawal of the FY 2007 IPPS proposed rule and the publication of a new proposed rule.

Response: We disagree with the commenter's claims that the data utilized in the development of the FY 2007 IPPS proposed rule were materially flawed, did not comply with the Federal Data Quality Act, and did not meet established OMB, Department and CMS guidelines for data quality. The data sources used in estimating the payment impacts from policy changes proposed in the FY 2007 IPPS proposed rule were the HCRIS files that contain Medicare cost report data, the MedPAR files that contain Medicare claims data, the OSCAR database, and the PSF (which is maintained by the fiscal intermediaries and used in paying Medicare claims). These are the best and most reliable data sources available to CMS for modeling the impacts of policy changes. We note that these same databases are used in modeling payment impacts under the LTCH PPS, the OPPS, the IRF PPS, and the IPF PPS, as well as other payment systems. We also note that the comment period to the FY 2007 IPPS proposed rule provided commenters with an opportunity to bring to our attention specific examples of incorrect or inaccurate data. In addition to our posting the impact files from the FY 2007 IPPS proposed rule on the CMS Web site, as always, commenters had access to the same CMS data files that we utilized through communication with our Office of Information Services (OIS).

The fact that the data we used in the development of the FY 2007 IPPS proposed rule were available and transparent to the public was attested by the detailed data analyses included with a significant number of the public comments we received on the FY 2007 IPPS proposed rule. Therefore, for the reasons stated above, we disagree with the commenter's assertion that the data used by CMS in the FY 2007 IPPS proposed rule does not meet the transparency and reproducibility standards. As is the case with any change in policy, we do not base policy decisions on mere assumptions, but rather we analyze the relevant data and any comments submitted in response to a proposed rule.

Comment: One commenter stated that it was unclear whether the weights published for CS DRGs included using the transfer-adjusted charges prior to calculating weights.

Response: We used the hospital's charge on the claim in the HSRVcc methodology. We presume the commenter is asking whether we adjusted the number of cases in setting the relative weights to reflect early transfer to either a post-acute or other acute care setting. We did use transfer-adjusted case counts when we applied the HSRVcc methodology for the relative weights that were shown in Table 5 of the IPPS proposed rule (71 FR 24272) and the “Consolidated severity adjusted DRG HSRVcc relative weights” provided on the CMS Web site at: http://www.cms.hhs.gov/AcuteInpatientPPS/FFD/list.asp#TopOfPage. The case mix index that we use to iterate the proposed FY 2007 HSRVcc weights did not reflect a transfer-adjusted case count. That is, we used the sum of all the case weights divided by the total number of cases unadjusted for transfers to post-acute or other acute care settings.

Comment: Many commenters stated that once a cost-based system is implemented, CMS should provide at least a three-year transition. They stated that a three-year transition is consistent with MedPAC's recommendation to implement the changes to the weights and DRG system over a transitional period. Commenters recommended that payments be made based on a blend of charge and cost-based weights culminating with full cost-based weights at the completion of the transition period.

Response: We have in the past provided for transition periods when adopting changes that have significant payment implications. Given the significant payment impacts upon some hospitals because of these changes to the DRG weighting methodology, we considered options to transition to cost-based weights. We believe the potential payment effects from the changes to the DRG relative weights can be mitigated by adopting a 3-year transition of the relative weights. During the first year of the transition, the relative weights will be based on a blend of 33 percent of the cost-based weights and 67 percent of the charge weights. In the second year of the transition, the relative weights will be based on a blend of 33 percent of the charge weights and 67 percent of the cost-based weights. In the third year of the transition, the relative weights will be based on 100 percent of the cost-based weights.

Comment: One commenter asserted that the proposed changes to improve payment accuracy and to provide payment equity between specialty and general hospitals do not address many of the differences between specialty and full-service hospitals. The commenter stated that hospitals should be reimbursed for the additional services that are required to operate a full-service hospital which are often unnecessary in a specialty hospital setting. The commenter acknowledged that CMS already provides some support to hospitals that serve a high percentage of Medicaid patients through disproportionate share payments. However, the commenter suggested that CMS also make add-on payments to the base DRG payment for expenses such as: operation of a full-service, 24-hour emergency department; operation of a trauma service, a burn unit, or other high cost medically necessary services; sponsoring ground and helicopter ambulance services; operation of 24-hour diagnostic services; provision of round the clock nursing services; and provision of other support services such as clinical pharmacists, nutritionists, case managers, and medical social workers. The commenter believed these add-on payments will encourage hospitals to maintain these services rather than promote specialty hospitals that may be able to operate at a lesser cost without some or all of these services.

Response: Medicare does pay for all of these services through either the IPPS or OPPS payment. We disagree that add-on payments are necessary for services that are commonly provided at many hospitals. The costs of these services will be incorporated in the IPPS or OPPS relative weights. Rather, we continue to believe that Medicare's IPPS payment system needs to be changed to make more equitable payment across all hospitals and decrease the incentive to profit from patient and DRG selection.

Comment: A few commenters stated that although the DRG payment changes proposed by CMS seek to address the proliferation of physician-owned, limited service hospitals in response to recommendations by MedPAC, they do not believe that these payment changes alone will remove the inappropriate incentives created by physician self-referral to limited-service hospitals. They stated that physicians will still have the ability and incentive to refer financially attractive patients to facilities they own, avoid serving low-income patients, and encourage utilization of profitable services. The commenters urged CMS to examine the investment structures of physician-owned, limited service hospitals and to continue the moratorium on issuing new provider numbers to physician-owned, limited service hospitals until the agency's strategic plan has been developed and the Congress has had the opportunity to consider the agency's final report on the topic.

Response: We are in the process of completing the Final Report to Congress and the Strategic and Implementing Plan on Specialty Hospitals, as required by section 5006 of the DRA. Section 5006 of the DRA requires us to consider, among other things, issues of bona fide investment and proportionality of investment with respect to physician investment in specialty (that is, cardiac, orthopedic or surgical) hospitals. Section 5006 of the DRA also provides that the suspension on enrollment of new specialty hospitals that we administratively instituted on June 9, 2005, shall expire upon the date we issue the final report, or, if the report is issued after August 8, 2006, it shall expire on October 8, 2006. We note that Congress has provided for a date certain for the end of the suspension on enrollment of new specialty hospitals. Furthermore, we have not identified a need at this time to continue the suspension beyond that date.

Comment: Many commenters stated that CMS's proposed HSRVcc methodology presented in the FY 2007 IPPS proposed rule failed to address issues of “charge compression.” The commenters explained that “charge compression” describes the common billing practice of hospitals applying higher percentage markups on lower cost items and lower percentage markups on higher cost items. The commenters noted that MedPAC explained that hospitals may reduce the mark-ups for higher-cost items to avoid “sticker shock.” As discussed below, many commenters believed that, to the extent “charge compression” exists, the proposed HSRVcc methodology would lead to systematic differences between estimates of costs and Medicare's payments. Therefore, the commenters believed that the proposal failed to accomplish CMS's stated goal of setting the DRG weights based on accurate cost determinations. If the proposed methodology is implemented, several commenters believed hospitals that perform a large volume of procedures requiring relatively costly supplies/procedures would be severely and unfairly penalized through inappropriately reduced Medicare DRG payments. The treatments they provide would be less likely to be provided, and consequently, Medicare beneficiaries’ access to care may be diminished. Therefore, the commenters stated that if CMS adopts a cost-based DRG weighting methodology, a more accurate measure of determining hospitals’ actual costs must be developed.

Many commenters believed that “charge compression” is a concern because the proposed HSRVcc methodology uses a single CCR for a variety of items and services in a department. Specifically, under the proposed HSRVcc methodology, we proposed to aggregate hospital-level departmental charges into 10 cost centers for each DRG, and then apply national average cost-center level CCRs to determine estimated costs. The commenters asserted that because most hospitals do not apply the same uniform percentage mark-up when setting the charges of each item in the department, the proposed HSRVcc methodology underestimates the cost of relatively more expensive items (particularly devices and implants) and overestimates the cost of relatively less expensive items. The commenters believed that the use of a single CCR for a variety of different items results in a systematic distortion of the estimated costs, and consequently the DRG relative weights that are used in determining the IPPS payment rates. Specifically, many commenters stated that the HSRVcc methodology has a disproportionate adverse impact on DRGs that include implantable technologies and devices, and in some cases would result in Medicare reimbursement that is less than the actual cost of the device.

Some commenters discussed cost data research that has been performed since the implementation of the OPPS to determine the causes and effects of “charge compression.” The commenters asserted that OPPS payment rates are also affected by charge compression. Specifically, one commenter recently commissioned research to investigate whether Medicare claims data provided statistical evidence of “charge compression.” (This research was summarized in an executive summary by Christopher Hogan of Direct Research, LLC. entitled “A Proposed Solution for Charge Compression.”) Many other commenters cited this recent research in their own comments, and recommended that the results of this research be used to develop an adjustment under the proposed HSRVcc methodology to account for “charge compression.” This analysis utilized the detailed coding of charges for supplies by revenue center on the Medicare claims data in the Standard Analytical Files (SAF) to divide the single cost-center CCR for “supplies and equipment” used under the proposed HSRVcc methodology into separate cost-center CCRs for 5 supplies subcategories (general supplies; implantables; sterile supplies; pacemakers and defibrillators; and all other supplies) based on a “strong statistical association between mix of charges for supplies (by revenue center) in a hospital and the [overall] supplies CCR in a hospital.” Using these data from all hospitals, a regression analysis yielded a single “set of CCR adjustments reflecting national average CCRs for [each of] the [five supplies] sub-categories.” This national-average set of adjustments is applied to each hospital (and combined with each hospital's actual supplies CCR) to determine an adjusted estimate of cost on each hospital's claim in the MedPAR file. The results of this research showed that this variation in CCRs across the supplies subcategories would result in weights for some DRGs being significantly different than under the HSRVcc methodology. In particular, the methodology advocated by Hogan would increase the relative weights “for DRGs with substantial charges in the implantable devices and pacemaker/defibrillator revenue centers.”

The commenters pointed out that the results of this research are consistent with previous analyses demonstrating “charge compression” in hospitals’ billing patterns. The commenters also noted that this research was conducted exclusively on Medicare claims data, without supplementation with any external data. The commenters believed that this research demonstrates that an adjustment for “charge compression” is possible. They further asserted that the research provides a solid analytical basis for a specific adjustment. The commenters advocated that we use the coefficients from this regression analysis to develop a “data-driven” adjustment to the CCRs for the supplies and equipment to address the distortion caused by “charge compression.”

Another commenter supported the idea of a “charge compression” adjustment but suggested that CMS should ensure appropriate stakeholder involvement before applying such a policy. Other commenters also advocated for the use of data from the SAF to analyze the relationship between costs and charges for non-implantable supplies and equipment to determine whether an adjustment to the medical-surgical supplies cost center on the MedPAR files to account for “charge compression” is also warranted.

As a result of the concerns discussed above, many commenters stated that any change toward a cost-based DRG weighting methodology under the IPPS must address the distortion caused by “charge compression” and must ensure that the methodology utilizes accurate cost determinations. Consequently, some commenters requested a delay in the implementation of the cost-based DRG weighting methodology until an adjustment for “charge compression” can be incorporated. In addition, some commenters stated that such an adjustment should also be used to address “charge compression” under the OPPS. Several commenters recommended that, in addition to including an adjustment for “charge compression,” the methodology for determining the cost-based DRG relative weights be developed without employing the HSRV methodology. However, a few other commenters endorsed the proposed HSRVcc methodology, stating that the “HSRVcc methodology more closely represents the cost of providing services than the current charge-based system.”

Several commenters referenced various research studies on this issue undertaken over the past 5 to 6 years. These commenters asserted that the research supports the existence of “charge compression” and its systemic distortion in payment rates. The commenters also stated that “although evidence of the effect of charge compression is not new,” research that could support an adjustment to offset charge compression was not available. However, according to the commenters, “research just completed now presents a solution.”

Response: We appreciate the commenters’ concerns regarding charge compression and its impact on the relative weight calculations under the proposed HSRVcc methodology. We are interested in further studying the analytic technique suggested in the comments of using a regression analysis to identify adjustments that could be made to the CCRs to account for charge compression. We note that the Hogan study's regression model was only applied to expensive medical supplies and devices and was not applied uniformly to develop potential adjustments that could be made to costs and charges across all revenue and cost centers that could potentially be subject to charge compression. If such a model were to be applied, we believe further analysis would have to be undertaken to determine whether it should apply to all costs and revenue centers. At this time, we intend to research whether a rigorous model should allow an adjustment for “charge compression” to the extent it exists. Accordingly, we have engaged a contractor to undertake a study on charge compression and review the statistical models provided to us by the commenters. To the extent that we find “charge compression” exists, we will further study potential models that could adjust for it so we can develop more accurate systems of cost-based weights to better reflect the relative costs of the different types of services provided under the IPPS. As suggested in the comments, we plan to fully involve appropriate stakeholders in future analysis of this issue to the extent feasible. Before implementing such an adjustment, we would fully describe our analysis and a potential proposed adjustment as part of the IPPS proposed rule for FY 2008.

Further, we intend to use the charge compression study that we will conduct over the next year as an opportunity to better understand the costs of medical devices. The United States faces a dilemna in health care. Although the rate-of-increase in health care spending slowed last year, costs are still growing at an unsustainable rate. One reason health care costs are rising so quickly is that most consumers of health care are frequently not aware of the actual cost of their care due to lack of transparency. We believe that cost, quality, and patient satisfaction information should be available across the spectrum of care.

Transparency of device pricing is a key aspect of consumer understanding of the cost of health care. We believe that the enhanced understanding of device pricing that will be brought about as part of our charge compression study will help accelerate the public release, in a consumer friendly fashion, of pricing information of medical devices. The public release of device pricing will help augment our overall efforts to empower consumers with better information on the health care they require.

In addition, we note that in order to mitigate the impact of adopting a revised methodology for calculating DRG weights, we are standardizing charges for MedPAR claims using the same methodology we have used in past years, rather than using the HSRV methodology. However, as discussed in detail in section II.E. of this preamble to the final rule, we are adopting our proposal to adjust charges to account for costs prior to establishing DRG weights. However, we anticipate undertaking further analysis of the hospital-specific methodology over the next year in conjunction with the research we are doing on charge compression. If our analysis suggests that an adjustment for charge compression should be applied and/or that the hospital-specific methodology will result in relative weights that more closely approximate the relative costs of care, we will propose further changes for FY 2008. In the interim, we are further mitigating the potential payment effects from the changes to the DRG relative weights by adopting a 3-year transition of the relative weights. During the first year of the transition, the relative weights will be based on a blend of 33 percent of the cost-based weights and 67 percent of the charge weights. In the second year of the transition, the relative weights will be based on a blend of 33 percent of the charge weights and 67 percent of the cost-based weights. In the third year of the transition, the relative weights will be based on 100 percent of the cost-based weights.

3. Refinement of DRGs Based on Severity of Illness

For purposes of the following discussions, the term “CMS DRGs” means the DRG system we currently use under the IPPS; the term “APR DRGs” means the severity DRG system designed by 3M Health Information Systems that currently is used by the State of Maryland; and the term “consolidated severity-adjusted DRGs (CS DRGs)” means the DRG system based on a consolidated version of the APR DRGs (as described in detail below). We discussed the CS DRGs in the FY 2007 IPPS proposed rule and solicited public comments on whether there are alternative DRG systems that could result in better recognition of severity than the CS DRGs we were proposing. As we made clear in the proposed rule, there are still further changes that are important to make to the CS DRG system before it is ready for adoption. In the remainder of this final rule, “CS DRGs” refers to the DRG system we analyzed and proposed for adoption in FY 2008. However, as we indicate below, we received a number of public comments about the proposed CS DRGs, potential alternatives, and a number of other issues related to our proposal. Below we summarize those comments, respond to the comments, and present our plans for adopting a severity-adjusted DRG system for FY 2008.

In the FY 2006 IPPS final rule (70 FR 47474), we stated that we would consider making changes to the CMS DRGs to better reflect severity of illness among patients. We indicated that we would conduct a comprehensive review of the CC list as well as consider the possibility of using the APR DRGs for FY 2007. We did not adopt APR DRGs for FY 2006 because such an adoption would represent a significant undertaking that could have a substantial effect on all hospitals. There was insufficient time between the release of the MedPAC report in March 2005 and the publication of the FY 2006 IPPS final rule for us to analyze fully a change of this magnitude. Instead, we adopted a more limited policy by implementing severity-adjusted cardiac DRGs.

After publication of the FY 2006 IPPS final rule, CMS contracted with 3M Health Information Systems to further analyze the MedPAC recommendations in support of our consideration of possible changes to the IPPS for FY 2007. Under one task of this contract, 3M Health Information Systems analyzed the feasibility of using a revised DRG system under the IPPS that is modeled on the APR DRGs Version 23 to better recognize severity of illness. The APR DRGs have been used successfully as the basis of Belgium's hospital prospective global budgeting system since 2002. The State of Maryland began using APR DRGs as the basis of its all-payer hospital payment system in July 2005. More than a third of the hospitals in the United States are already using APR DRG software to analyze comparative hospital performance. Many major health information system vendors have integrated this system into their products. Several State agencies utilize the APR DRGs for the public dissemination of comparative hospital performance reports. APR DRGs have been widely applied in policy and health services research. In addition to being used in research by MedPAC, the APR DRGs also contain a separate measure of risk of mortality that is used in the Quality Indicators of the Agency for Healthcare Research and Quality, the Premier Hospital Quality Incentive Demonstration discussed in section IV.B. of this preamble, and the Joint Commission on Accreditation of Healthcare Organizations (JCAHO) hospital accreditation survey process (Shared Visions-New Pathways).

Below we present a comparison of the CMS DRG system and the APR DRG system.

a. Comparison of the CMS DRG System and the APR DRG System

The CMS DRG and APR DRG systems have a similar basic structure. There are 25 MDCs in both systems. The DRG assignments for both systems are based on the reporting of ICD-9-CM diagnosis and procedure codes. Both DRG systems are composed of a base DRG that describes the reason for hospital admission and a subdivision of the base DRG based on other patient attributes that affect the care of the patient. For surgical patients, the base DRG is defined based on the type of procedure performed. For medical patients, the base DRG is defined based on the principal diagnosis. In Version 23.0 of the CMS DRG system, there are 367 base DRGs and 526 total DRGs. In Version 23 of the APR DRG system, there are 314 base DRGs and 1,258 total APR DRGs. Some of the base DRGs in the two systems are virtually identical. For example, there is no significant difference between the base DRG under both systems for medical treatment of congestive heart failure. For other base DRGs, there are substantial differences. For example, in the CMS DRG system, there are two base DRGs for appendectomy (simple and complex); in the APR DRG system, there is only one base DRG for appendectomy (the relative complexity of the patient is addressed in the subsequent subdivision of the base DRG into severity of illness subclasses).

The focus of the CMS DRGs is on complexity. Complexity is defined as the relative volume and types of diagnostic, therapeutic, and bed services required for the treatment of a particular illness. Thus, the focus of payment in the CMS DRG system reflects the relative resource use needed by the patient in one DRG group compared to another. Resource use is generally correlated with severity of illness but intensive resource use does not necessarily indicate a high level of severity in every case. It is possible that some patients will be resource-intensive and require high-cost services even though they are less severely ill than other patients. The CMS DRG system subdivides the base DRGs using age and the presence of a secondary diagnosis that represents a CC. The age subdivisions primarily relate to pediatric patients (those who are less than 18 years of age). Patients are assigned to the CC subgroup if they have at least one secondary diagnosis that is considered a CC. The diagnoses that are designated as CCs are the same across all base DRGs. The subdivisions of the base CMS DRGs are not uniform: Some base DRGs have no subdivision; some base DRGs have a two-way subdivision based on the presence of a CC; and other base DRGs have a three-way subdivision based on a pediatric subdivision followed by a CC subdivision of the adult patients. In addition, some base DRGs in MDC 5 (Diseases and Disorders of the Circulatory System) have a subdivision based on the presence of a major cardiovascular condition or complex diagnosis.

The APR DRG system subdivides the base DRGs by adding four severity of illness subclasses to each DRG. Under the APR DRG system, severity of illness is defined as the extent of physiologic decompensation or organ system loss of function. The underlying clinical principle of APR DRGs is that the severity of illness of a patient is highly dependent on the patient's underlying problem and that patients with high severity of illness are usually characterized by multiple serious diseases or illnesses. The assessment of the severity of illness of a patient is specific to the base APR DRG to which a patient is assigned. In other words, the determination of the severity of illness is disease-specific. High severity of illness is primarily determined by the interaction of multiple diseases. Patients with multiple comorbid conditions involving multiple organ systems are assigned to the higher severity of illness subclasses. The four severity of illness subclasses under the APR DRG system are numbered sequentially from 1 to 4, indicating minor (1), moderate (2), major (3), and extreme (4) severity of illness.

The APR DRG system does not subdivide base DRGs based on the age of the patient. Instead, patient age is used in the determination of the severity of illness subclass. In the CMS DRG system, the CC list is generally the same across all base DRGs. However, there are CC list exclusions for secondary diagnoses that are related to the principal diagnosis. In the APR DRG system, the significance of a secondary diagnosis is dependent on the base DRG. For example, an infection is considered more significant for an immune-suppressed patient than for a patient with a broken arm. The logic of the CC subdivision in the CMS DRG system is a simple binary split for the presence or absence of a CC. In the APR DRG system, the determination of the severity subclass is based on an 18-step process that takes into account secondary diagnoses, principal diagnosis, age, and procedures. The 18 steps are divided into three phases. There are six steps in Phase I, three steps in Phase II, and nine steps in Phase III.

The diagram below illustrates the three-phase process for determining patient severity of illness subclass.

Under the CMS DRG system, a patient is assigned to the DRG with CC if there is at least one secondary diagnosis present that is a CC. There is no recognition of the impact of multiple CCs. Under the APR DRG system, high severity of illness is primarily determined by the interaction of multiple diseases. Under the CMS DRG system, patients are assigned to an MDC based on their principal diagnosis. While the principal diagnosis is generally used to assign the patient to an MDC in the APR DRG system, there is a rerouting step that assigns some patients to another MDC. For example, lower leg amputations can be performed for circulatory, endocrine, or musculoskeletal principal diagnoses. Instead of having three separate amputation base DRGs in different MDCs as is done in the CMS DRG system, the APR DRG system reroutes all of these amputation patients into a single base APR DRG in the musculoskeletal MDC. The CMS DRG system uses death as a variable in the DRG definitions but the APR DRG system does not. Both DRG systems are based on the information contained in the Medicare Uniform Bill. The APR DRG system requires the same information used by the current CMS DRG system. No changes to the claims form or the data reported would be necessary if CMS were to adopt APR DRGs or a variant of them.

The CMS DRG structure makes some DRG modifications difficult to accommodate. For example, high severity diseases that occur in low volume are difficult to accommodate because the only choice is to form a separate base DRG with relatively few patients. Such an approach could lead to a proliferation of low-volume DRGs. Alternatively, these cases may be included in DRGs with other patients that are dissimilar clinically or in costs. Requests for new base DRGs formed on the use of a specific technology may also be difficult to accommodate. Base DRGs formed based on the use of a specific technology would result in the payment weight for the DRG being dominated by the price set by the manufacturer for the technology.

The structure of the APR DRGs provides a means of addressing high severity cases that occur in low volume through assignment of the case to a severity of illness subclass. However, the APR DRG structure does not currently accommodate distinctions based on complexity. Technologies that represent increased complexity, but not necessarily greater severity of illness, are not explicitly recognized in the APR DRG system. For example, in the CMS DRGs, there are separate DRGs for coronary angioplasty with or without insertion of stents. The APR DRGs do not make such a differentiation. The insertion of the stent makes the patient's case more complex but does not mean the patient is more severely ill. However, the inability to insert a stent may be indicative of a patient's more advanced coronary artery disease. Although such conflicts are relatively few in number, they do represent an underlying difference between the two systems. If Medicare were to adopt a severity DRG system based on the APR DRG logic but assign cases based on complexity as well as severity as we do under the current Medicare DRG system, such a distinction would represent a departure from the exclusive focus on severity of illness that currently forms the basis of assigning cases in the APR DRG system.

Section 1886(d)(4) of the Act specifies that the Secretary must adjust the classifications and weighting factors at least annually to reflect changes in treatment patterns, technology, and other factors that may change the relative use of hospital resources. Therefore, we believe a method of recognizing technologies that represent increased complexity, but not necessarily greater severity of illness, should be included in the system. We plan to develop criteria for determining when it is appropriate to recognize increased complexity in the structure of the DRG system and how these criteria interact with the existing statutory provisions for new technology add-on payments. In the FY 2007 IPPS proposed rule, we invited public comments on this particular issue.

Another difference between the CMS DRG system and the APR DRG system is the assignment of diagnosis codes in category 996 (Complications peculiar to certain specified procedures). The CMS DRG system treats virtually all of these codes as CCs. With the exceptions of complications of organ transplant and limb reattachments, these complication codes do not contribute to the severity of illness subclass in the APR DRG system. While these codes could be added to the severity logic, the appropriateness of recognizing codes such as code 998.4 (Foreign body accidentally left during a procedure) as a factor in payment calculation could create the appearance of incentives for less than optimal quality. Although there is no direct recognition of the codes under the 996 category, the precise complication, in general, can be coded separately and could contribute to the severity of illness subclass assignment.

Comment: Some commenters strongly supported including the complication codes (996.00-999.9) when assigning a patient to a severity-adjusted DRG because the codes represent pre-existing or predictably higher risks upon admission for difficult patients who are typically referred to regional centers. The commenters stated that failure to do so will create new incentives for adverse admission selection and underpay hospitals that treat difficult patients. The commenters stated that the 996 codes include some complications that should never be paid (for example, wrong site surgery and instruments left in the patient). However, the commenters indicated that these kinds of complications likely constitute less than one-half of one percent of all complications and revising the DRG system so that all 996 codes are not paid will provide incentives to hospitals to avoid admitting patients that are at high risk because of a pre-existing condition or other circumstance. Another commenter stated that all infections should be removed as complicating conditions under the DRG system.

Response: The discussion in this section of the proposed rule noted that 996 codes are used in assigning a patient to a CMS DRG but not to an APR DRG. Although the discussion in this section of the proposed rule did indicate that using these codes to assign a patient to a DRG may raise questions about incentives for less than optimal quality, the discussion was only intended to note the differences that currently exist between the CMS and the APR DRGs. The commenters raised issues that require further study. We will consider quality of care issues and payment incentives as we consider how to implement section 5001(c) of Pub. L. 109-171 with respect to hospital acquired conditions, including infections. There is a more detailed discussion of this provision of the law in a later section of this final rule.

Table B below summarizes the differences between the two DRG systems:

To illustrate the differences between the two DRG systems, we compare in Table C below four cases that have been assigned to CMS DRGs and APR DRGs. In all four cases, the patient is a 67-year-old who is admitted for diverticulitis of the colon and who has a multiple segmental resection of the large intestine performed. ICD-9-CM diagnosis code 562.11 (Diverticulitis of colon (without mention of hemorrhage)) and ICD-9-CM procedure code 45.71 (Multiple segmental resection of large intestine) would be reported to capture this case. In both DRG systems, the patient would be assigned to the base DRG for major small and large bowel procedures. These four cases would fall into two different CMS DRGs and four different APR DRGs. We include Medicare average charges in the table to illustrate the differences in hospital resource use.

Case 1: The patient receives only a secondary diagnosis of an ulcer of anus and rectum (ICD-9-CM diagnosis code 569.41). Under the CMS DRG system, the patient is assigned to base DRG 149 (Major Small and Large Bowel Procedures Without CC). Under the APR DRG system, the patient is assigned to base DRG 221 (Major Small and Large Bowel Procedures) with a severity of illness subclass of 1 (minor).

Case 2: The patient receives a secondary diagnosis of an ulcer of anus and rectum and an additional secondary diagnosis of unspecified intestinal obstruction (ICD-9-CM diagnosis code 560.9). Under the CMS DRG system, the patient is assigned to DRG 148 (Major Small and Large Bowel Procedures With CC). Under the APR DRG system, the patient is assigned to base DRG 221 and the severity of illness subclass increases to 2 (moderate).

Case 3: The patient receives multiple secondary diagnoses of an ulcer of anus and rectum, unspecified intestinal obstruction, acute myocarditis (ICD-9-CM diagnosis code 422.99), and atrioventricular block, complete (ICD-9-CM diagnosis code 426.0). Under the CMS DRG system, the patient is assigned to DRG 148. Under the APR DRG system, the patient is assigned to base DRG 221 and the severity of illness subclass increases to 3 (major).

Case 4: The patient receives multiple secondary diagnoses of an ulcer of anus and rectum, unspecified intestinal obstruction, acute myocarditis, atrioventricular block, complete, and the additional diagnosis of acute renal failure, unspecified (ICD-9-CM diagnosis code 584.9). Under the CMS DRG system, the patient is assigned to DRG 148. Under the APR DRG system, the patient is assigned to base DRG 221 and the severity of illness subclass increases to 4 (extreme).

The largest significant difference in average charges is seen in case 4 where the average charge under the APR DRG assigned to the patient ($130,750) is more than double the average charge under the CMS DRG assigned to the patient ($59,519).

b. CS DRGs for Use in the IPPS

APR DRGs were developed to encompass all-payer patient populations. As a result, we found that, for the Medicare population, some of the APR DRGs have very low volume. MedPAC noted that the larger number of DRGs under a severity-weighted system might mean that CMS would be faced with establishing weights in many categories that have few cases and, thus, potentially creating unstable estimates. While volume is an important consideration in evaluating any potential consolidation of APR DRGs for use under the IPPS, we believe that hospital resource use and clinical interpretability also need to be taken into consideration. For example, any consolidation of severity of illness subclasses within a base DRG should be restricted to contiguous severity of illness subclasses. Thus, it would not be reasonable clinically to combine severity of illness subclasses 1 and 4 solely because both consist of low-volume cases. We analyzed consolidating APR DRGs by either combining the base DRGs or the severity of illness subclasses within a base DRG. For consolidation across base APR DRGs, we considered patient volume, similarity of hospital charges across all four severity of illness subclasses and clinical similarity of the base APR DRGs. For consolidations of severity of illness subclasses within a base DRG, we considered patient volume and the similarity of hospital charges between severity of illness subclasses. In considering how to consolidate severity of illness subclasses, we believed it was important to use uniform criteria across all DRGs to avoid creating confusing and difficult to interpret results. That is, we were concerned about inconsistencies in the number of severity levels across different DRGs.

The objective to simultaneously take into consideration patient volume and average charges often produced conflict. Table D below contains the overall patient volume and average charge by APR DRG severity of illness subclass. While severity of illness subclass 4 (extreme) has the lowest patient volume of 5.80 percent, we found that the dramatically different average charges between severity of illness subclass 3 (major) and subclass 4 (extreme) patients of approximately $32,426 and $81,952, respectively, would make it difficult to consolidate severity of illness subclass 3 and 4 patients. Conversely, we found that, while the average charge difference between severity of illness subclass 1 (minor) and 2 (moderate) patients was much smaller, of approximately $17,649 and $20,021, respectively, the majority of patient volume (68.08 percent) is in these two subclasses. Thus, low patient volume and small average charge differences rarely coincided.

There were also few opportunities to consolidate base DRGs. For base DRGs in which there was a clinical basis for considering a consolidation, there were usually significant differences in average charges for one or more of the severity of illness subclasses. APR DRGs already represented a considerable consolidation of base DRGs (314) compared to CMS DRGs (367). Thus, we expected that further base DRG consolidation would be difficult.

We reviewed the patient volume and average charges across APR DRGs and found that medical cases assigned severity of illness subclass 4 within an MDC have similar average charges. We observed the same pattern in average charges across severity of illness subclass 4 surgical patients within an MDC. The data suggest that, in cases with a severity of illness of subclass 4, the severity of the cases had more impact on hospital resource use than the reason for admission (that is, the base APR DRG within an MDC). Thus, we believe that, within each MDC, the severity of illness subclass 4 medical and surgical patients, respectively, could be consolidated into a single group.

In some MDCs, it was not possible to consolidate into a single medical and a single surgical severity of illness subclass 4 group. In these MDCs, more than one group was necessary. For instance, Table E below contains the patient volume and average charges for severity of illness subclass 4 cases in MDC 11 (Diseases and Disorders of the Kidney and Urinary Tract). Taking into consideration volume and average charges, except for APR DRG 440 (Kidney Transplant), surgical cases assigned severity of illness subclass 4 in MDC 11 could be consolidated into a single group having 5,492 patients and an average charge of $107,258. However, we decided not to include kidney transplant patients in this severity of illness subclass 4 due to their very high average charges (approximately $203,732 or more than $100,000 greater than other patients in MDC 11 having a severity of illness subclass 4). Average charges within the consolidated severity of illness subclass 4 surgical DRG in MDC 11 show some variation but are much higher than the corresponding average charges for the severity of illness subgroup 3 patients of $48,863. Thus, our analysis suggests that the data support maintaining three severity of illness levels for each base DRG in MDC 11; a separate severity of illness subclass 4 for all patients other than those having kidney transplant; and a separate DRG for kidney transplants.

The consolidation of severity of illness subclass 4 APR DRG into fewer groups was done for all MDCs except MDC 15 (Newborn and Other Neonates With Conditions Originating in the Perinatal Period), MDC 19 (Mental Diseases and Disorders), and MDC 20 (Alcohol/Drug Use and Alcohol/Drug Induced Organic Mental Disorders). In the 22 MDCs in which the severity of illness subclass 4 consolidation was applied, the number of separate severity of illness subclass 4 groups was reduced from 262 to 69.

For MDC 14 (Pregnancy, Childbirth, and Puerperium), the base APR DRGs were consolidated from 12 to 6. Severity of illness subclass 1 through 3 were retained, and severity of illness subclass 4 was consolidated into a single APR DRG, except for cesarean section and vaginal deliveries, which were maintained as separate APR DRGs. This consolidation reduced the total number of obstetric APR DRGs from 48 to 22.

The Medicare patient volume in MDC 15 was very low, allowing for a more aggressive consolidation. For MDC 15, we consolidated 28 base APR DRGs into 7 base CS DRGs. For each of the 7 consolidated base MDC 15 DRGs, we combined severity of illness subclasses 1 and 2 into one DRG and severity of illness subclass 3 and 4 into another DRG. This consolidation reduced the total number of MDC 15 DRGs from 112 in the APR DRG system to 14 CS DRGs.

In MDC 19, we consolidated 12 base DRGs into 4 base DRGs. We retained the 4 severity of illness subclasses in MDC 19 for each of the 4 base DRGs. In MDC 20, the base APR DRG for patients who left against medical advice has severity of illness subclass 1 and 2 consolidated and severity of illness subclass 3 and 4 consolidated. The remaining 4 base DRGs were consolidated into 1 base DRG with 4 severity of illness subclasses.

We did not consolidate any of the pre-MDC subclass 4 APR DRGs such as Heart Transplant. As explained earlier, pre-MDC DRGs are DRGs to which cases are directly assigned on the basis of ICD-9-CM procedure codes. These DRGs are for liver and/or intestinal transplants, heart and/or lung transplants, bone marrow transplants, pancreas transplants, and tracheotomies. For the pre-MDC DRGs, except for Bone Marrow Transplant, we consolidated severity of illness subclasses 1 and 2 into one DRG. In addition, the three base APR DRGs for Human Immunodeficiency Virus (HIV) with multiple or major HIV-related conditions had severity of illness subclasses 1 and 2 consolidated.

In total, we reduced 1,258 APR DRGs to 861 CS DRGs. In Appendix C of this proposed rule, we present the 861 unique combinations of CS DRGs.

Table F below includes a description of the consolidations that we did within each individual MDC and includes information about the total number of DRGs that were eliminated from the APR DRGs to develop the CS DRGs.

Appendix D of the FY 2007 IPPS proposed rule (71 FR 24433) showed the crosswalk of each CS DRG to its respective APR DRG. We numbered the DRGs sequentially and incorporated the severity of illness subclass into the DRG description. However, within the range of sequential numbers used for an MDC, we retained some unused numbers to allow for future DRG expansion. By using a three-digit number for the CS DRGs, we also avoid the need for reprogramming of computer systems that would be necessary to accommodate a change from the current three-digit DRG number to separate fields for the base CS DRG number and the severity of illness subclass.

Severity DRGs represent a significant change from our current DRG system. In addition to changing the way claims are grouped, severity DRGs introduce other issues requiring additional analysis, including possible increases in reported case-mix and changes to the outlier threshold. Our analysis of these issues is outlined further in the next section.

Comment: A number of commenters suggested further refinements that need to be made to the CS DRGs to account for complexity as well as severity. Commenters recommended that CMS create a “task force” to analyze situations in which the complexity of the patients is not always appropriately recognized by the proposed CS DRGs. One commenter stated that the severity system is flawed because it does not capture resource utilization or the utility of technologies that would be more appropriate for beneficiaries.

The commenters also provided examples of base DRG assignments under the current CMS DRGs that are different than those under the CS DRG. For instance, one commenter indicated that high dose interleukin-2 (HD IL2) is used to treat otherwise terminal cancer patients with metastatic renal cell cancer and melanoma. HD IL2 can evoke an immune response that eradicates the tumor and provides a potential opportunity for recovery. In the FY 2004 IPPS final rule, CMS created a new procedure code for HD IL2 therapy and assigned these patients to DRG 492. The commenter reported improved access to HD IL2 therapy as a result of these changes. However, the commenter was concerned that these patients could potentially be assigned to a number of different DRGs under the CS DRGs with a weighted average reduction in the relative weight of 58 percent. The commenter suggested revising the CS DRG to take into account the complexity associated with providing HD IL2 therapy. Other commenters noted:

• Some patients in need of ventricular assist devices (VAD) are currently paid in the same group as heart transplant patients using the CMS DRGs. Other heart assist devices are assigned to DRG 525 (Other Heart Assist Implant). These patients will be paid in the same group as implantable cardiac defibrillator patients under the CS DRGs. The commenters noted that it is possible that payment for these kinds of cases could decline by more than 70 percent under the proposed rule. The commenter believed that the assignment under the CS DRGs will not recognize higher resources associated with treating VAD patients relative to those in need of implantable cardiac defibrillators.
• Bare metal and drug-eluting coronary stents would be assigned to the same CS DRG eliminating the distinction currently made for these two different kinds of stents in the CMS DRGs. The commenters noted that CMS created separate DRGs for drug eluting and bare metal stents to recognize the higher costs of drug eluting stents.
• Defibrillator device replacement cases are currently assigned to DRG 551 (Permanent Cardiac Pacemaker Implant With Major Cardiovascular Diagnoses or AICD Lead or Generator). The commenters were concerned that these cases would be assigned to the DRGs for Permanent Cardiac Pacemaker Implant With & W/O AMI, Heart Failure or Shock and the cases would revert back to classification based on presence or absence of heart failure, AMI, or shock, rather than an MCV.
• Patients receiving tPA thrombolytic therapy for stroke are currently assigned to DRG 559 (Acute Ischemic Stroke With Use of a Thrombolytic Agent). CMS revised the DRGs in FY 2006 to provide a separate DRG for stroke patients being treated with a reperfusion agent. According to the commenter, these patients will be paid in the same group with all stroke cases under CS DRGs undoing the change that CMS made in FY 2006 according to the commenter.
• In FY 2006, CMS created separate DRGs for the revision of hip or knee replacement (DRG 545, Revision of Hip or Knee Replacement) to distinguish the higher resources associated with revisions from original replacements. Under CS DRGs, these cases would be assigned to the same group as the original replacement (bilateral or single) of the specific joint. The commenters were concerned that CMS' proposal to adopt cCS DRGs will undo a proposal that it adopted just 1 year ago.
• Combined anterior/posterior spinal fusion cases are currently assigned to DRG 496 (Combined Anterior/Posterior Spinal Fusion). This procedure requires two separate incisions and turning the patient over during surgery. The commenter expressed concern that under the CS DRG system, these cases would be paid in the same group as all spinal fusions and the new DRGs would not recognize higher costs associated with treating these patients.
• The APR DRG and CS DRG systems do not have DRGs for lung transplants alone or combined kidney/pancreas transplants. The commenter suggested that there should be separate DRGs for these transplants in addition to liver/intestinal transplants. The commenter indicated that lung transplants alone have lower costs and should not be in the same DRG as combined transplants.

Response: In the vast majority of clinical situations, severity of illness and treatment complexity are directly related and are therefore addressed in the CS DRGs. As discussed in the proposed rule, there are a number of clinical situations, primarily related to the use of specific technologies, in which low severity patients receive care with high treatment complexity and cost. We acknowledge that further refinements are needed to the proposed CS DRG system before it will be ready for adoption. In the FY 2007 IPPS proposed rule, we noted a number of concerns we had with adopting the CS DRGs in FY 2007 (71 FR 24027). Among them was our concern that we might need additional time to refine the CS DRGs to better account for complexity as well as severity. The commenters have brought some important issues to our attention that we believe should be carefully considered before we adopt the CS DRGs. We will consider these issues if we were to make further modifications to the CS DRGs and propose adopting them for FY 2008. However, as we indicate elsewhere in this final rule, we have engaged a contractor to assist us with completing an evaluation of alternative DRG systems that may better recognize severity than the current CMS DRGs and meet other criteria that would make them suitable to adopt for purposes of payment under the IPPS. We expect to complete this evaluation of alternative DRG systems quickly this fall as part of moving forward on adopting a revised DRG system that better recognizes severity in the IPPS rulemaking for FY 2008. It is possible that some of the alternatives that we evaluate for better recognizing severity in the DRGs will be based on the current CMS DRGs. If we were to develop a clinical severity concept that uses the current CMS DRGs as the starting point, it is possible that the issues raised by the commenters will no longer be a concern. If, however, we were to propose adopting the CS DRGs for FY 2008, we would consider the issues raised by the commenters as we make further refinements to this DRG system so it accounts for complexity as well as severity as a proxy for relative resource use.

Comment: One commenter suggested a way of accounting for therapeutic complexity when assigning a patient under the CS DRGs. The commenter indicated that the patient should be assigned to a severity of illness subclass based on whether they received a separately identifiable technology that provides a clinical benefit and results in significantly higher case costs independent of severity level relative to the base DRG. The commenter also recommended that complexity levels be superimposed on the proposed severity of illness levels, such that either severity or complexity, or a combination of the two, would increase the classification of a case. The classifications would be defined as severity of illness or complexity (1-4).

Response: We will further consider how to incorporate complexity into the assignment of a patient to a severity of illness subclass under either the CS DRGs if we propose to adopt them in FY 2008 or the alternative DRG system that we will consider once we complete our evaluation of potential DRG systems. It may be possible to assign a case to a severity of illness subclass under either the CS DRGs, the alternative system we plan to evaluate or even underrefined CMS DRGs by using the procedures or services that are provided to the patient as a measure of resource use (that is, complexity). We agree that the use of a separately identifiable procedure or technology may be useful in determining the assignment of a patient to a specific subclass of a base DRG much like what occurs today under the CMS DRGs when assigning patients with placement of a bare metal or drug-eluting stent to separate DRGs.

Comment: Some commenters were concerned that CMS did not propose to adopt the already widely used APR DRGs endorsed by MedPAC, but rather proposed to adopt CMS'-developed CS DRGs. Some commenters stated that the CMS analysis that resulted in the CS DRGs is skewed because Medicare uses a truncated list of diagnosis and procedure codes. The commenter noted that CMS does not use comparable data to what 3M uses for the complete APR DRGs. Another commenter stated that the APR DRGs are the most advanced DRG classification system available yielding the most clinically homogenous groupings and the greatest predictive power. This commenter believed that it provides a sound basis for developing CS DRGs.

Response: MedPAC did not endorse using the APR DRGs. However, MedPAC's analysis that led to their recommendation to refine the current DRGs to more fully account for difference in severity of illness among patients was based on the APR DRGs. Even though MedPAC's analysis was based on the APR DRGs, it recognized that CMS would have to consider a number of different factors when making decisions in the design of a DRG system. For instance, MedPAC noted that the large number of DRGs might mean that CMS would be faced with establishing weights in many categories that have few cases and thus potentially creating unstable estimates. To avoid creating refined DRGs with unstable relative weights, MedPAC recommended that the Secretary should be selective in adopting fine clinical distinctions similar to those reflected in the APR DRGs. Refining the DRGs will require carefully weighing the benefits of more accurate and economic distinctions against the potential for instability in relative weights based on a small number of cases. We do not believe that MedPAC expected that we would adopt the APR DRGs without any changes.

Comment: Some commenters stated concerns with merging of dissimilar patient groups in the CS DRG system. Combining clinically dissimilar groups across the severity dimension has the potential to render the groups far less clinically meaningful. It is anticipated that such groups would have to be restructured frequently as treatment patterns change for primarily very ill patients. Some commenters stated that it seems that more categories may have been consolidated than necessary, giving up clinical and statistical homogeneity unnecessarily. It was noted that this is especially important if the CS DRGs are envisioned as part of the basis for evolving efforts towards value-based purchasing where such measures as post-admission complications and readmissions need to be evaluated on a risk-adjusted basis. An alternative approach was suggested to keep the patient groups separate from a classification perspective, but merge from a payment analysis perspective.

Response: As discussed above, the CS DRGs are based on the APR DRG system. The APR DRG system is comprised of 314 base DRGs, which are divided into four severity of illness subclasses. We believe that the APR DRG greatly improve recognition of resource use and clinical similarity of patients. However, in our analysis of the APR DRG system, we observed that cases assigned severity of illness subclass 4 within an MDC have similar average charges. Furthermore, our clinical consultants frequently considered the severity of illness subclass 4 patients across DRGs within an MDC to have a closer clinical resemblance than to lower severity patients in their respective DRGs because, in severely ill patients, comorbidities have a greater impact on severity than the reason for admisssion. Treatment patterns will evolve for these multiple comorbidities leading to severity level 4 (sepsis, shock, acute renal failure, among others). However, to the extent that these multiple comorbidities will change (for example, better treatment of septic shock so that this occurs less frequently) they should do so equally across all patients within an MDC. With respect to the comment about maintaining more DRG groups for purposes other than payment under the IPPS, we proposed to adopt the CS DRGs only for Medicare inpatient hospital payment. We chose to consolidate the APR DRGs to increase administrative simplicity, minimize the impact on existing claim processing systems, and avoid having multiple DRGs with low case volumes and similar weights. The commenter's suggestion would essentially result in many more DRGs having exactly the same weight. Therefore, we do not see a need to adopt the commenter's suggestion. However, a hospital or any other entity can use an alternative patient classification system for the other purposes suggested in the comment.

Comment: Some commenters stated that the CS DRGs are problematic because they were not designed to accommodate non-Medicare populations. The commenters indicated that many hospitals use DRGs for quality and other outcome measurements and that the proposed CS DRGs may not be clinically appropriate for these purposes.

In addition, another commenter stated that private health insurance company contracts use the CMS DRG relative weights as the payment basis for inpatient services delivered to members under private health insurance plans. The commenter stated that because these contracts are typically negotiated based on a fairly static assumption of CMS DRGs (including classification and weights), the proposed redistribution will disrupt virtually every contract because of the varying services consumed by members covered under private health insurance. The commenter urged CMS to provide a greater lead time in implementing changes to the DRG system and relative weight methodology to allow health insurers more time to model the impact of the methodological changes to their hospital contracts.

Response: We acknowledge that Medicare DRGs are sometimes used by non-Medicare payers for their own purposes. However, CMS' primary focus of updates to the Medicare DRG classification system is on changes relating to payment for services furnished to Medicare beneficiaries, not the obstetric, pediatric, or neonatal population. Cases involving these patients are found far less frequently among Medicare beneficiaries than in the general population. In fact, we applied consolidations to the APR DRGs to develop the CS DRGs to recognize that the APR DRGs were developed to accommodate all patient populations and there would be many DRGs with few Medicare cases or insufficient differences in the relative weights to warrant us maintaining a separate DRG. We encourage other payers that use Medicare's DRG system for payment to make appropriate modifications for patient populations that are found infrequently among Medicare beneficiaries such as neonates and children. Again, as we stated above, a hospital or any other entity can use an alternative patient classification system for purposes other than Medicare payment.

In response to the commenter's concern with regard to the impact on private health insurance plans, we are improving our relative weight methodology to make Medicare payments more accurate. We utilize Medicare specific data to calculate the relative weights designed to pay Medicare costs. We have a fiduciary responsibility to administer the trust fund in order to provide quality care for our beneficiaries and that, not private payer contracts, is our foremost concern. However, as we noted earlier in this section, we are postponing the implementation of the HSRV methodology while we study its impact on charge compression. Instead, we are using a more similar approach to calculating the IPPS relative weights that is used in the OPPS. That is, rather than using a hospital-specific relative weighting methodology, we are standardizing charges to remove relevant payment factor adjustments and then adjusting those charges to costs using national cost center CCRs.

In addition, we are adopting a 3-year transition of the relative weights. We believe this transition may also mitigate any potential impacts to private payer contracts from the changes to the DRG relative weights. During the first year of the transition, the relative weights will be based on a blend of 33 percent of the cost-based weights and 67 percent of the charge weights. In the second year of the transition, the relative weights will be based on a blend of 33 percent of the charge weights and 67 percent of the cost-based weights. In the third year of the transition, the relative weights will be based on 100 percent of the cost-based weights.

Comment: One commenter suggested that CMS seek further refinements to the DRGs for mental services. The commenter suggested that these DRGs have been underpaid for many years.

Response: We will consider whether the psychiatric DRGs need further refinements as we proceed to refine the DRG system to better recognize severity for FY 2008. We note that the application of cost-based weights will increase Medicare's payments for the psychiatric DRGs in FY 2007.

Comment: Some commenters inquired how other prospective payment systems such as the IPF PPS and LTCH PPS that rely upon the IPPS DRG classifications would be affected by the changes to adopt CS DRGs.

Response: We did not propose any changes to the DRG classifications systems used under the IPF PPS or the LTCH PPS in the IPPS proposed rule. However, we acknowledge that these PPSs use the IPPS DRG classifications to make payment determinations. Furthermore, we note that the refinements we are adopting to the current CMS DRG system to better recognize severity (which are discussed in detail in section II.C.7. of this final rule) will be applicable under the IPF PPS and LTCH PPS, just as past annual updates to the IPPS DRG classifications). We will need to consider whether corresponding changes need to be made to these other payment systems once final decisions are made about how DRG classification will occur under the IPPS in the future. Payment rate and policy changes to the IPF PPS and LTCH PPS went into effect for RY 2007 on July 1, 2006. These PPSs are using the Version 23 IPPS GROUPER for the first 3 months of RY 2007 (July 2006 through September 2006). Consistent with the IPPS, the IPF PPS will use Version 24 of the IPPS GROUPER, effective October 1, 2006. No further changes will be made to the IPF PPS until next July. Under the LTCH PPS, changes to the LTC-DRGs were proposed for FY 2007, based on the proposed Version 24 IPPS GROUPER (71 FR 24049 through 24068), and changes to the LTC-DRGs that will be effective October 1, 2006, based on the finalized Version 24 IPPS GROUPER (presented in this final rule) are discussed in section II.F. of the preamble of this final rule. Any changes to the DRG classification systems for these prospective payment systems would be undertaken through notice and comment rulemaking in their respective proposed rules.

Comment: One commenter stated that it was not clear how the judgment was made for the MDC 11 severity subclass 4 example shown that these average charge values were sufficiently similar to consolidate. The commenter suggested that CMS provide further information about the criteria and considerations it used to judge categories as low volume and potentially unstable and to judge the mean charges (or costs) as sufficiently similar to warrant consolidation. One commenter expressed concern about the consolidations related to obstetrics and psychiatric care services.

Response: As discussed above, the CS DRGs are based on APR DRGs that are divided into severity subclasses 2 through 4 subclasses which greatly increase the resource and clinical similarity of the patients. Furthermore, as discussed above, our clinical consultants frequently considered the level 4 severity patients across DRGs within an MDC to have a closer clinical resemblance than to lower severity patients in their respective DRGs. In consolidating the severity level 4 patients in an MDC, volume was a primary consideration along with the extent of clinical difference. For example, in MDC 11 severity level 4, kidney transplants were kept in a separate group and not consolidated with the other MDC 11 surgical DRGs because of the clinical distinctiveness of patients having a major organ transplant.

Comment: One commenter expressed concern that patients may need to suffer adverse consequences in order for the case to be assigned to a higher severity level. The commenter believed that the severity grouping should reflect complexity and patient benefit as well and should allow for an increased severity/complexity level even without adverse patient consequences.

Response: The current DRG system assigns a CC status to most patients with a complication or adverse event that occurs after admission. Although in the CS DRGs post admission complications can result in an increase in a patient's severity level, patients are primarily assigned to the higher severity levels (levels 3 and 4) based on the presence of multiple serious comorbidities in multiple organ systems rather than a single adverse event. Thus, unlike the current DRGs in which a single post admission complication can place the patient in a higher paying DRG, the CS DRGs in general require multiple significant problems to be present in order for a higher severity level to be assigned. In general, these patients will be more costly to treat. The system does not reward “adverse” consequences as suggested by the commenter but instead recognizes severity of illness will also be associated, at least in part, with resource use.

Patients are increasingly admitted to the hospital at high severity of illness. Adverse consequences can and do occur within the hospital. However, some of those consequences are unavoidable (particularly for patients who are admitted at a high severity of illness). Section 5001(c) of Pub. L. 109-171 requires that, beginning in FY 2009, we select diagnosis codes associated with at least two conditions that result in assignment of a higher weighted DRG and that reasonably could be prevented through the application of evidence-based guidelines. Beginning with discharges in FY 2009, section 5001(c) requires that we not assign cases to higher weighted DRGs based on the presence of these preventable conditions. Section 5001(c) also mandates that, for discharges on or after October 1, 2007, we require a hospital to include the secondary diagnosis of a patient at admission as part of the information required to be reported by a hospital for payment purposes. We believe that the concerns of the commenter will be addressed when we implement section 5001(c) of Pub. L. 109-171.

Comment: A number of comments supported CMS' goal of improving payment accuracy. However, the commenters stated that the need for and best approach to changing the patient classification system has not been objectively demonstrated. One commenter provided a sophisticated statistical analysis that it asserted confirms MedPAC's conclusion that changes are needed to improve payment accuracy. However, this commenter suggested the greatest improvement in cost-margin consistency resulted from switching the basis for the DRG weights from charges to cost and neither the HSRVcc methodology nor the CS DRGs improved payment accuracy. Other commenters indicated that more careful analysis is needed, along with greater access to the details of the CS DRG methodology. The commenters identified the following concerns:

• Validation. The commenters indicated that it is unclear whether there is a need for a new patient classification system. The commenters stated that the implication of moving from a resource-based system to a severity-based payment system must be more fully explored and understood. They indicated that CMS provided no analysis that shows that the proposed changes result in an improved hospital payment system compared to the existing DRG system or APR DRGs.
• Budget neutrality adjustment. The commenter indicated that the proposed rule did not address an adjustment for improved documentation and coding or even a methodology for determining one. The commenter suggested that CMS not apply an adjustment for more comprehensive documentation and coding that increases perceived but not real case mix until there is evidence that one is needed. The commenter requested that CMS monitor actual changes in coding and documentation practices associated with implementation of inpatient payment reforms to determine if any base payment adjustments are needed rather than adjust payments in anticipation of such changes.
• Availability of the GROUPER. Many commenters stated that the proprietary nature and lack of transparency of the proposed CS DRG GROUPER are concerns. The current DRG GROUPER logic has been in the public domain since the inception of IPPS. Without the new GROUPER logic, the commenters believed that it is virtually impossible for anyone to thoroughly analyze the system and comment. The commenters urged that CMS make any new classification system widely available to the public on the same terms as the current DRG system. Some commenters stated that CMS should provide the GROUPER for the CS DRGs and open a new public comment period. Several commenters were concerned about the cost of the GROUPER if the CS DRGs were implemented.
• Too few diagnoses and procedures considered. The commenters are concerned that the current CMS GROUPER does not use all diagnosis and procedures that affect a patient's severity of illness and/or the resources utilized. The commenters believed that the number of secondary diagnoses may be an important factor in determining differences in patient characteristics.

Response: With respect to the comment about the need for a new patient classification system, the proposed rule noted that we modeled the CS DRGs and observed a 12-percent increase in the explanatory power (or R-square statistic) of the DRG system to explain hospital charges. That is, we found more uniformity among hospital total charges within the CS DRGs than we did with Medicare's current DRG system (71 FR 24027). Thus, we believe that there is clear evidence that improvements can be made to the current DRG system that will reduce heterogeneity among patients within a given DRG. While this statistic indicates that the current CMS DRG system can be refined to improve payment accuracy, we agree that it does not necessarily mean we should adopt the system we proposed. As suggested by the commenters, there are a number of other evaluation criteria that we need to consider before deciding whether to adopt the CS DRGs or a potential alternative. We describe these criteria in more detail below. With respect to the comments about a budget neutrality adjustment to account for potential improvements in documentation and coding, we discuss the comments and our responses on this issue more fully in the next section of this final rule. The comment about the availability of the GROUPER is related to a number of detailed comments we received about the potential for Medicare to adopt a proprietary DRG system. We have provided a more detailed description of these comments and our responses below. With respect to the comment about fully utilizing all of the diagnosis and procedure codes submitted on the claim, we note that CMS does not process codes submitted electronically on the 837i electronic format beyond the first 9 diagnosis codes and the first 6 procedure codes. While HIPAA requires CMS to accept up to 25 ICD-9-CM diagnosis and procedure codes on the HIPAA 837i electronic format, it does not require that CMS process that many diagnosis and procedure codes. As suggested by the commenters, there may be value in retaining additional data on patient conditions that would result from expanding Medicare's data system so it can accommodate additional diagnosis and procedure codes. We will consider this issue while we contemplate refinements to our DRG system to better recognize patient severity. However, extensive lead time is required to allow for modifications to our internal and contractors' electronic systems in order to process and store this additional information. We are unable to move forward with this recommendation without carefully evaluating implementation issues. One issue that we expect to consider in deciding whether to adopt such a major systems change is how frequently information beyond the ninth diagnosis code and sixth procedure code affects DRG assignment. Given the cost of an infrastructure change to accommodate this request, we want to be certain that there are sufficient benefits to justify the costs. Again, we will continue to carefully evaluate this request to expand the process capacity of our systems.

Comment: Some commenters stated that the CS DRG grouping methodology based on average charges is inconsistent with the proposed changes to adopt cost relative weights. The commenters recommended using the HSRVcc methodology to determine cost-based weights for consolidating the APR DRGs into CS DRGs.

Response: As explained above, we are not adopting the HSRVcc methodology for FY 2007 because of our concerns about the interaction of charge compression with the hospital-specific portion of the cost weight methodology. Instead, we are setting relative weights based on the estimated cost of the DRGs where cost is determined by applying the national average CCRs to the standardized charges for each DRG in each of the 13 cost centers. In general, when we consider whether to further distinguish types of cases within a DRG in order to create a new DRG or to reassign these cases to a different DRG, we are comparing cases that are clinically similar. Therefore, it is possible or even likely that these cases will be using the same mix of routine and ancillary services and the results of the analysis will be similar whether the cases are compared based on average costs or charges. That is, the cases will be using services that have comparable charge markups over costs and the analysis will produce the same conclusion whether the comparison between cases is based on costs or charges. The major differences between cost and charge weights will occur when comparing across clinically dissimilar services that use a different mix of routine and ancillary services with variable markups. For this reason, we believe that we can continue to do our initial evaluation of potential DRG changes using average charges. Given the complexity associated with developing cost-based weights, we believe our preliminary analysis for evaluating whether to make a DRG change should use charges as a proxy for costs. However, we will consider the commenters' suggestion and, to the extent feasible, consider whether it is possible to evaluate potential DRG changes using costs as well as charges.

Comment: Numerous comments expressed concerns about the use of a proprietary DRG classification system. The commenters indicated that the current DRG GROUPER logic has been in the public domain since the inception of the IPPS. Many commenters noted that the source code, logic and documentation for the current DRG system can be purchased through the National Technical Information Service. The commenters stressed the importance of maintaining transparency within the DRG system (that is, any new DRG system should be available to the public on the same terms as the current one). The commenters stated that any methodology used for the Medicare GROUPER must not be based on a proprietary system. One commenter questioned how future DRG refinements would be made if the underlying system is owned by 3M.

A number of commenters were concerned that it was not possible to thoroughly analyze the proposed CS DRGs and provide comments without the GROUPER logic. Other commenters stated that limited information on the proposed CS DRGs hampered their ability to conduct modeling of the new system. Some commenters raised serious concerns allowing CMS to assign the CS DRG without hospitals having the ability to group the case themselves. According to the commenters, without the CS DRG information, revenue and patient receivables cannot be recorded accurately. The commenters stated that hospitals must have the ability to accurately estimate payments in evaluating strategic initiatives, business plans, budgets, marketing, staffing, and other critical decisions. Commenters noted that CMS provided a link to a web tool on the 3M Web site that allowed hospitals to conduct their own analyses of the impact of moving to CS DRGs. However, these commenters stated that the reality was that if a hospital does not have its own APR DRG GROUPER software, it can only obtain CS DRG information one case at a time by entering specific diagnostic and procedure codes.

Several commenters stated that if CS DRGs are adopted and the GROUPER remains proprietary, they would be limited in their ability to educate and assist hospitals in use of the new system. One commenter indicated that the current 3M product is proprietary and not available in the public domain for hospitals or their software vendors who develop and support their patient account billing and case management software. The commenter also stated that it does not have any access to the underlying codes, conditions and edits utilized by 3M with its product and as a result could not accurately comment on the interaction between severity and complexity associated with individual claims in contrast to resource consumption. The commenter stated that, although hospitals are not required to have a GROUPER, hospitals that hold compliance as a top priority rely on a grouper/encoder to ensure that claims meet all edits prior to submission.

Several commenters stated that a single company's monopoly over the DRG system would be costly to hospitals. The commenters indicated that it would be more difficult to obtain the system to integrate it into hospitals' existing systems. The commenters reported that Maryland hospitals report a GROUPER price of $20,000 per hospital with the ultimate price varying based on criteria such as whether it is used on a mainframe or personal computer. Another commenter expressed concerns that only 3M would be providing access to the GROUPER. The commenter stated that with over 4,000 hospitals requiring a new severity-adjusted DRG GROUPER, it is not feasible or reasonable to expect that one vendor could service all the hospitals nationally in the few months between the posting of the final IPPS rule and an October 1, 2006 implementation. The commenter stated that having 3M maintain control of the GROUPER software limits access by other software vendors to begin reprogramming of the many computer systems that would need to be loaded with the CS DRGs that is currently incompatible with the CMS DRGs. The commenter stated that there will need to be sufficient time between making the GROUPER available and implementation so that hospitals can test their systems, and study the impact on their facilities.

Another commenter stated that it offered software that hospitals and health plans utilize in managing the billing, coding, and payment for inpatient hospital services under the DRGs. The development of software related to Medicare's DRG system by private companies is possible only because the current DRG methodology is available in the public domain. The commenter also noted that the public can obtain full access to the details underlying the CMS DRG system by purchasing information and software from the National Technical Information Service for a nominal charge in a timely manner well in advance of the implementation of changes. The commenter noted the information was available to all of the public simultaneously and no company currently has a competitive advantage in producing DRG products. The commenter added that CMS currently engages in an open and comprehensive discussion about the structure of the DRG methodology through a variety of mechanisms including notices published in the Federal Register. CMS releases sufficient detail about its methodology in electronic formats to enable providers, health plans, and vendors to develop and validate their own computer programs. The commenter expressed concern that unfettered access to the underpinnings of the DRG system would not continue to be available under the CMS proposal to adopt CS DRGs. The commenter suggested the following criteria that a new DRG system should meet in order to be adopted by Medicare:

• Software distribution comparable to what is currently made available, which includes:
• GROUPER source code which produces all pertinent return information;
• All underlying tables that drive the GROUPER with documentation;
• A complete set of test cases to validate the functioning of the software;
• Complete system and user documentation;
• Contact people who can and will respond to questions in a timely fashion;
• The right to redistribute the methodology to business partners and consultants;
• The right to translate source code to other technology environments and to integrate it into other systems;
• Pre-releases of software and documentation well in advance of planned implementations; and
• An open inclusive process for considering future enhancements.

The commenter indicated that the agency must also ensure that whatever refinement methodology is adopted is open to public discussion and scrutiny, now and on an ongoing basis. The commenter stated that transparency is critical to advancing affordability in our health care system.

Response: With respect to making information available for the public to analyze the proposed DRGs, we were cognizant of this issue and attempted to provide as much information as possible that would allow the public an opportunity to comment meaningfully on the proposed CS DRGs. We provided the following data files on the CMS Web site at no cost to the public to assist with understanding our proposed rule:

• Provider Specific File.
• Impact file for IPPS FY 2007 Proposed Rule.
• CCRs and Weighting Factors.
• DRG Relative Weights.
• CS DRG HSRVcc relative weights.
• CAH List for FY 2007 Proposed Rule.

In addition to this information, we made available for purchase both the FY 2004 and FY 2005 MedPAR data that were used in simulating the policies in the IPPS proposed rule. We also discussed the proposed rule in at least two national teleconferences that were open to the public. One of these calls was a Special Hospital Open Door call that was scheduled for 1 and 1/2 hours and was completely devoted to explaining the IPPS proposed rule and answering questions from the public. There were over 1,100 calls into this national teleconference. Finally, we were able to provide access to a Web tool on 3M's Web site that would allow an end user to build case examples using the proposed CS DRGs. While the commenters are correct that these case examples could only be analyzed one at a time, the tool did provide a detailed explanation of how the severity of illness was assigned and the demographic and diagnostic information that went into that determination. Further, other information about the CS DRGs and APR DRGs were available at that Web site, including access to the APR DRG definitions manual.

We acknowledge the many comments suggesting that the logic of Medicare's DRG system should continue to remain in the public domain as it has since the inception of PPS. We also acknowledge the commenters' concern about the impact of moving to a proprietary system and the potential for limiting public access to the underlying GROUPER logic relative to the current CMS DRGs. We note that the issues associated with using a proprietary DRG system were well illustrated in a public comment that we received from the Maryland Health Services Cost Review Commission (HSCRC). Maryland adopted the APR DRGs in June 2004. The commenter noted that “despite the advance notice, a number of hospitals had not acquired the APR DRG GROUPER until near the time for full implementation to begin. In addition to acquiring the GROUPER, hospitals had to deal with issues of integrating the GROUPER with other hospital systems, which was at times difficult with proprietary systems.” The commenter further noted that Maryland has 47 acute care hospitals and “moving the nation's entire hospital industry to a new system in a short period is likely to be much more difficult.” The commenter indicated that “CMS has the opportunity to avoid some of the transition issues the HSCRC faced by placing the CS DRG logic in the public domain or by requiring open licensing of the GROUPER at reasonable rates.” The commenter noted that consultants and vendors to hospitals have struggled to obtain access to the GROUPER as they advised their clients.

The public comments and Maryland's experience with APR DRGs have led to many commenters recommending that Medicare should adopt a new DRG system that is in the public domain. As we evaluate alternative severity classification systems, we will use public access to the system as an important element in evaluating whether each system can be adopted by Medicare. We will continue to strive to promote transparency in our decision making as well as in future payment and classification systems, as we have done in the past.

Comment: A number of commenters suggested that a more straightforward approach to achieving the same or similar objective would be for CMS to refine the current DRG classification system by retaining the current base DRGs (eliminating the current paired DRGs with and without CC) and adding 3-4 levels of severity, rather than using APR DRGs. This option would preserve the many policy decisions that CMS has made over the last 20 years that are already incorporated into the DRG system and yet adjust hospital payments to reflect the cost of care based on patient needs and conditions. Other commenters suggested designating certain DRGs as device-dependent to ensure that device costs are appropriately reflected in the claims file data. Some commenters suggested that CMS retain the current DRG system but revise the CC list as an alternative approach to better recognizing severity of illness in the DRG system.

Several commenters stated that CMS did not conduct an objective study of the CS DRGs although alternatives for the APR DRG system are readily available. These commenters asked whether CMS considered adopting an alternative DRG system that could also better recognize severity.

Two commenters proposed alternative severity of illness systems to the APR DRG system. One commenter suggested that we use the Refined DRG (RDRG) severity of illness system which is supported by Health Systems Consultants, Inc, that contains 1,274 groups with 350 base DRGs. The commenter explained that each of the medical base DRGs is divided into three severity classes and each of the surgical base DRGs is divided into four severity classes. In addition, there are neonate groups based on birth weight, seven DRGs that do not have severity classes and an early death group in each MDC created to remove low outliers according to the commenter. The commenter noted that the research for the RDRG system was undertaken between 1986 and 1989 under a Health Care Financing Administration (now CMS) cooperative agreement. The commenter indicated that the RDRG system has been updated annually using the current CMS complications and comorbidities list since 1989. Solucient, LLC has also used the previous HCFA DRG severity work to develop a risk adjusted DRG system which they refer to as Refined Diagnosis Related Group (R-DRG). Solucient also reports that they have updated their system annually with ICD-9-CM code changes. Another commenter noted that HSS/Ingenix has developed an all-payer severity-adjusted DRG system (APS-DRGs) which contains 1,130 case-mix cells with 376 consolidated DRGs plus 2 error categories. The commenter indicated that, outside of MDC 15, all consolidated DRGs are divided uniformly into three severity levels. The commenter also indicated that the number of severity levels within MDC 15 depends upon the consolidated DRG in the APS-DRG system.

One commenter stated that based on their analysis none of the off-the-shelf Version 23 DRG systems is the best alternative. Rather, it was recommended that a hybrid system be created which would combine the best features of each system. The commenter stated that the proposed CS DRG system or the current CMS DRG system would be the preferred systems to modify. One commenter stated that the use of objective, physiologic data on admission to enhance claims data significantly improves the accuracy of any severity stratification. The commenter suggested that CMS conduct one or more demonstration projects studying claims data enhanced with objective, time-stamped electronically captured laboratory results as an alternative approach for severity adjustment for payment and quality assessment purposes.

Response: The approach suggested in the comments to incorporating severity measures into the current CMS DRG system may be a viable option that we will evaluate in the coming year. With respect to the comment that we undertake demonstration projects to study alternative ways of better recognizing severity in the DRG system, we are concerned that such an endeavor could not be completed in time for FY 2008 implementation. We believe it is very important to make improvements to the DRG system to better recognize severity rapidly and there are a number of different ways that improvements in payment accuracy can be achieved without undertaking a lengthy demonstration project. As suggested by the commenters, much research has already been completed on alternative DRG systems. We believe it is likely that at least one of these systems (or potentially a system that we develop ourselves based on our own prior research) will be suitable to achieve our goal of improvements in payment accuracy by FY2008. We are currently in the process of engaging a research contractor to evaluate the 3M Severity of Illness DRG products along with the other DRG severity systems that have come to our attention during the comment process.

As indicated above, we will use public access to the system as an important element in evaluating whether each system can be adopted by Medicare. With respect to the CS DRGs and potentially the other systems described in the public comments, there may be licensing issues. We proposed to use the CS DRGs beginning in FY 2008. While they were developed under a contract with the Federal government, the CS DRGs are essentially a variant of the APR DRGs that are copyrighted by 3M. The APS-DRGs are a proprietary product owned by HSS/Ingenix, a division of United Health Care. However, HSS/Ingenix has indicated that, should we decide to adopt their product, it would make its DRG system available to the public under the same terms as the current CMS DRGs (that is, the source code, logic and documentation can be purchased through the National Technical Information Service). The RDRG system is supported by Health Systems Consultants.

There are other issues of note with respect to the DRG systems mentioned in the comments and Medicare's efforts to adopt a DRG system that better recognizes severity. In the late 1980's, CMS (then HCFA) funded a Yale University contract for the development of refined severity DRGs. The severity DRGs developed under this contract formed the basis for most of the severity DRG systems available today, including the Ingenix APS-DRGs, the 3M APR DRGs, the Health Systems Consultants RDRGs and the Australian government's AR-DRGs. In the mid-1990's, CMS (then HCFA) also adapted the Yale system and developed a potential severity DRG system, which was described in the Health Care Financing Review. Although the APR DRGs have departed from the Yale approach to a greater extent than have the other systems, both the 3M product and the APS-DRGs were derived from the 1989 Yale severity system that is in the public domain. Given that the Yale system is in the public domain and CMS considered adopting a severity DRG system based on it in the mid 1990's, we will also consider updating our prior work part of our initiative to identify and implement a severity DRG system for use by Medicare in FY 2008. Consistent with the sentiment expressed in the public comments, this option would have the advantage of using the current DRGs as a starting point and retaining the benefit of the many DRG decisions we have made in recent years. The DRG system we considered in the mid-1990's used a base DRG with 3 levels of severity depending upon whether the patient had no CC, a CC, or a major CC. During this past winter, CMS began a comprehensive review of over 13,000 diagnosis codes to determine whether they should be classified as CCs when present as a secondary diagnosis. Under this option, we could continue this review of the CC list, classifying them into one of the three categories described above in conjunction with updating the severity DRG system that we considered in mid-1990's.

c. Changes to CMI From a New DRG System

After the 1983 implementation of the IPPS DRG classification system, CMS observed unanticipated growth in inpatient hospital case-mix (the average relative weight of all inpatient hospital cases) that is used as proxy measurement for severity of illness.

There are three factors that determine changes in a hospital's CMI:

(1) Admitting and treating a more resource intensive patient-mix (due, for example, to technical changes that allow treatment of previously untreatable conditions and/or an aging population);

(2) Providing services (such as higher cost surgical treatments, medical devices, and imaging services) on an inpatient basis that previously were more commonly furnished in an outpatient setting; and

(3) Changes in documentation (more complete medical records) and coding practice (more accurate and complete coding of the information contained in the medical record).

Changes in CMI as a result of improved documentation and coding do not represent real increases in underlying resource demands. For the implementation of the IPPS in 1983, improved documentation and coding were found to be the primary cause in the underprojection of CMI increases, accounting for as much as 2 percent in the annual rate of CMI growth observed post-PPS.

We believe that adoption of CS DRGs would create a risk of increased aggregate levels of payment as a result of increased documentation and coding. MedPAC notes that “refinements in DRG definitions have sometimes led to substantial unwarranted increases in payments to hospitals, reflecting more complete reporting of patients' diagnoses and procedures.” MedPAC further notes that “refinements to the DRG definitions and weights would substantially strengthen providers' incentives to accurately report patients' comorbidities and complications.” To address this issue, MedPAC recommended that the Secretary “project the likely effect of reporting improvements on total payments and make an offsetting adjustment to the national average base payment amounts.”

The Secretary has broad discretion under section 1886(d)(3)(A)(vi) of the Act to adjust the standardized amount so as to eliminate the effect of changes in coding or classification of discharges that do not reflect real changes in case-mix. While we modeled the changes to the DRG system and relative weights for the proposed rule to ensure budget neutrality, we are concerned that the large increase in the number of DRGs will provide opportunities for hospitals to do more accurate documentation and coding of information contained in the medical record. Coding that has no effect on payment under the current DRG system may result in a case being assigned to a higher paid DRG under a system that better recognizes severity. Thus, more accurate and complete documentation and coding may occur under a DRG system that better recognizes severity because it will result in higher payments than the current CMS DRGs. In the FY2007 IPPS proposed rule, we solicited comments on this issue.

Comment: One commenter suggested that CMS should delay implementation of the proposed changes to the DRG system until it conducts nationwide coding and documentation education, particularly to physicians. The commenter also suggested that CMS should find a method to provide physicians who practice in hospitals with web-based documentation training and incentives document correctly.

Response: The proposed CS DRG system is based on the reporting of current ICD-9-CM diagnosis and procedure codes. The proposed changes do not require any changes for hospitals or physicians in how they code or document information in the medical record. For this reason, we do not believe there is a need for any changes to education and training that occurs with respect to documentation and coding.

Comment: Several commenters expressed concern that the proposed rule did not provide any type of analysis to justify or support the need for an adjustment to the IPPS rates for anticipated changes in case mix from a new DRG system. These commenters noted that CMS did not provide a specific adjustment amount in the proposed rule. The commenters stated their view that it is the responsibility of CMS to provide adequate notice and the opportunity for meaningful public comments in response to such a specific proposal before any adjustment can be applied. One commenter recognized that CMS is authorized to make adjustments for changes in coding that are likely to occur. However, absent strong evidence, they urged CMS to avoid making negative adjustments to the standardized amount for anticipated increases in case mix. Another commenter provided two suggestions to CMS. The first suggestion was for CMS to share its thought process on how the standardized amount would be adjusted and allow the public an opportunity to provide comments on this basic set of criteria. The second suggestion was that CMS should make a commitment to adjust future base payment levels if it is determined that the initial adjustment projections are inaccurate. Another commenter stated that any adjustment to the standardized amount in an attempt to account for increased documentation and coding is unnecessary and unwarranted. The commenter asserted that it is virtually impossible to objectively distinguish real changes in case mix from those that occur due to improved coding and documentation. This commenter stated claims are coded using the official coding guidelines that are the same regardless of the DRG system being used. Another commenter requested that CMS not overestimate the growth in CMI as a result of improved coding. This commenter asserted there are many needs for accurate data collection in a hospital setting and coders do not stop reviewing a medical record after locating the first CC that assigns the patient to a higher weighted DRG. The commenter maintained that several hospitals ask coders to assign codes to many of the non-invasive procedures that do not affect DRG assignment. This same commenter also stated they believe the increase in CMI will not be as significant as CMS anticipates.

One commenter representing the State of Maryland shared the state's experience with case mix index changes after adoption of the APR DRG system. The commenter stated correct coding resulting in maximum reimbursement under the CMS DRGs could understate a hospital's case mix under the APR DRGs. Facilities that have tried to improve their coding productivity by seeking to maximize reimbursement under Medicare may not obtain an accurate representation of its patient's severity of illness under APR DRGs. According to the commenter, hospitals have a financial incentive to improve their clinical documentation and to code more completely when APR DRGs (or CS DRGs which are based on APR DRGs) are used for reimbursement.

The commenter also indicated that case mix growth exceeded four percent for the State's hospitals on average, as they began to prepare for the full transition to APR DRGs. Case mix growth in this current fiscal year is about the same. As such, the State has established a policy for FY 2006, limiting the amount of case mix growth experienced for each hospital until the coding patterns become stable. In addition, an appeals process for hospitals with services that generate rising case mix growth due to complexity has also been established.

Response: We appreciate the commenters' concerns and feedback regarding potential adjustments to the national standardized amount to account for improvements in documentation and coding that may cause the case-mix index to increase absent real case-mix growth. The commenters are correct that we did not propose a specific adjustment for improved documentation and coding. As stated in the proposed rule, we were soliciting comments on the possibility of changes in the case mix index as a result of the increase in the number of DRGs within the proposed CS DRGs. We will continue to analyze this issue as we evaluate alternative DRG systems that may better recognize severity of illness for implementation in FY 2008. We acknowledge the commenters' request to provide an opportunity for public comment before CMS adopts a specific adjustment to the standardized amounts for improved documentation and coding. As stated earlier, we intend to propose further changes to better recognize severity in the DRG system for FY 2008. If we decide to make an adjustment to the standardized amount to account for improvements in documentation and coding, we will provide the specific level adjustment and the data and analysis underlying it in a proposed rule that will allow for an opportunity for public comment.

We disagree with the commenters that suggested there is no need for an adjustment to the IPPS standardized amounts to account for improvements in documentation that increase case mix and, therefore, payments. As presented above and in the proposed rule, Medicare's experience since the original inception of the IPPS and long-standing research provide substantiation that improvements in documentation and coding that increase case-mix and payment will occur when the opportunity arises through the expansion of the DRG system. Further, the comment representing the State of Maryland made clear that when CS DRGs “are used for reimbursement, hospitals have the financial incentive to improve their clinical documentation and to code administrative records more completely.” MedPAC also noted that “adopting our recommended refinements to the DRG definitions and weights would substantially strengthen providers' incentives to accurately report patients' comorbidities and complications.”

Comment: One commenter stated that, in its experience, a change to the severity of illness grouping logic will result in an increase to the rate of change in case-mix. Because any effect will not be revenue neutral, the commenter questioned if and how CMS intends to address the change in case-mix, for example, regulating the change or setting a cap for hospitals. The commenter indicated that case-mix could rapidly decline as well as rapidly increase at the hospital-specific level and asked if CMS had a mechanism to address that issue, as well. The commenter also recommended that hospitals with improved case mix due to improved coding accuracy and internal documentation should be entitled to the full CMI benefit.

Response: We appreciate the commenter's concern and agree that the severity of illness grouping logic will affect case-mix. Also, we have known since the development of a PPS for capital payments that changes in case-mix affect capital payments to certain hospitals as much, or more than, operating payments. However, we do not know, at this point, the extent and direction of the impact to case-mix that the severity of illness grouping logic would have, or how rapidly the changes to case-mix would occur. When a decision is made regarding implementing the severity logic, we will be carefully scrutinizing the data and a myriad of variables to ascertain its effect and whether or not adjustments or interventions are necessary.

4. Effect of CS DRGs on the Outlier Threshold

In its March 2005 Report to Congress on Physician-Owned Specialty Hospitals, MedPAC recommended that Congress amend the law to give the Secretary authority to adjust the DRG relative weights to account for the differences in the prevalence of high-cost outlier cases. MedPAC recommended DRG-specific outlier thresholds that would be financed by each DRG rather than through an across-the-board adjustment to the standardized amounts. Furthermore, in comments that MedPAC submitted during the comment period for the FY 2006 IPPS proposed rule, MedPAC stated its belief that the current policy makes DRGs with a high prevalence of outliers profitable for two reasons: 1) These DRGs receive more in outlier payments than the 5.1 percent that is removed from the national standardized amount; and 2) the relative weight calculation results in these DRGs being overvalued because of the high standardized charges of outlier cases. MedPAC also noted that, under its recommendations, outlier thresholds in each DRG would reduce the distortion in the relative weights that comes from including the outlier cases in the calculation of the weight and would correct the differences in profitability that stem from using a uniform outlier offset for all cases. MedPAC added that its recommendation would help make relative profitability more uniform across all DRGs.

In the FY 2006 IPPS final rule (70 FR 47481), we responded to MedPAC's recommendation on outliers by noting that a change in policy to replace the 5.1 percent offset to the standardized amount would require a change in law. However, because the Secretary has broad discretion to consider all factors that change the relative use of hospital resources in the calculation of the DRG relative weights, we stated we would consider changes that would reduce or eliminate the effect of high-cost outliers on the DRG relative weights. At this time, we have not completed a detailed analysis of MedPAC's outlier recommendation because we do not have the authority to adopt such a change under current law. Instead, we have focused our resources on analyzing MedPAC's recommendations with respect to adopting severity DRGs and calculating cost-based HSRV weights that can be adopted without a change in law. While we intend to study MedPAC's recommendation in more detail at a future date, we note that changes to the DRG system that better recognize severity would have important implications for the outlier threshold. In the proposed rule, we analyzed how the outlier threshold would be affected by adopting the CS DRGs.

Using FY 2004 Medicare charge data, 3M Health Information Systems simulated the effect of adopting CS DRGs in conjunction with HSRVcc weights (described) on the FY 2006 outlier threshold using the same estimation parameters used by CMS in the FY 2006 final rule (that is, the charge inflation factor of 14.94 percent) (70 FR 47494). Under these assumptions, 3M Health Information Systems estimated that the outlier threshold would be reduced from $23,600 under the current system to $18,758 under the CS DRGs with HSRVcc weights. By increasing the number of DRGs to better recognize severity, the DRG system itself would provide better recognition for cases that are currently paid as outliers. That is, many cases that are high-cost outlier cases under the current DRG system would be paid using a severity of illness subclass 3 or 4 under the CS DRGs and could potentially be paid as nonoutlier cases.

Comment: Some commenters noted that there was only a limited discussion of the CS DRGs' effect on the outlier threshold and no information about application of the postacute care transfer payment policy. Some commenters inquired how policy areas such as outliers and new technology will be affected by the proposed DRG changes.

Response: We will consider further the application of the postacute care transfer payment policy as we make changes to the DRG system. With respect to outliers, we discussed this issue in the proposed rule. We noted that better recognition of severity in the DRG system will result in some cases that are currently paid as outliers becoming nonoutliers. Under current law, we are required to establish an estimated outlier threshold so that between 5 and 6 percent of estimated IPPS payments are made as outlier payments. Our longstanding policy has been to set the outlier threshold so that estimated outlier payments equal 5.1 percent of estimated IPPS payments. If we were to continue this longstanding policy, we would expect DRG refinements that better recognize severity to lead to a reduction in the outlier threshold. In the proposed rule, using the same data and assumptions used for the FY 2006 final rule, we estimated that adoption of the CS DRGs would reduce the outlier threshold from $23,600 to $18,758.

Comment: One commenter recommended that CMS continue to provide the additional payment for blood clotting factor administered to hemophiliac inpatients in the future even if severity-adjusted DRGs are implemented.

Response: Section 1886(a)(4) of the Act excludes the costs of administering blood clotting factors to inpatients with hemophilia from the definition of “operating costs of inpatient hospital services.” Therefore, under the statute, payment for blood clotting factor provided to hemophiliac inpatients is not included in Medicare's IPPS payment and is paid separately. For this reason, we will continue to apply Medicare's policy of paying separately for blood clotting factor provided to hemophiliac inpatients.

5. Impact of Refinement of DRG System on Payments

In the FY 2007 IPPS proposed rule (71 FR 24020), using the FY 2004 MedPAR claims data, we simulated the payment impacts of moving to the CS DRG GROUPER and the alternative HSRVcc method for developing HSRV weights. These payment simulations did not make any adjustments for changes in coding or case-mix. For purposes of this analysis, estimated payments were held budget neutral to estimated FY 2006 payments because we have a statutory requirement to make any changes to the weights or GROUPER budget neutral. Based on the results of this impact analysis, in the FY 2007 IPPS proposed rule, we proposed to adopt both the HSRVcc weighting methodology for FY 2007 and the CS DRGs for FY 2008. Later in the proposed rule (71 FR 24028) and in the Appendix A—Regulatory Impact Analysis (71 FR 24404), we modeled the effect of only adopting HSRVcc relative weights using the FY 2005 MedPAR claims data applying the traditional statutory budget neutrality requirements.

For reasons described in more detail above, we are adopting cost-based weights in this final rule. However, we are not adopting our proposal to standardize charges on MedPAR claims using HSRVs until we further research issues related to charge compression. Further, as described in more detail above, we are modifying our proposed plan to adopt the CS DRG system for FY 2008. Rather, we will evaluate the CS DRGs along with the other DRG severity systems that have come to our attention during the comment process and consider updating the work we did to develop a severity DRG system in the mid-1990's before adopting a system that better recognizes severity for FY 2008.

In the proposed rule, we presented the impact of the proposed changes on specific high volume DRGs. For comparison purposes, in the following table we are showing the percent changes in weight for these DRGs presented in the proposed rule and the percent changes in weights for these DRGs under the policies we are finalizing in this rule:

We received a number of comments, which we discuss below, expressing concern over the magnitude of the changes we proposed to the relative weight methodology and the effects on the DRG weights. As shown in this table above, the impact of the transitional cost based weights computed without using the HSRVcc method of standardization is significantly less than the impacts projected in the proposed rule. As a further demonstration of the manner in which our final policy mitigates the impacts of the proposed rule, we are presenting the following two tables showing the number of DRGs experiencing percent gains and losses in their relative weights in the proposed and final rules. We also are showing the number of providers experiencing percent gains and losses in case mix due to the proposed and final changes. As shown in the tables, the more extreme percent changes are greatly reduced with our final policies.

For additional comparison purposes between the proposed and final rule relative weights and DRG changes, the following table shows the estimated payment impacts on case mix change by hospital group that we projected for the proposed rule and also shows the estimated payment impacts that we are finalizing in this rule.

We are discussing specific comments and responses relevant to our impact analysis below. The changes that we are adopting in this final rule are illustrated in our regulatory impact analysis.

Comment: Some commenters expressed concern that the proposed rule discusses the impact of moving to CS DRGs using FY 2004 inpatient claims rather than FY 2005 claims to estimate impact. Some commenters stated that using 2 separate years of claims data to show the impact of major changes made it impossible to assess the overall impact of the changes with any reasonable level of confidence.

Response: Because of the long lead time to develop the methodology and our proposed rule, we used the FY 2004 MedPAR data to calculate HSRVcc weights and model the CS DRGs for purposes of the analysis shown on pages 24007-24011, 24020-24026 of the FY 2007 IPPS proposed rule (71 FR24007-24011, 24020-24026). At the time we were developing provisions of the proposed rule, FY 2005 MedPAR data were unavailable to us. Given the public interest in prompt publication of the rule, we decided not to replicate all of the analysis that we provided in section II.C. of the proposed rule based on the FY 2004 data once the new FY 2005 data became available to us. We believed delaying publication of the proposed rule to revise our analysis so all of the payment impacts were shown based on FY 2005 data was not in the public interest. Once we developed the methodology and the analysis for the proposed rule, we calculated the relative weights using the HSRVcc methodology that we were proposing to adopt for FY 2007 using the FY 2005 MedPAR. We modeled the HSRVcc relative weights using the FY 2005 MedPAR because we would be using these data to calculate actual relative weights that would be used to determine FY 2007 hospital payments. We believed it was important to model our FY 2007 proposal as closely to how payments would be determined to provide the most meaningful opportunity for public comment. For purposes of providing the payment impacts shown on pages 24028-24030 and the Appendix A—Regulatory Impact Analysis (71 FR24404) and the methodological description shown on pages 24044-24049 of the proposed rule, we used FY 2005 MedPAR data. We disagree with the commenters that providing separate analyses using 2 years of data makes it more difficult to understand and assess the payment impacts. Rather, we believe that providing these analyses makes it easier to understand how relative weights will change solely as a result of updating the data.

Comment: MedPAC was pleased that CMS proposed three of MedPAC's four recommended changes to the IPPS system. However, the MedPAC expressed concern the proposal not to implement the severity changes until FY 2008. They stated that it is important to correct for differences in patients' severity concurrently with the corrections for charging distortions. MedPAC believed that all of the proposed policy changes to the IPPS should happen concurrently. MedPAC stated that failure to adopt all of the changes would leave some payment distortions in place, thereby continuing to favor some kinds of patients over others. According to MedPAC, adopting all of the policies would create the most accurate payments and prevent hospitals from facing unjustified shifts in their payments that may occur under partial adoption of the payment reforms. MedPAC stated that concerns about giving hospitals time to adapt to the changes may be better managed by implementing all changes in FY 2007 and then giving hospitals a transition period. Another commenter asked that CMS implement both of these proposed changes in FY 2007 for the following reasons:

• MedPAC's analysis revealed significant inaccuracy in the current payment system and recommended implementation of both the new severity-refined DRGs and a revised method for the weights at the same time.
• It is inequitable to remove the subsidy provided by the overpayments for cardiac and orthopedic surgery prior to correcting the underpayments for the most severely ill patients.
• It is not reasonable to ask that some hospitals experience financial losses from implementing the new weights this year if implementing severity would offset some or all of these losses. To stagger implementation will cause providers to experience unnecessary payment fluctuation between FY 2007 and FY 2008.

The commenter further added that a delay is not beneficial to taxpayers as hospitals will have more time to up-code and increase their Medicare payments. Many commenters agreed with MedPAC that the cost weights and severity-adjusted DRGs should be implemented simultaneously. However, these commenters suggested implementation no sooner than FY 2008 to limit sharp fluctuations in payments to hospitals from year to year. Many commenters opposed a two-step implementation, whereby CMS would implement cost-based weights in one year and a new DRG system to better account for patient severity in a subsequent year. They noted that each of these two major reforms significantly redistributes payments, often in off-setting directions. They stated that large swings in payments between the two reforms would create unnecessary volatility and have a profound impact on hospitals' ability to plan effectively, especially for necessary major medical equipment purchases and other capital expenditures. Therefore, they recommended that CMS implement both cost-based weights and severity-adjusted DRGs concurrently. While some commenters urged CMS to implement both payment reforms concurrently in FY 2007, other commenters advised delaying until at least FY 2008 to allow enough time to improve the proposed methodologies and underlying cost data to ensure accuracy of payments. Some commenters stated that the cost-based weights methodology should be implemented after the severity adjusted DRG methodology.

Response: Although we are not adopting the CS DRGs this year, we agree that it is important to smooth the transition for our current DRG system to a more accurate payment system. As indicated above, we have decided to adopt traditional cost-based weights for FY 2007 without the HSRV part of the methodology and we are making refinements that will create 20 new CMS DRGs, modify 32 others across 13 different clinical areas involving 1,666,476 cases that would improve the CMS DRG system's recognition of severity of illness for FY 2007. We believe it is appropriate to take steps toward transitioning the IPPS to a severity based DRG system for FY 2007 by applying some of the severity logic from our proposal to the CMS DRGs where appropriate. By revising the CMS DRGs, we are offering hospitals an interim step toward severity DRGs. Hospitals would be able to take advantage of the improved recognition of severity within the context of the more familiar CMS DRGs. This interim step affords us the opportunity to adopt some of the more basic components of a severity DRG system, such as specific splits in DRGs that lead to groups with greater resource utilization.

Comment: Some commenters were concerned that CMS has not taken into account all of MedPAC's recommendations for reforming the IPPS.

Response: We believe the commenters were expressing concern that we did not analyze MedPAC's recommendation to adjust the relative weights to account for differences in the prevalence of outlier cases. As explained above, we placed most of our attention and resources on the recommendations related to refinement of the current DRGs to more fully account for differences in severity of illness among patients as we do not have the statutory authority to make the specific changes to our outlier policy that MedPAC recommended. While we have not made MedPAC's recommendation regarding outliers a central focus of our analysis, we do intend to examine this issue in more detail over the next year.

Comment: One commenter stated that the annual impact of the changes to the proposed CS DRG system will reduce payments for its institution by an additional $2.7 million per year. The commenter suggested that community, not for profit hospitals be exempt from these proposed changes as this is not the group of hospitals that were the intended target of these changes. One commenter stated that the efforts to address issues identified in the MedPAC report should begin and end with the specialty hospital subset and should not occur in conjunction with payment systems at large for all other hospital facilities.

A few commenters urged CMS to further analyze and evaluate the impact of the proposed HSRVcc methodology on access to Centers of Excellence. They noted that the proposed changes are particularly significant for large volume hospitals and may have a negative impact on the Centers of Excellence. Any negative impact to these Centers could impede beneficiary access to high quality services. Several commenters stated that although CMS' intent may have been to eliminate reimbursement incentives for specialty hospitals to select the most profitable cases, the proposed methodology appears to negatively affect all hospitals serving the most prevalent diagnoses (cardiology, orthopedic joint replacement, and neurosurgery) within the Medicare population. The commenters stated that efforts to address issues identified in the MedPAC report should be limited to specialty hospitals. The payment systems at large that affect all other hospital facilities should not be changed. These commenters suggested that CMS address the reimbursement incentives of specialty hospitals by implementing a separate payment system for specialty hospitals, rather than implement a proposed policy that could negatively impact all hospitals. Several commenters suggested implementing the proposal only for specialty hospitals while deferring the proposed payment reforms for full-service hospitals to afford more time to study the implications of the HSRVcc as a method of general applicability. Another commenter stated that care for Medicare beneficiaries in rural areas will be adversely affected by the proposed adoption of HSRVcc weights because of the dramatic impact on specialized services provided by rural referral centers that are not available at other smaller hospitals in rural communities. The commenter suggested that the future viability of these specialized services may be at risk. Therefore, the commenter recommended that CMS recognize the unique impact of the proposed changes on rural referral centers by excluding these hospitals from the change.

Response: Payments under a prospective payment system are predicated on averages. Therefore, we do not believe it would be appropriate to exclude certain hospital groups from implementation of the changes we are adopting to use cost-based weights or better recognize severity in the DRG system. While these changes are expected to reduce incentives for hospitals to “cherry pick” or treat only the most profitable patients, the objective of these proposed revisions is to improve the accuracy of payments, leading to better incentives for hospital quality and efficiency and ensure that payment rates relate more closely to patient resource needs. Even though few hospitals will have a large increase or decrease in overall Medicare payments, there may be a significant increase or decrease in payment for individual cases within a hospital. Under certain circumstances, the current DRG system benefits hospitals that focus on treating less severely ill patients. Adjusting payment for the severity of the patient will remove the incentives to systematically choose one patient over another. Currently, the DRGs overpay for some types of cases and underpay for others because the relative weight system is based on charges and the DRG system does not sufficiently distinguish more or less resource intensive patients based on severity of illness. The changes we are making to account for costs in the DRG relative weights and improve recognition of severity within the DRG system will significantly increase payment accuracy at both the patient and hospital level.

For these reasons, we believe these changes should apply to all hospitals paid using the IPPS, regardless of whether a hospital is a specialty hospital or a rural referral center. We have made significant changes to our proposal and the impacts shown in this final rule may be very different for an individual hospital than those we showed in the proposed rule. The impact on any specific hospital will depend on the types of cases it treats.

Comment: Several commenters stated that in order to analyze and comment, a crosswalk between the current DRGs and the severity DRGs should be made available.

Response: As indicated earlier, we provided a number of resources during the comment period to assist commenters in analyzing our proposal. We provided a number of data files listed earlier on the CMS Web site at no cost to the public. In addition to this information, we made available for purchase both the FY 2004 and FY 2005 MedPAR data that were used in simulating the policies in the FY 2007 IPPS proposed rule. We also provided access to a Web tool on 3M's Web site that would allow an end user to build case examples using the proposed CS DRGs.

Comment: One commenter stated that the best estimates on a hospital specific basis, of the incremental effects on payment of CMS' changes to the DRG system should be published in the FY 2007 IPPS final rule. The commenter also suggested that CMS release impact files by hospitals far in advance of any implementation.

Response: Information to determine hospital-specific impacts is available on the CMS Web site at: http://www.cms.hhs.gov/AcuteInpatientPPS/FFD/list.asp#TopOfPage. Click on: “Acute Inpatient—Files for Download http://www.cms.hhs.gov/AcuteInpatientPPS/FFD/list.asp.” For the proposed rule impact file, click on “Impact file for IPPS FY 2007 Proposed Rule http://www.cms.hhs.gov/AcuteInpatientPPS/FFD/itemdetail.asp?filterType=none&filterByDID=-99&sortByDID=2>&sortOrder=ascending&itemID=CMS061736.” Similar information for the final rule will also be available on the CMS Web site shortly after the publication of this final rule. We note that some level of familiarity with data concepts and Medicare payment variables will be necessary for hospitals to use these files and simulate a payment analysis for their own facility. Using the latest data available at the time this final rule was prepared, we estimated impacts by category of hospital, and the tables displaying these impacts are published in the impact section of this final rule. Space limitations preclude us from being able to provide hospital-level impacts. In addition, to the extent that adjustments for providers such as the IME adjustment, DSH adjustment, and/or operating and capital CCRs may be updated for FY 2007 subsequent to the publication of this final rule, the actual impacts on individual providers may differ slightly from those we estimated. We believe that by providing the payment variables and other information electronically on the CMS Web site, hospitals have the flexibility to simulate and develop their own impact analyses that may be better suited to their needs than any analysis CMS would do at the hospital level.

Comment: Some commenters stated that CMS needs to extend the comment period to allow hospitals additional time to evaluate the effects of these proposed changes.

Response: One of the reasons that we proposed adopting the CS DRGs for FY 2008 was to give hospitals more than the 60-day public comment period and the additional 60-day delay between the publication of the final rule and implementation on October 1, 2006, to fully understand and plan for the change to the CS DRG system. As indicated earlier, we are not adopting CS DRGs for FY 2007. Therefore, we do not see a need to extend the 60-day public comment period. Although we are not extending the 60-day public comment period, we will involve hospitals and other stakeholders in our plans for moving to a severity DRG system for FY 2008. We are interested in public input on the types of criteria that we should consider and how to evaluate improved payment accuracy as we consider changes to the DRG system to better recognize severity of illness.

Comment: Some commenters encouraged CMS to review the cost/benefit of implementing the cost-based weight methodology and a severity-adjusted DRG system in conjunction with changes to the CMS UB04 claim form and the adoption of ICD-10-CM. The commenters suggested that implementing these changes simultaneously could help alleviate the additional cost of multiple system upgrades both for the hospital and the fiscal intermediaries. Some commenters stated that CMS should conduct a single independent study to determine the impact that implementation of this methodology will have on coding and billing productivity or hospital cash flow. Some commenters stated that implementing the significant DRG changes proposed by CMS is only a temporary solution until a more refined DRG system can be adopted with more specific clinical classification systems such as ICD-10-CM and ICD-10-PCS that will be capable of fully recognizing a patient's severity of illness and the services provided to treat that condition.

Response: We believe that it is important to improve the payment accuracy in the hospital IPPS by implementing these changes when appropriate. The IPPS payment reforms that we have proposed do not require information system changes for hospitals similar to those that will be required for adoption of ICD-10 or a new HIPAA compliant transaction system. The relative weights are merely one component in a payment formula for calculating Medicare's IPPS payment rate. Although there will be increases and decreases in the relative weights that are used in the payment formula for different DRGs, this payment change does not require hospitals to make any computer system changes. Similarly, the changes to adopt a severity DRG system will also not necessarily require hospitals to make any upgrades to their computer systems. The proposed DRG system or any alternative that we consider would use the same ICD-9-CM diagnosis and procedures codes as the current CMS DRGs. Although it seems likely that hospitals will want to acquire the DRG system that Medicare will use, we do not expect that substituting one DRG GROUPER for another should be burdensome and require upgrades to hospital information systems. With regard to the comment that a more refined DRG system can only be adopted with more specific classification systems such as ICD-10-CM and ICD-10-PCS, the Secretary is evaluating whether we should adopt ICD-10.

Comment: One commenter supported the decision to use the CS DRGs, noting that use of a 3-digit DRG number would avoid the undue health programming costs that move limited financial resources away from initiatives focused on improving quality care and access to health care. However, the commenter also indicated that the number of digits in the DRG number should not be a factor in choosing the best severity classification system.

Response: We appreciate the commenter's support for our proposal as well as the comment that the DRG classification system used by Medicare should not be dependent upon the number of digits in the DRG number. We will consider any information system infrastructure issues as we evaluate alternative DRG systems.

Comment: Several commenters stated that the reasons CMS gave in the proposed rule for not implementing CS DRGs for FY 2007 are valid. The commenters stated that they are all the more valid because hospitals now would have less time to prepare if CMS were to implement its proposed severity adjusted DRGs this October 1.

Response: We agree. The proposed change to adopt CS DRGs represents a major change to how hospitals are paid for Medicare inpatient services. We will not be implementing the CS DRGs for FY 2007. However, we do plan to evaluate potential alternative DRG systems that better recognize severity than the current CMS DRGs for FY 2008.

Comment: One commenter suggested that the CS DRG system's reliance on 3M's proprietary APR DRG grouping logic and software may not be in compliance with Pub. L. 104-113, the National Technology Transfer and Advancement Act of 1995. The commenter recommended that we participate in the formation of expert committees with a proven consensus standards body to develop a standardized DRG classification and severity-adjustment system for the IPPS.

Response: We appreciate the commenter's support for the use of a consensus standards body to develop a severity-adjusted DRG system. The National Technology Transfer and Advancement Act of 1995 directs Federal agencies to use voluntary consensus standards in lieu of government-unique standards, except where inconsistent with law or otherwise impractical. As we move toward implementing a severity-adjusted DRG system, we will carefully consider whether it would be appropriate to involve a voluntary consensus standards body in the process.

Comment: Some commenters stated a transition (blended) period with stop loss protections should be provided over a period of one to three years. Other commenters suggested a longer transition period given the magnitude of payment distribution across DRGs and hospitals. The commenters believe that the transition approach would be consistent with many other major changes that have been implemented gradually over the years, including the capital prospective payment system. The commenters suggested that a minimum of 1 year should be allowed for the development of software systems to handle these changes.

Response: We agree that these changes should be implemented over a transitional period. As we indicated earlier, we are revising the current DRG system to better recognize severity (which is discussed in detail in section II.C.7. of the preamble of this final rule) and are also adopting cost-based weights for FY 2007. We are providing for a transition period of 3 years with the relative weights becoming an increasing blend of costs weights as the transition proceeds. We also believe that the 20 new DRGs we are adopting for 2007 will improve the transition from our current system to a more sophisticated severity DRG system in FY 2008.

Comment: One commenter noted that MedPAC recommended excluding statistical and high cost outliers from the computation of the DRG weights in order that the weights reflect the average cost of the inlier case only. MedPAC further recommended shifting the financing of the outlier pool from all cases to cases in the DRGs with the highest prevalence of outliers. The commenter noted that outlier cases occur most frequently in high-weighted DRGs. Therefore, MedPAC's proposal of accounting for the high prevalence of outliers in the DRGs would compound the weight compression caused by the HSRV methodology. The commenter believed that each proposal by MedPAC (to exclude statistical and high cost outliers from the computation of the DRGs) would exacerbate payment inaccuracies, and the two proposals combined would be deleterious. The commenter stated that it would further analyze MedPAC's proposal to test their theory empirically.

Another commenter was also concerned about MedPAC's recommendation to adjust the DRGs to account for the prevalence of high-cost cases. The commenter explained that reducing the relative weights to finance the outlier pool will adversely affect payment for hospitals specializing in the most complex patients. Hospitals may be discouraged from developing the capacity to treat high cost outliers and responding to the needs of their community according to the commenter. Meanwhile, the commenter suggested that hospitals that have the capacity to treat the highest cost and most complex cases may abandon such an infrastructure because it will be too costly to maintain.

One commenter supported MedPAC's proposal and believed that implementing MedPAC's proposal would support the goal of achieving payment accuracy. The commenter explained that the current system provides double reward for DRGs with a high prevalence of outliers. The commenter recommended that CMS seek legislative authority to implement MedPAC's proposal of DRG specific outlier thresholds.

Another commenter was supportive of MedPAC's recommendation and noted that MedPAC stated in a letter to CMS that “failure to adopt any of (MedPAC's) recommendations would leave some payment distortions in place, thereby continuing to favor some patients over others.” Therefore, the commenter recommended that CMS implement all of MedPAC's recommendations simultaneously when Congress has granted CMS authority to adopt MedPAC's outlier recommendation.

One commenter was concerned that CMS provided only “minimal” analysis of the effect of the DRG refinements on the outlier threshold. Noting that the 5.1 percent set aside for outlier payments could be significantly reduced with the adoption of severity DRG refinements, the commenter believed that implementation of severity DRGs is premature until the Secretary determines whether statutory changes are needed to determine the percentage of total IPPS payments that should be made as outliers.

One commenter recommended that, even though CMS does not have the authority to change the outlier policy, it should review creating DRG-specific or day outliers under a severity DRG system. Another commenter recommended that CMS reduce payments for outliers and eventually eliminate them upon implementing severity DRGs.

Response: We thank the commenters for taking the time to comment on MedPAC's recommendation. As noted above, we do not have the statutory authority to implement MedPAC's recommendation, and, therefore, we placed most of our attention and resources on the recommendations related to refinement of the current DRGs to more fully capture differences in severity of illness among patients. However, we intend to examine MedPAC's recommendation regarding outliers in more detail in the future and will consider the comments we received on the FY 2007 IPPS proposed rule.

6. Conclusions

As we describe in more detail below, we believe that adopting cost-based weights and making improvements to the DRG system to better recognize severity has the potential to result in significant improvements to Medicare's IPPS payments. This final rule implements a cost weight methodology effective for FY 2007. Further, we are creating 20 new CMS DRGs and modifying 32 others across 13 different clinical areas involving 1,666,476 cases that would improve the CMS DRG system's recognition of severity of illness for FY 2007. Further, as suggested by MedPAC and others, we are adopting these changes over a transition period while we plan further improvements to the IPPS for FY 2008.

In developing our proposed and final policies, we considered a range of alternatives outlined below, and we solicited comments on both the proposal and the alternatives. We asked commenters to consider both the CS DRGs and alternative severity adjustment methods for accounting for severity more comprehensively in the DRG payment system. For example, under one alternative in the proposed rule, we would implement the CS DRGs in FY 2007 along with the HSRVcc weighting methodology. In this event, as discussed above, to maintain budget neutrality, we would also implement in FY 2007 an adjustment to the standardized amounts to eliminate the effect of changes in coding or classification of discharges that do not reflect real changes in case-mix. Although we did receive comments in support of this idea, many commenters requested that we not adopt the CS DRGs and the HSRVcc weights for FY 2007. Many of these commenters suggested delaying implementation of both proposals until at least FY 2008. Under another alternative, we would have adopted and implemented CS DRGs in FY 2008. Although we did receive comments in support of this idea, we also received many comments raising important concerns about licensing and proprietary issues potentially associated with use of the CS DRGs. The commenters asked us not to adopt the CS DRGs unless we could make them available on the same terms as the current CMS DRGs. Yet other commenters objected to our proposed implementation of the CS DRGs unless we evaluated alternatives and better justified why there is a need to adopt a revised DRG system. Under yet another alternative, we would consider partially implementing the CS DRGs in FY 2007 and complete implementation in FY 2008. However, we noted that there were practical difficulties associated with partial implementation of CS DRGs because cases in a single DRG under the current CMS DRG system may group to multiple DRGs and MDCs under the CS DRG system. Conversely, cases that group to multiple MDCs and DRGs under the current system may group to a single MDC and DRG under the current CS DRG system. We did not receive any comments supporting the idea of partial adoption of the CS DRGs.

In the FY 2007 IPPS proposed rule, we discussed in some detail an alternative to partially adopting CS DRGs that would apply a clinical severity concept to an expanded set of DRGs in FY 2007. For example, we have received correspondence that raised the concern that hospitals may have incentives under the current DRG system to avoid severely ill, resource-intensive back and spine surgical cases (as discussed in section II.D.3.b. of the proposed rule; the correspondence specifically requested that we apply a clinical severity concept to DRG 546). In the proposed rule, we noted that other surgical DRGs may not accurately recognize case severity. Because of the frequency of DRG use and the potential for risk selection, we pointed out that certain DRGs may be particularly important in creating a financial incentive for hospitals to select a less severely ill patient whose case would be assigned to the same DRG as a more severely ill patient.

Therefore, while we proposed to adopt the CS DRGs in FY 2008, we were considering whether to make more limited changes to the current DRG system to better recognize severity of illness in FY 2007. In the FY 2006 IPPS final rule (70 FR 47474 through 47478), we took steps to better recognize severity of illness among cardiovascular patients. For all DRGs except cardiac DRGs, we currently distinguish between more and less complex cases based on the presence or absence of a CC. However, the diagnoses that we designate as CCs are the same across all base DRGs. Because the CC list is not dependent on the patient's underlying condition, CCs may not accurately recognize severity in a given case. The changes we made in FY 2006 to the cardiac DRGs significantly improved recognition of severity between patients by distinguishing between more and less severe cases based on the presence or absence of a MCV. In the proposed rule, we indicated that we were considering whether a similar approach applied to other DRGs would improve payment.

Much like the approach we took last year to identify MCV conditions that represented higher severity in cardiovascular patients, in the proposed rule, we indicated that we planned to examine which conditions identified more severely ill cases in selected MDCs and DRGs. We solicited comments as to whether it would be appropriate to adopt these types of limited changes in FY 2007 as an intermediate step to adopting CS DRGs in FY 2008. There were a number of comments that suggested we should make improvements to our current DRG system rather than adopting the CS DRGs. A number of comments expressed support for using the current DRG system as the starting point for revising the DRG system to better recognize severity to avoid losing the many positive changes that have been made over the years to the CMS DRGs. We also encouraged commenters to send us suggestions regarding potential changes that could be made to the current DRG system to better recognize severity of illness. As indicated below, some commenters did provide us with specific suggestions for how we could revise the current DRGs.

In the FY 2007 IPPS proposed rule, we also discussed an additional alternative under which we would implement the CS DRGs in FY 2007 and the HSRVcc methodology in FY 2008. We did receive one comment supporting this idea. However, as we have discussed elsewhere, we believe that we should not adopt CS DRGs in FY 2007, but rather evaluate severity DRG systems for adoption in FY 2008.

With respect to the relative weight calculations, we believe that adopting HSRVcc weights has the potential to significantly improve payment equity between DRGs. As MedPAC notes, a “survey of hospitals' charging practices suggests that hospitals use diverse strategies for setting service charges and raising them over time.” MedPAC found that data from the Medicare cost reports indicate that hospital markups for ancillary services (for example, operating room, radiology, and laboratory) are generally higher than for routine services (for example, intensive care unit and room and board). Thus, MedPAC has concluded that the relative weights for DRGs that use more ancillary services may be too high compared to other DRGs where the routine costs account for a higher proportion of hospital costs. Although we agree with MedPAC's conclusion, the public comments raised important issues about the effect of charge compression on the relative weights using the HSRVcc methodology. These commenters argued that the HSRV calculation exacerbates the effect of charge compression or the practice of hospitals applying higher percentage markups on lower cost items and lower percentage markups on higher cost items. As we indicated above, we have engaged a contractor to assist us with studying whether charge compression is an actual phenomenon and how it affects the HSRV methodology. As part of this analysis, we will study an adjustment for charge compression suggested in the public comments and will consider adopting HSRV weights in the future. Nevertheless, in the interim, we believe it is important to adopt a methodology for calculation of DRG relative weights that takes costs into account. We have revised the CCRs that we used to develop cost-based weights based on the public comments. Although they do not show the same differentials indicated in the proposed rule, they continue to support MedPAC's conclusion that a system based on charges pays too much for some types of cases and pays too little for others. As indicated above, we summarized hospital-level cost and charge information to 2 routine and 11 ancillary departmental cost centers and found that national average routine cost center CCRs ranged from 50 percent (intensive care unit days) to 56 percent (routine days), while ancillary cost center CCRs ranged from 16 percent (anesthesiology) to 46 percent (labor and delivery room).

MedPAC also found that relative profitability ratios were higher among cardiovascular surgical DRGs than the medical DRGs. We believe the relative profitability of the surgical cardiovascular DRGs has been an important factor in the development of specialty heart hospitals. Our payment impact analysis indicates that this issue will be addressed by adopting cost-based weights. Moving from the current system of charge-based weights to cost-based weights increases payment in the medical DRGs relative to the surgical DRGs. We expected this result, given that routine costs will generally account for a higher proportion of total costs in the medical DRGs than in the surgical DRGs. In the proposed rule, we estimated that all of our combined changes would, on average, increase the medical DRG weights by approximately 7.3 percent while reducing the surgical DRG weights by approximately 6.9 percent. Implementing the cost-based weights without utilizing the HSRV standardization method under the 3-year transition period where the weights for FY 2007 will be based on 33 percent of the cost-based weight and 67 percent of the charge weight will lessen the effects of redistribution between medical and surgical DRGs. In this final rule, we estimate that the increase in the average medical DRG weight will be 0.9 percent and that the decrease in the average surgical DRG weight will be 1.2 percent. The pattern of increasing medical weights and decreasing surgical weights still holds true. However, by adopting the cost based weights in a transition period, we are mitigating the larger swings in payments that our proposed policies adopted in full would have caused.

Although adopting HSRVcc weights would result in the most significant improvement in hospital payment-to-cost ratios among the changes to the IPPS recommended by MedPAC, we have concerns about implementing this methodology until we can further study whether the relative weights might be affected by charge compression. For this reason, we are adopting cost-based weights without HSRV for FY 2007. However, we will consider applying the HSRV methodology in subsequent years if our analysis of charge compression suggests the issue is not a concern or, if appropriate, we can apply an adjustment that would account for its effects.

Based on our analysis, we concur with MedPAC that the CS DRGs would account more completely for differences in severity of illness and associated costs among hospitals. MedPAC observed some modest improvements in hospitals' payment-to-cost ratios from adopting APR DRGs. We modeled the CS DRGs discussed above and observed a 12-percent increase in the explanatory power (or R-square statistic) of the DRG system to explain total hospital charges. That is, we found more uniformity among hospital total charges within the CS DRG system than we did with Medicare's current DRG system. While we believe the CS DRG system that we described above has the potential to improve the IPPS, we have the following concerns about adopting it for FY 2007:

• Further adjustments are needed to the proposed DRG system. In the proposed rule, we indicated that further adjustments need to be made to the proposed CS DRGs to account for situations where less severely ill patients may be more resource-intensive because they need expensive medical technology. The CS DRGs assign a patient to a DRG based on severity of illness but do not recognize increased complexity due to the types of services/technology provided. In addition, the CS DRGs do not incorporate many of the changes to the base DRG assignments that have been made over the years to the CMS DRGs. There was significant interest in the public comments in either revising the CS DRGs to reflect these changes or use the CMS DRGs at the starting point to better recognize severity. The public comments provided a number of examples where we need to consider whether further changes are needed to the CS DRGs before they are ready for implementation.
• Use of a proprietary DRG system. The commenters raised valid points about adopting a proprietary DRG system, including concerns about the availability, price and transparency of logic of the APR DRGs that are currently in use in Maryland. The CS DRGs are a variant of the APR DRG system. As we evaluate alternative severity classification systems, we will use public access to the system as an important element in evaluating whether each system can be adopted for Medicare. We will continue to strive to promote transparency in our decisionmaking as well as in future payment and classification systems as we have done in the past.
• No alternatives have been evaluated. We have not evaluated alternative DRG systems that could also better recognize severity. We have received comments suggesting that alternative DRG systems can better recognize severity than the CS DRGs. It appears that all of the DRG systems that were raised in the public comments as potential alternatives to the CS DRGs are proprietary systems. However, it is possible that we could use one of these systems if it were made available in the public domain on the same terms as the current CMS DRGs. Further, as discussed above, CMS (then HCFA) did work on developing a severity DRG system in the mid-1990's. It is possible that we could update this work and adopt a system that better recognizes severity based on the current CMS DRGs for FY 2008 that does not raise the licensing issues that are involved with using prioprietary systems.

Therefore, for the reasons indicated above, we are not adopting the CS DRGs for FY 2007. However, we are creating 20 new CMS DRGs and modifying 32 others across 13 different clinical areas involving 1,666,476 cases that would improve the CMS DRG system's recognition of severity of illness for FY 2007. Furthermore, as discussed earlier, we have engaged a contractor to assist us with evaluating alternative DRG systems that were raised as potential alternatives to the 3M Severity of Illness DRG products in the public comments. Finally, we will consider the review that we have undertaken of the 13,000 codes on the CC list as part of making further refinements to the current CMS DRGs to better recognize severity of illness based on the work that CMS (then HCFA) did in the mid-1990's to adopt severity DRGs. Again, we expect to complete this work in time for proposing changes to the DRG system to better recognize severity of illness by FY 2008.

7. Severity Refinements to CMS DRGs

In response to the FY 2007 IPPS proposed rule, we received a number of public comments that supported the refinement of the current CMS DRGs so that they better capture severity. Several commenters supported the expanded use of a clinical severity concept similar to the approach used in FY 2006 to refine the cardiac DRGs. One commenter urged CMS to expand the set of DRGs to which this clinical severity concept would apply, including the DRGs that capture the implanting of defibrillators. Another commenter expressed support for additional modifications to the current DRGs to better capture severity and complexity of patients. Another commenter recommended that CMS start with the current DRG system and provide overlays for severity, complexity and patient benefit. One commenter suggested that CMS develop severity levels within all of the existing DRGs (or pairs of DRGs, in the cases where CC or MCV splits now exist), or identify specific DRGs that may be most appropriate for severity adjustments. Several commenters recommended specific adjustments to better capture severity for septicemia, headache, and mechanical ventilation patients. (The DRG recommendations are discussed below under the specific DRG topic.)

We recognize the importance of having a classification system that recognizes cases that utilize greater resources and have higher levels of severity of illness. While we discussed moving to a new DRG system such as the CS DRGs for FY 2007, we stated that we were also interested in improving the current DRGs so that they better capture patients with greater severity of illness as early as FY 2007. We solicited comments in the proposed rule on whether it would be appropriate in FY 2007 to apply a clinical severity concept to an expanded set of DRGs, similar to the approach we used in FY 2006 to refine cardiac DRGs based on the presence or absence of an MCV.

We believe it is appropriate to move in a direction toward a DRG system that better recognizes severity. Our strategy involves following recommendations received as part of public comments and implementing some of the severity logic in the proposed CS DRGs in the CMS DRGs where appropriate. By doing so, we would be taking an interim step toward better recognizing severity in the DRG system. Hospitals would be able to take advantage of a portion of improved severity logic in the proposed CS DRGs within the context of the more familiar CMS DRGs. This interim step would also afford hospitals a more detailed understanding of some of the basic types of DRG logic used in the proposed CS DRG system. Obviously, we were not able to adopt some of the more sophisticated logic involved in the 18 steps included in the proposed CS DRG system. However, we were able to adopt some of the more basic components such as specific splits in DRGs that lead to groups with greater resource utilization.

We began our process of adopting some of the severity logic within the proposed CS DRGs by first comparing the current CMS DRGs to the base DRGs in the proposed CS DRGs to identify areas where improvements could be made to better account for severity of illness and resource utilization. We used two general approaches to evaluate potential DRG changes. First, we analyzed where the assignment of a case to a DRG differed under the CMS DRGs and the proposed base CS DRGs. Second, we analyzed whether there was a list of “major conditions” that could be used to revise any DRGs to better recognize severity, similar to the changes to the cardiovascular DRGs involving MCVs we established in last year's final rule. We used the diagnoses listed as “major” or “extreme” under the proposed CS DRGs for this review. The changes described below will result in better recognition of severity in the current DRG system and, like the changes we made last year to reform the cardiovascular DRGs based on MCVs, represent an excellent next step in refining the Medicare inpatient hospital payment system so our payments are better targeted to specific patients based on their costs of care.

We began our review by focusing on the cardiac and orthopedic DRGs because of our concerns that cardiac, orthopedic, and surgical hospitals have taken advantage of opportunities in the DRG system to specialize in the least complex and most profitable inpatient cases. However, with respect to orthopedic and surgical specialty hospitals, we considered that they have very small inpatient volume and the issues that are leading to their creation are generally unrelated to profit opportunities in the IPPS. Although we did review the orthopedic DRGs, we generally did not find opportunities within the current DRG system to make further refinements for severity of illness. We were also unable to find a strong basis to subdivide further most of the cardiovascular DRGs. In last year's IPPS rule, we already made significant changes to the DRG system to better account for severity of illness in the DRGs frequently performed by cardiac hospitals. As mentioned earlier, this DRG change involved splitting some cardiac DRGs based on the presence or absence of an MCV. We then conducted a comparison of the base DRGs in the CMS DRG system and proposed CS DRGs. We analyzed data to identify specific CMS DRGs with wide ranges in charges that had been subdivided or in other ways modified under the proposed CS DRGs. As stated earlier, this process did not allow CMS to use the more sophisticated logic involved in the proposed CS DRGs to differentiate groups with greater severity. However, we were able to identify a group of DRGs that could be created to better align our payments based on severity of illness. We used our own analysis along with specific recommendations received during the comment period to develop further severity refinements to the current DRGs.

We identified 20 new CMS DRGs involving 13 different clinical areas that would improve the CMS DRG system's recognition of severity of illness. Twelve of the new DRGs are medical and 8 are surgical. The 20 new DRGs are constructed through a combination of approaches used in the proposed CS DRGs to refine the base DRGs such as:

• Subdividing existing DRGs through the use of diagnosis codes.
• Subdividing DRGs based on specific surgical procedures.
• Selecting cases with specific diagnosis and/or procedure codes and assigning them to a new DRG which better accounts for their resource use and severity.

We also modified 32 DRGs to better capture differences in severity. The new and revised DRGs were selected from 40 current DRGs which contain 1,666,476 cases and represent a number of body systems. In creating these 20 new DRGs, we are deleting 8 existing DRGs and modifying 32 existing DRGs. The specific DRG changes are described below:

a. MDC 1 (Diseases and Disorders of the Nervous System)

(1) Nervous System Infection Except Viral Meningitis

Under our current DRG system, all nervous system infections except viral meningitis are assigned to CMS DRG 20 (Nervous System Infection Except Viral Meningitis). By combining all nervous system infections except viral meningitis into one DRG, we are grouping together patients with wide ranges of severity. Under our proposed CS DRGs, there are separate DRGs that distinguish bacterial infection and tuberculosis from other infections of the nervous system. The CS DRGs divided these cases in order to better recognize severity. The codes which describe bacterial infection and tuberculosis are listed below.

We then divided the cases within CMS DRG 20 based on the presence or absence of bacterial infections and tuberculosis of the nervous system. Our medical advisors support dividing these cases in this manner to better recognize severity of illness. The data indicated that these are two distinctly different groups with significant differences in severity. The bacterial and tuberculosis infection group had average charges of $47,034 compared to the $36,507 average charges for cases with other types of infection of the nervous system. Clearly these charge data support the fact that the bacterial and tuberculous infection group has a significantly greater degree of severity. The chart below illustrates these data:

The data support the creation of two separate DRGs for these two groups of patients. Therefore, we are deleting DRG 20 and creating the following two new DRGs:

• DRG 560 (Bacterial & Tuberculosis Infections of Nervous System).
• DRG 561 (Non-Bacterial Infections of Nervous System Except Viral Meningitis).

The ICD-9-CM diagnosis codes assigned to each new DRG are as follows.

The new DRG 560 will have principal diagnosis codes listed in the following table.

The new DRG 561 will have principal diagnosis codes listed in the following table.

(2) Seizure and Headache

Comment: One commenter stated that the current DRGs do not adequately capture the severity of patients with more severe types of headaches. The commenter further noted that seizures and headaches represent distinctly different levels of severity, yet they are grouped together in the CMS DRGs:

• CMS DRG 24 (Seizure & Headache Age >17 with CC).
• CMS DRG 25 (Seizure & Headache Age >17 without CC).
• CMS DRG 26 (Seizure & Headache Age 0-17).

The commenter stated that more severely ill patients, such as those with intense migraine headaches, should be differentiated from other patients in the DRG. The commenter suggested splitting these DRGs into two or more new DRGs to better capture severity. Alternatively, the commenter suggested that CMS examine how the APR DRG system handles these types of cases.

Response: Under both the APR DRGs and our proposed CS DRGs, seizure and headache cases are assigned to separate DRGs while these cases are grouped together in the CMS DRGs. Both severity DRG systems recognize different levels of severity for these two groups of patients. Our medical advisors found that seizure and headache patients are clinically different, with seizure patients having a higher level of severity. We also analyzed data for patients with seizures versus those who are admitted with headaches and found that seizure cases have higher average charges than headaches. We did not have enough cases to analyze potential DRG changes for DRG 26. As the chart below shows, seizure patients age greater than 17 have average charges of $17,125 with CC and $10,540 without CC. Headache patients greater than 17 years of age have average charges of $11,618. The data did not support creating a split for headache patients greater than 17 years with and without CC. The difference in average charges for these groups was only $2,596 ($12,591 with CC as compared to $9,995 for those without a CC).

The data also support creating separate DRGs for seizure and headache patients greater than 17 years of age. The data further support an additional split for seizure patients based on the presence of a complication or comorbidity (CC). Seizure cases with a CC have $6,585 greater average charges compared to cases without a CC. The data are less compelling for creating a split based on the presence of a CC for headache cases, since the difference in average charges is only $2,596.

The clinical data and our medical advisors support the creation of separate DRGs for these two groups of patients. Therefore, we are deleting the following DRGs:

• DRG 24 (Seizure & Headache Age >17 with CC).
• DRG 25 (Seizure & Headache Age >17 without CC).

We are creating the following three new DRGs:

• DRG 562 (Seizure Age >17 with CC).
• DRG 563 (Seizure Age >17 without CC).
• DRG 564 (Headaches Age >17).

The ICD-9-CM codes and DRG logic for cases assigned to these new DRGs will be as follows.

New DRG 562 will have the following principal diagnosis codes and age greater than 17 years with a CC.

New DRG 563 will have the principal diagnosis codes listed above for DRG 562, age greater than 17 years, but no complication/comorbidity.

New DRG 564 will have the principal diagnosis codes listed as follows and an age greater than 17 years.

b. MDC 4 (Diseases and Disorders of the Respiratory System): Respiratory System Diagnosis With Ventilator Support

Medical patients who are treated with mechanical ventilation for respiratory failure are currently assigned to DRG 475 (Respiratory System Diagnosis with Ventilator Support). This DRG includes patients who are on a mechanical ventilator for only a few hours as well as patients who are on mechanical ventilation for several days. The proposed CS DRGs divide these patients into two groups, those on ventilator support for 96 or more hours and those on ventilator support for less than 96 hours. The CS DRGs recognize the difference in severity between these two groups of patients. Our medical advisors agree that medical patients who are treated with mechanical ventilation for respiratory failure for 96 or more hours in most cases are more severely ill than patients who are treated with mechanical ventilation for fewer than 96 hours. A review of these cases illustrates a significant difference in average charges for patients on ventilator support for 96 or more hours which supports the greater severity of these patients. The chart below shows that patients on ventilator support for 96 or more hours have average charges of $83,058 compared to $38,300 for patients on ventilator support for less than 96 hours, a difference of $44,758 in charges. The following chart summarizes these data.

The proposed CS DRGs do a much better job of identifying patients on ventilator support who have higher levels of severity and utilize significantly more resources. Therefore, we will adopt the approach used under the CS DRG system and split these patients based on whether or not the patients are on mechanical ventilation for 96 hours. We are deleting DRG 475 and creating the following two new DRGs:

• DRG 565 (Respiratory System Diagnosis with Ventilator Support 96+ Hours).
• DRG 566 (Respiratory System Diagnosis with Ventilator Support < 96 Hours).

The DRG logic for these two new DRGs is as follows.

New DRG 565 will have a respiratory system diagnosis and procedure code 96.72 (Continuous mechanical ventilation for 96 consecutive hours or more).

New DRG 566 will have a respiratory system diagnosis and the following procedure codes:

96.70 (Continuous mechanical ventilation of unspecified duration).

96.71 (Continuous mechanical ventilation for less than 96 consecutive hours).

c. MDC 6 (Diseases and Disorders of the Digestive System)

(1) Major Esophageal Disorders and Major Gastrointestinal and Peritoneal Infections

The proposed CS DRGs assign major esophageal disorders to a single DRG because these disorders have been shown to have a higher level of severity than do other types of esophageal disorders. Under the current CMS DRGs these disorders are dispersed throughout 8 separate DRGs. The conditions included in the list of major esophageal disorders are described in the table below. The proposed CS DRGs also assign specific gastrointestinal and peritoneal infections that represent a high level of severity into a single DRG. These conditions are assigned to the same group of eight CMS DRGs mentioned above within CMS' current DRGS. The conditions considered gastrointestinal and peritoneal infections are described in the table below.

Our data show that the two groups of cases assigned to major esophageal disorders and to the gastrointestinal and peritoneal infections represent significantly greater severity levels and have higher average charges than do other cases in the eight CMS DRGs. The eight current CMS DRGs to which these two groups of higher severity cases as assigned are as follows:

• CMS DRG 174 (G.I. Hemorrhage with CC).
• CMS DRG 175 (G.I. Hemorrhage without CC).
• CMS DRG 182 (Esophagitis, Gastroenteritis & Miscellaneous Digestive Disorders Age >17 with CC).
• CMS DRG 183 (Esophagitis, Gastroenteritis & Miscellaneous Digestive Disorders Age >17 without CC).
• CMS DRG 184 (Esophagitis, Gastroenteritis & Miscellaneous Digestive Disorders Age 0-17).
• CMS DRG 188 (Digestive System Diagnoses Age >17 with CC).
• CMS DRG 189 (Digestive System Diagnoses Age >17 without CC).
• CMS DRG 190 (Digestive System Diagnoses Age 0-17).

As can be seen from the tables above, cases assigned to these eight DRGs without a major esophageal disorder or a major gastrointestinal disorder and peritoneal infection have average charges ranging from $9,845 to $18,194. The average charges for major esophageal disorders are $18,410, while average charges for major gastrointestinal disorders and peritoneal infections are $20,861. Removing these higher severity cases from the eight DRGs does not have a significant impact on the DRG weights for the remaining cases. Most of the higher severity cases are being removed from DRG 182. There were 282,619 cases in this DRG. By removing the two new groups of cases, the DRG has 243,563 cases remaining. The average charge for DRG 182 with the remaining cases decreases from $14,269 to $13,124. Therefore, the impact on the remaining cases is not that significant. However, reassigning cases with major esophageal and gastrointestinal disorders and peritoneal infections to two new DRGs has the effect of creating two groups which have higher levels of severity and use significantly greater resources. Our medical advisors agree that these two groups represent higher levels of severity and that it is appropriate to move these two groups of cases out of their existing assignments and into the following two new DRGs:

• DRG 571 (Major Esophageal Disorders)
• DRG 572 (Major Gastrointestinal Disorders and Peritoneal Infections)

We are creating new DRG 571 with the following ICD-9-CM diagnosis codes (removing them from DRGs 174, 175, 182, 183, 184, 188, 189, and 190):

We are creating new DRG 572 with the following ICD-9-CM diagnosis codes (removing them from DRGs 182, 183, 184, 188, 189, and 190):

(2) Principal or Secondary Diagnosis of Major Gastrointestinal Diagnosis

We examined the diagnosis codes assigned to MDC 6 for severity using the proposed CS DRGs and created a list of diagnosis codes that are identified as major or extreme in the APR DRGs or the consolidated severity DRGs. We refer to this set of higher severity diagnosis codes as Major Gastrointestinal Diagnoses. The list of higher severity diagnosis codes considered to be a Major Gastrointestinal Diagnosis is provided in the table below showing new DRG 569.

We then examined DRGs 148 and 149 (Major Small & Large Bowel Procedures with and without CC, respectively) and DRGs 154 through 156 (Stomach, Esophageal & Duodenal Procedures Age >17 with and without CC and Age 0-17, respectively) when these Major Gastrointestinal Diagnoses were present as either a principal or secondary diagnosis. In general, these Major Gastrointestinal Diagnoses represent or are associated with the reason for performing the surgical procedure in DRGs 148 and 149 and DRGs 154 through 156 and are the most serious diagnoses that necessitate surgery. As the following tables illustrate, the presence of these Major Gastrointestinal Diagnoses identifies patients with a higher level of severity. The presence of these Major Gastrointestinal Diagnoses leads to significantly higher average charges for these two groups of surgical patients, particularly for cases currently assigned to DRGs 148 and 154 which are the surgical procedures that include the presence of a CC. The surgical patients with Major Gastrointestinal Diagnoses would not only be considered to have a greater level of severity and be more expensive, they would also be assigned to the surgical DRG that includes a CC. The tables below show that patients in DRG 148 with a Major Gastrointestinal Diagnosis have average charges of $70,001.16 compared to average charges of $43,809.03 when a Major Gastrointestinal Diagnosis is not present. The difference in charges for cases in DRG 149 was not as great. The difference in average charges was $29,103.84 for DRG 149 when a Major Gastrointestinal Diagnosis was present and $23,077.84 when it was not. The number of cases with a Major Gastrointestinal Diagnosis was significantly larger for DRG 148 (58,153 cases compared to only 1,822 in DRG 149). Similar findings occur for DRGs 154, 155, and 156. Cases with a Major Gastrointestinal Diagnosis occur with significantly greater numbers in DRG 154 (9,924 compared to only 357 in DRG 155 and none in DRG 156). The average charges for cases with a Major Gastrointestinal Diagnosis were $84,270.92 for DRG 154, and only $29,193.81 for DRG 155.

Our medical advisors agree that these gastrointestinal surgical patients with a Major Gastrointestinal Diagnosis are more severely ill and represent patients with a higher level of severity. They support subdividing cases in DRG 148 and 154 based on the presence of a Major Gastrointestinal Diagnosis to better capture patients with higher level of severity. A summary of these changes is provided below.

We are deleting DRG 148 and creating the following two new DRGs:

• DRG 569 (Major Small & Large Bowel Procedures with CC with Major Gastrointestinal Diagnosis)
• DRG 570 (Major Small & Large Bowel Procedures with CC without Major Gastrointestinal Diagnosis)

The DRG logic for new DRGs 569 and 570 is as follows.

New DRG 569 will have a principal diagnosis from MDC 6 and one of the following codes as either the principal or secondary diagnosis. This DRG will also have an operating room procedure from current DRG 148 and a Complication/Comorbidity (as defined in CMS DRG GROUPER Version 24.0).

New DRG 570 will have an operating room procedure code from current CMS DRG 148 and a principal diagnosis from MDC 6, except for a principal or secondary diagnosis listed above in the Major Gastrointestinal Diagnosis list and will have a Complication/Comorbidity.

We also are deleting DRG 154 and creating two new DRGs as follows:

• DRG 567 (Stomach, Esophageal & Duodenal Procedures Age >17 with Complication/Comorbidity with Major Gastrointestinal Diagnosis)
• DRG 568 (Stomach, Esophageal & Duodenal Procedures Age >17 with Complication/Comorbidity without Major Gastrointestinal Diagnosis)

New DRG 567 will have a principal diagnosis from MDC 6 with either a principal or secondary diagnosis of a Major Gastrointestinal Diagnosis (see list of Major Gastrointestinal Diagnoses listed above). New DRG 567 will also have an operating room procedure from current CMS DRG 154 and a CC. New DRG 568 will have a principal diagnosis from MDC 6, except it will not have a principal or secondary diagnosis from the list of Major Gastrointestinal Diagnoses. It will also have an operating room procedure from current CMS DRG 154 and a CC.

d. MDC 11 (Diseases and Disorders of the Kidney and Urinary Tract): Major Bladder Procedures

Under our proposed CS DRGs, cases with a major bladder procedure were found to have a higher level of severity than were cases with other types of bladder procedures. Therfore, cases with a major bladder procedure are assigned to a single DRG in the CS DRGs. The procedures classified as a major bladder procedure are as follows:

The CMS DRGs assign these cases to one of the five following DRGs:

• DRG 303 (Kidney, Ureter & Major Bladder Procedures for Neoplasm).
• DRG 304 (Kidney, Ureter & Major Bladder Procedures for Non-Neoplasm with CC)
• DRG 305 (Kidney, Ureter & Major Bladder Procedures for Non-Neoplasm without CC)
• DRG 308 (Minor Bladder Procedures with CC)
• DRG 309 (Minor Bladder Procedures without CC)

Our medical advisors support creating a new DRG for major bladder procedures because they represent cases with higher levels of severity, are clinically different, and use greater resources. We examined data on cases containing a major bladder procedure and determined they represent cases with a higher level of severity and utilize significantly more resources than other cases within the DRGs where they are currently assigned. Cases with a major bladder procedure had average charges of $53,434 compared to $14,976 to $38,119 for other cases within the five DRGs where the patient did not have a major bladder procedure. The tables below illustrate these data.

Therefore, we are moving these procedures out of their current DRGs (DRG 303, 304, 305, 308, and 309) and into new DRG 573 (Major Bladder Procedures). A summary of these changes is as follows:

We are renaming the following three DRGs:

• DRG 303—“ Kidney and Ureter Procedures for Neoplasm”
• DRG 304—“ Kidney and Ureter Procedures for Non-Neoplasm With CC”
• DRG 305—“ Kidney and Ureter Procedures for Non-Neoplasm Without CC”

We are removing the following procedure codes from DRG 303-305, 308, and 309 and assigning them to new DRG 573. New DRG 573 will contain the following procedure codes.

e. MDC 16 (Diseases and Disorders of the Blood and Blood Forming Organs and Immunological Disorders): Major Hematological and Immunological Diagnoses

Under our proposed CS DRGs, major hematological and immunological diagnoses were found to identify cases with a higher level of severity. They are assigned to a single DRG under the CS DRGs. The diagnoses considered to be major hematological and immunological diagnoses include the following conditions:

These conditions are currently assigned to the following four CMS DRGs:

• DRG 395 (Red Blood Cell Disorders Age >17)
• DRG 396 (Red Blood Cell Disorders Age 0-17)
• DRG 398 (Reticuloendothelial & Immunity Disorders with CC)
• DRG 399 (Reticuloendothelial & Immunity Disorders without CC)

Our medical advisors agree that major hematological and immunological disorders are found in patients with significantly greater levels of severity and are different from other conditions in the four DRGs where they are assigned. Our data analysis shows that major hematological and immunological diseases identify patients with significantly greater levels of severity. They are more resource intensive than other conditions assigned to these four DRGs. Cases with major hematological and immunological conditions had average charges of $21,276 compared to $11,066 to $18,791 for the other conditions where these cases are currently assigned. Most of the nonhematological and immunological cases (96,557) are assigned to DRG 395 and have an average charge of $12,977.

We are creating a new CMS DRG 574 (Major Hematologic/Immunologic Diagnoses Except Sickle Cell Crisis and Coagulation Disorders). We are removing the codes mentioned in the table above from DRGs 395, 396, 398, and 399 and assigning them to new DRG 574. We also are assigning the new diagnosis codes indicated by an asterisk (*) to new DRG 574. These new codes also capture major hematological and immunological conditions and were created to provide more detail than the current codes in this section of ICD-9-CM. The DRG assignments for these new codes are also shown in Table 6A of the Addendum to this final rule.

f. MDC 18 (Infections and Parasitic Diseases (Systemic or Unspecified Sites)): O.R. Procedure for Patients With Infectious and Parasitic Diseases

Under the APR DRG system, cases in DRG 415 (O.R. Procedure for Infectious and Parasitic Diseases) are subdivided based on the presence or absence of one of the following principal diagnosis codes, which we are referring to as Postoperative or Post-Traumatic Infection:

• 958.3, Posttraumatic wound infection, not elsewhere classified
• 998.51, Infected postoperative seroma
• 998.59, Other postoperative infection
• 999.3, Infection complicating medical care, not elsewhere classified

The APR DRG system found cases with one of the above infection codes to represent a higher level of severity. Our medical advisors examined cases in the current CMS DRG system in DRG 415 and found that the presence of one of these infection codes as a principal diagnosis led to significantly higher levels of severity. Charge data also support this conclusion. The following table illustrates our findings.

As can be seen from the above table, cases in DRG 415 with a principal diagnosis except for postoperative or post-traumatic infection have average charges of $74,964.28. Cases with a principal diagnosis of postoperative or postπtraumatic infection have average charges of $43,154.68, or $31,809.60 less. Therefore, cases without one of the four infection codes, 958.3, 998.51, 998.59, and 999.3, have significantly higher severity levels than do cases that contain one of the four infection codes.

Accordingly, we are deleting DRG 415 and divide the cases into two new DRGs as follows:

• DRG 578, Infectious and Parasitic Diseases with O.R. Procedure
• DRG 579, Postoperative or Post-traumatic Infection with O.R. Procedure

Cases will be assigned to new DRG 578 if they were previously in DRG 415, but do not contain one of the following principal diagnosis codes:

• 958.3, Posttraumatic wound infection, not elsewhere classified
• 998.51, Infected postoperative seroma
• 998.59, Other postoperative infection
• 999.3, Infection complicating medical care, not elsewhere classified

Cases will be assigned to DRG 579 if they were previously assigned to DRG 415 and contain one of the four principal diagnosis codes listed above.

g. Severe Sepsis

Comment: As an alternative to the proposed CS DRGs, commenters recommended a new DRG to identify patients with severe sepsis associated with respiratory failure requiring mechanical ventilation. One commenter suggested using an approach to better recognize severity of illness that is similar to the change CMS implemented in the FYa2006 final rule for major cardiovascular conditions (MCVs). This approach involved examining the MCVs which could be present as either a principal or secondary diagnosis leading to greater severity of illness and resource consumption. Another option suggested by two commenters involved modifying DRGa416 (Septicemia Age >17) so that it would be split based on mechanical ventilation greater than 96 hours (code 96.72). The commenter stated that patients on mechanical ventilation for greater than 96 hours have a greater severity of illness than do those who are not on mechanical ventilation for 96 or more hours. Another commenter recommended considering mechanical ventilation as a pre-MDC DRG on the basis of the mechanical ventilation greater than 96 hours procedure code (96.72) to better recognize patients with a greater severity level. This commenter also provided an option to add systemic infections (038.x) as an acceptable principal diagnosis for DRG 475 when reported in conjunction with mechanical ventilation or tracheostomy. One commenter maintained that the clinical reason to address a new DRG for severe sepsis is related to proper recognition and treatment for this group of patients with a greater degree of severity. This commenter stated clinicians are getting better at understanding the importance of early recognition and treatment. As sepsis presents with organ dysfunction, treatments must be prompt or mortality rapidly increases according to the commenter.

Response: We analyzed data for patients in DRG 416 and 417 who are on mechanical ventilation for 96 or more hours. The following table shows our findings.

The data clearly show that DRG 416 septicemia patients who are on mechanical ventilation for 96 or more hours have a significantly greater severity of illness level and use greater resources than do other patients in DRG 416. Those patients on mechanical ventilation for 96 or more hours had average charges of $94,994 compared to $25,709 for other patients in DRG 416. We found no cases in DRG 417 with patients who reported mechanical ventilation for 96 or more hours. Therefore, we agree with the commenters that patients in DRG 416 who are on long term mechanical ventilation of 96 or more hours have greater severity of illness and use significantly greater resources. These patients should be assigned to a separate DRG to better reflect their higher severity level. Because we have no data on patients in DRG 417, we are not modifying that DRG at this time. Because the data on DRG 416 are compelling, we are deleting DRG 416 and splitting these cases into two new DRGs based on whether or not the patient is on mechanical ventilation for 96 or more hours. These two new DRGs are as follows:

• DRG 575 (Septicemia with Mechanical Ventilation 96 + Hours Age >17)
• DRG 576 (Septicemia without Mechanical Ventilation 96 + Hours Age >17)

Cases will be assigned to DRG 575 when they have a principal diagnosis from current DRG 416 and code 96.72 (Continuous mechanical ventilation for 96 consecutive hours or more). Cases will be assigned to DRG 576 when they have a principal diagnosis from current DRG 416 and do not have code 96.72.

We note that this DRG split is similar to the change we are making in MDC 4, for DRG 475 which was discussed earlier. The creation of these two new DRGs is distinct from the request to create a separate DRG for severe sepsis, which is discussed in section II.D.7. of this final rule.

D. Changes to Specific DRG Classifications

1. Pre-MDCs

a. Heart Transplant or Implant of Heart Assist System: Addition of Procedure to DRG 103

Based on public comments, we are assigning an additional procedure code to DRG 103 (Heart Transplant or Implant of Heart Assist System) under the pre-MDCs. In the FY 2006 IPPS final rule (70 FR 47297), we addressed suggestions concerning the placement of codes for external heart assist systems in DRG 103. Although we found that charges associated with code 37.65 (Implant of external heart assist system) were more than $100,000 lower than the average charges for all cases in DRG 103, we found that there was a subgroup of patients who were comparable in resource use and length of stay to other cases included in DRG 103. Those patients received both the external heart assist device (code 37.65) and later had the device removed (code 37.64, Removal of heart assist system) after a lengthy period of rest and recovery of their native hearts. We note that commenters provided external data indicating that survival rates are improving for patients receiving more advanced versions of these devices. In addition, commenters provided information indicating that longer periods of support with the external heart assist device are improving patients' survival chances and opportunity to be discharged with their native heart. These data show a 50-percent survival rate with an average total length of stay of 43 days for all AMI heart recovery patients. On average, a surviving patient will receive 31 days of average support time followed by an additional 38 days in the hospital after the device is removed. Based on information considered from a later year than our MedPAR data, it is clear that patients weaned from the external heart assist system have longer lengths of stay and are very different from the average patients having this procedure that were in our FY 2004 data.

Given the newness of this procedure and the latest generation of this device, the Medicare charge data included a limited number of patients having the device implanted and removed. However, the Medicare charge data did support that patients receiving both an implant and removal of an external heart assist system in a single hospital stay had an average length of stay exceeding 50 days and average charges of $378,000 that are more comparable to patients in DRG 103 than DRG 525 (Other Heart Assist System Implant). Accordingly, in FY 2006, we revised DRG 103 so that both implantation and removal of an external heart assist device in the same hospitalization would group to DRG 103.

However, we did not consider those cases where an external heart assist system is switched during a hospitalization, and replaced with another external heart assist system, that is subsequently removed. The ICD-9-CM coding structure specifies that the replacement of the system be coded to 37.63 (Repair of heart assist system), and not to 37.65. These cases are assigned to DRG 525 not DRG 103 even though the cases are comparable in resources expended, length of stay, etc., to other patients where the device is implanted and explanted during the same hospital stay.

Based on public comments, we believe that DRG 103 should be revised to take this situation into account. Therefore, we are reconfiguring DRG 103 in the following manner: Those patients who have both the replacement of an external heart assist system (code 37.63) and the explantation of that system (code 37.64) prior to the hospital discharge will be assigned to DRG 103.

By making this change, Medicare will be making higher payments for patients who receive both a replacement and an explant of an external heart assist system during a single hospital stay. Our intent in making this change is to recognize the higher costs of patients who have a longer length of stay and are discharged alive with their native heart. Cases in which a heart transplant also occurs during the same hospitalization episode will continue to be assigned to DRG 103.

b. Pancreas Transplants

On July 1, 1999, we issued coverage policy that specified that pancreas transplants were only covered when performed simultaneously with or after a Medicare covered kidney transplant. A noncoverage policy for pancreas transplant remained in effect for patients who had not experienced end stage renal failure secondary to diabetes. On July 29, 2005, we opened a national coverage determination (NCD) to determine whether pancreas transplant alone, that is, without a kidney transplant, is a reasonable and necessary service for Medicare beneficiaries. On April 26, 2006, we published the NCD for pancreas transplants on our Web site at: http://www.cms.hhs.gov/mcd/viewncd.asp?ncd_id=260.3&_version=3&basket=ncd%3A260%2E3% 3A3%3APancreas+Transplants. The NCD specifies the limited circumstances where the evidence is adequate to conclude that pancreas transplant alone is reasonable and necessary for Medicare beneficiaries.

Medicare coverage of pancreas transplants alone is limited to transplants in those facilities that are Medicare-approved for kidney transplantation. A listing of approved transplant centers can be found at: http://www.cms.hhs.gov/ESRDGeneralInformation/02_Data.asp#TopOfPage. The CMS NCD includes several criteria for the coverage of pancreas transplants alone, including having a diagnosis of Type I diabetes. (We refer readers to section 260.3 of the Medicare National Coverage Manual for the entire language of the NCD.)

Because we had issued a proposed NCD and a final NCD was not expected to be completed until late April 2006 (after completion of the proposed rule), we used the FY 2007 IPPS proposed rule to indicate the coding changes that we would make to DRG 513 (Pancreas Transplant) in FY 2007 if Medicare's final decision memorandum would have continued the program's national noncoverage of pancreas transplants (71 FR 24030). In addition, we also indicated the conforming changes that we would make to the MCE “NonCovered Procedure” edit if Medicare coverage was established for pancreas transplants alone. That discussion was included in section II.D.6. of the preamble of the proposed rule (71 FR 24039), which described proposed changes to the MCE.

Because the April 2006 Medicare final decision memorandum stated that the performance of pancreas transplants alone is reasonable and necessary for Medicare beneficiaries in limited circumstances, the logic for the determination of patient case assignment to DRG 513 in the FY 2006 GROUPER program needs to be modified to remove the requirement that patients also have kidney disease. Therefore, because the NCD was finalized, we are modifying DRG 513 to consist of the following logic: List A (the diabetes codes) of the required principal or secondary diagnosis codes remains the same, as does the required operating room procedures (codes 52.80 (Pancreatic transplant NOS), and 52.82, (Homotransplant of pancreas)). List B is removed from the logic; the following codes will no longer be required as a principal or secondary diagnosis:

• 403.01, Hypertensive kidney disease, malignant, with chronic kidney disease
• 403.11, Hypertensive kidney disease, benign, with chronic kidney disease
• 403.91, Hypertensive kidney disease, unspecified, with chronic kidney disease
• 404.02, Hypertensive heart and kidney disease, malignant, with chronic kidney disease
• 404.03, Hypertensive heart and kidney disease, malignant, with heart failure and chronic kidney disease
• 404.12, Hypertensive heart and kidney disease, benign, with chronic kidney disease
• 404.13, Hypertensive heart and kidney disease, benign, with heart failure and chronic kidney disease
• 404.92, Hypertensive heart and kidney disease, unspecified, with chronic kidney disease
• 404.93, Hypertensive heart and kidney disease, unspecified, with heart failure and chronic kidney disease
• 585.1, Chronic kidney disease, Stage I
• 585.2, Chronic kidney disease, Stage II (mild)
• 585.3, Chronic kidney disease, Stage III (moderate)
• 585.4, Chronic kidney disease, Stage IV (severe)
• 585.5, Chronic kidney disease, Stage V
• 585.6, End stage renal disease
• 585.9, Chronic kidney disease, unspecified
• V42.0, Organ or tissue replaced by transplant, kidney
• V43.89, Organ or tissue replaced by other means, other organ or tissue, other

We note that DRG 513 remains in the pre-MDC hierarchy.

Comment: Five commenters supported the proposed coding changes to DRG 513 and the MCE.

Response: We appreciate the support of the commenters. Accordingly, as the NCD for pancreas transplants alone was approved, in this final rule, we are adopting the changes as described above to DRG 513 and the MCE logic.

2. MDC 1 (Diseases and Disorders of the Nervous System)

a. Implantation of Intracranial Neurostimulator System for Deep Brain Stimulation (DBS)

Deep-brain stimulation (DBS) is designed to deliver electrical stimulation to the subthalamic nucleus or internal globus pallidus to ameliorate symptoms caused by abnormal neurotransmitter levels that lead to abnormal cell-to-cell electrical impulses in Parkinson's disease and essential tremor. DBS implants for essential tremor are unilateral, with neurostimulation leads on one side of the brain. DBS implants for Parkinson's disease are bilateral, requiring implantation of neurostimulation leads in both the left and right sides of the brain.

The implantation of a full DBS system requires two types of procedures. First, surgeons implant leads containing electrodes into the targeted sections of the brain where neurostimulation therapy is to be delivered. Second, a neurostimulator pulse generator is implanted in the pectoral region and extensions from the neurostimulator pulse generator are then tunneled under the skin along the neck and connected with the proximal ends of the leads implanted in the brain. Hospitals stage the two procedures required for a full-system DBS implant.

In FY 2005, to better account for these two types of procedures, we revised procedure code 02.93 (Implantation or replacement of intracranial neurostimulator lead(s)) for the lead placement and created three new procedures codes for the pulse generator: 86.94 (Insertion or replacement of single array neurostimulator pulse generator); 86.95 (Insertion or replacement of dual array neurostimulator pulse generator); and 86.96 (Insertion or replacement of other neurostimulator pulse generator). We published the new procedure codes and revised procedure code titles in Tables 6B and 6F of the FY 2005 IPPS final rule (69 FR 49627 and 49641).

In FY 2006, we made further refinements to the pulse generator codes to identify rechargeable pulse generators. We published the new procedure codes and revised procedure code titles in Tables 6B and 6F of the FY 2006 IPPS final rule (70 FR 47637 and 47639). The current list of pulse generators codes are:

• 86.94 (Insertion or replacement of single array neurostimulator pulse generator, not specified as rechargeable);
• 86.95 (Insertion or replacement of dual array neurostimulator pulse generator, not specified as rechargeable);
• 86.96 (Insertion or replacement of other neurostimulator pulse generator);
• 86.97 (Insertion or replacement of single array neurostimulator rechargeable generator); and
• 86.98 (Insertion or replacement of dual array neurostimulator rechargeable generator).

Kinetra® is an implantable dual array neurostimulator pulse generator that is approved for a new technology add-on payment through FYA2006. For more information about the new technology add-on payment, please refer to section II.G.3.a. of this preamble.

Medtronic, the manufacturer of Kinetra®, argues that the new technology add-on payment provision is designed to recognize the higher costs of new medical innovations for the initial period the technology is available on the market, and until the associated costs and charges related to the technology are available in the MedPAR database and can be used to recalibrate the DRG weights. Medtronic also argues that, once a technology is no longer eligible for new technology add-on payments, the new technology add-on payment provision is designed to support the reclassification of the technology to other clinically coherent DRGs with comparable resource costs.

With the conclusion of the new technology add-on payment, Medtronic is concerned that Kinetra® will be inadequately paid in DRG 1 (Craniotomy Age >17 With CC) or DRG 2 (Craniotomy Age >17 Without CC) under MDC 1. Medtronic recommended that CMS reassign the full-system Kinetra® implants to DRG 543 (Craniotomy with Implant of Chemo Agent or Acute Complex CNS Principal Diagnosis) under MDC 1. To accommodate this recommendation, procedure codes 02.93 and 86.95 would have to be reassigned to DRG 543 and the title for DRG 543 would have to be revised to “Craniotomy with Implantation of Major Device or Acute Complex CNS Principal Diagnosis.” Medtronic argued that DRG 543 would be a “clinically-consistent DRG that more appropriately reflects the resource utilization associated with full-system [deep brain stimulation] procedures.” Medtronic also emphasized that its proposal would only apply to full-system Kinetra® implants when both the leads and generators are implanted during a single inpatient stay and procedure codes 02.93 and 86.95 both appear on the claim. Medtronic believes the current DRG assignment is appropriate for partial system implants.

Medtronic provided an analysis of FY 2004 MedPAR data. Procedure code 86.95 was not created until FY 2005 so Medtronic used procedure codes 02.93 and 86.09 (Other incision of skin and subcutaneous tissue) to identify the full system. It identified 193 cases assigned to DRG 1 with average charges of approximately $69,155, and 532 cases assigned to DRG 2 with average charges of approximately $56,113.

In the FY 2007 IPPS proposed rule we indicated that we have reviewed the latest data for the full-system DBS implants assigned to DRG 1 or DRG 2 in the FY 2005 MedPAR file. We identified cases with procedure codes 02.93 and 86.95 for full-system dual array cases. We also identified cases with reported codes 02.93 and 86.96 for those full-system cases where the type of pulse generator was not specified. The following table displays our results:

These data showed that approximately one-quarter of the full-system dual array neurostimulator pulse generator cases are assigned to DRG 1 and approximately three-quarters of these cases are assigned to DRG 2. In both DRGs, the average length of stay was shorter for the full-system array neurostimulator pulse generator cases than for all other cases. However, the average charges for the full-system dual array neurostimulator pulse generator cases are approximately $18,000 and $27,000 higher than the average charges for DRGs 1 and 2, respectively. The average charges for these cases in DRG 1 are comparable to those for DRG 543. However, the more commonly occurring cases in DRG 2 have average charges that are less than those in DRG 543 by nearly $12,000. We reviewed all of the procedures that will result in a case being assigned to DRGs 1 and 2. Unlike the full-system DBS implants, we believe for most of the cases assigned to these DRGs, there will be no device cost to the hospital. For this reason, we believe the higher average charges and lower length of stay for cases involving full-system dual array neurostimulator pulse generators are likely accounted for by the cost of the device. While it is possible that the cost of the device itself will make the full-system DBS implants more expensive than other cases in the DRG, the hospital's charge markup may also explain the higher charges but lower average length of stay. As indicated in section II.G.3.a. of this final rule, the national average CCR for medical equipment and supplies is approximately 34 percent. Thus, the actual cost to the hospital of the case including the full-system dual array neurostimulator pulse generator may be much lower than the charges would suggest.

With respect to whether the cost of the technology itself, absent a charge markup, makes the case more expensive, in the FY 2007 IPPS proposed rule, we stated that we intended to address this issue as we make further refinements to the DRG system to address severity of illness as discussed in section II.C. of this preamble.

Comment: Several commenters opposed CMS’ proposed decision to retain the current assignment of implantable dual array neurostimulator pulse generator cases in DRGs 1 and 2. Several commenters stated that CMS should recognize the higher resources associated with this technology and reassign implantable dual array neurostimulator pulse generator cases to DRG 543. Two commenters disagreed with CMS’ statements that markups associated with Kinetra® may overstate the total charges of the implant procedure. Medtronic submitted information on charge compression in which the company contends that it conclusively finds the hospital charge markups for implantable devices are in fact significantly lower than for other, lower cost supplies and equipment. Medtronic and one other commenter argued that the total charges found in the FY 2005 MedPAR data associated with implantable dual array neurostimulator pulse generator procedures may be understated relative to other procedures in DRG 1, DRG 2 and DRG 543 and that reassignment of this technology to DRG 543 is fully warranted. The commenters stated that the implementation of the CS DRGs should be deferred to at least FY 2008 and not be a factor in CMS’ decision to make DRG reassignments this year.

Response: With regard to the issue of charge compression, we are studying this issue in our effort to improve payment accuracy in the IPPS. The average charges for the 51 cases in DRG 1 where the patient received a dual array neurostimulator are $17,426 or 31 percent higher than the rest of the cases in DRG 1. The average charges are comparable to those for DRG 543 ($73,020 for dual array neurostimulator cases and $71,138 for DRG 543).

The average charges for the 146 cases in DRG 2 are $26,623 or 81 percent higher than the rest of the cases in DRG 2 and only $12,000 less than the average charges for DRG 543. Based on these data, we believe that the dual array neurostimulator cases will be more accurately paid in DRG 543 than DRGs 1 and 2. We will be implementing this change to the DRG assignment for the full-system dual array neurostimulator cases for FY 2007. Implantable dual array neurostimulator pulse generator procedure cases reported with ICD-9-CM procedure codes 02.93 and 86.95 will be reassigned to DRG 543. We are changing the DRG title for DRG 543 to “Craniotomy With Major Device Implant or Acute Complex CNS Principal Diagnosis.”

b. Carotid Artery Stents

Background: Stroke is the third leading cause of death in the United States and the leading cause of serious, long-term disability. Approximately 70 percent of all strokes occur in people age 65 and older. The carotid artery, located in the neck, is the principal artery supplying the head and neck with blood. Accumulation of plaque in the carotid artery can lead to stroke either by decreasing the blood flow to the brain or by the plaque breaking free and lodging in the brain or other arteries leading to the head. The percutaneous transluminal angioplasty (PTA) procedure involves inflating a balloon-like device in the narrowed section of the carotid artery to reopen the vessel. A carotid stent is then deployed in the artery to prevent the vessel from closing or restenosing. A distal filter device (embolic protection device) may also be present, which is intended to prevent pieces of plaque from entering the bloodstream.

Effective July 1, 2001, Medicare covered PTA of the carotid artery concurrent with carotid stent placement when furnished in accordance with the FDA-approved protocols governing Category B Investigational Device Exemption (IDE) clinical trials. PTA of the carotid artery, when provided solely for the purpose of carotid artery dilation concurrent with carotid stent placement, was considered to be a reasonable and necessary service only when provided in the context of such clinical trials and, therefore, was considered a covered service for the purposes of those trials. Performance of PTA in the carotid artery when used to treat obstructive lesions outside of approved protocols governing Category B IDE clinical trials remained noncovered until the release of the October 12, 2004 NCD for PTA of the carotid artery in post-approval studies. This decision extended coverage of PTA in the carotid artery concurrent with placement of an FDA-approved carotid stent for an FDA-approved indication when furnished in accordance with the FDA-approved protocols governing post-approval studies. On March 17, 2005, CMS released an NCD that extended coverage to patients at high risk for carotid endarterectomy (CEA) who also have symptomatic carotid artery stenosis ≥70 percent. Procedures must be performed in CMS-approved facilities and with FDA-approved carotid artery stent(s) with distal embolic protection. (Section 20.7 of the NCD manual which discusses this decision may be viewed at the Web site: http://www.cms.hhs.gov/manuals/downloads/ncd103c1_Part1.pdf.

Placement of a carotid artery stent in patients who have had a disabling stroke (modified Rankin scale ≥3) is excluded from coverage.

We established codes for carotid artery stent procedures for use with discharges occurring on or after October 1, 2004, for inpatients who were enrolled in an FDA-approved clinical trial and who were using on-label FDA-approved stents and embolic protection devices. These codes are as follows:

• 00.61 (Percutaneous angioplasty or atherectomy of precerebral (extracranial vessel(s)); and
• 00.63 (Percutaneous insertion of carotid artery stent(s)).

We assigned procedure code 00.61 to four MDCs and seven DRGs. The most likely clinical scenario is that in which cases are assigned to MDC 1 (Diseases and Disorders of the Nervous System) in DRGs 533 (Extracranial Procedures with CC) and 534 (Extracranial Procedures without CC). Other DRG assignments can be found in Table 6B of the Addendum to the FY 2005 IPPS final rule (69 FR 49624). Code 00.63 is not considered a procedure code itself and should be used in combination with code 00.61.

Based on the results of evaluation of PTA and carotid stents for our FY 2006 final rule (70 FR 47300, August 12, 2005), we did not find sufficient evidence to warrant a DRG change at that time.

We again reviewed the PTA and insertion of a carotid stent(s) for the FY 2007 proposed rule, as manufacturer representatives suggested that we assign all carotid stenting cases to DRG 533 only, bypassing DRG 534. As we indicated in the FY 2007 IPPS proposed rule (71 FR 24032), we reviewed the FY 2005 MedPAR data on all cases in DRGs 533 and 534 and on those cases containing code 00.61 in combination with 00.63. The following table displays those results:

We found that 5.5 and 5.1 percent of the cases in DRGs 533 and 534, respectively, involved placement of a carotid artery stent. In DRG 533, the average length of stay was 19.4 percent shorter for the carotid stenting cases than for all other cases. In DRG 534, the average length of stay was 11.6 percent shorter for the carotid stenting cases than for all other cases. However, the average charges for the carotid stent cases were higher by $6,968 in DRG 533 and $7,804 in DRG 534. We reviewed all of the procedures that would result in a case being assigned to DRGs 533 and 534. Unlike the carotid artery stent placements, we believe that, for most of the other cases assigned to these DRGs, there will be no device cost to the hospital. For this reason, we believe the higher average charges and lower length of stay for the cases involving carotid artery stents could be accounted for by the cost of the device. We discussed the possibility that the cost of the device itself makes the stent cases more expensive than other cases in the DRG, and that the hospital's charge markup may also explain the higher charges but lower average length of stay. We also suggested that we intended to address this issue as we make further refinements to the CS DRG system previously described. The use of a carotid stent or stents may increase complexity and resource use even though the patient is not necessarily more severely ill. We indicated that we believed that the CS DRG system we proposed would need to be further refined to assign cases based on complexity as well as severity to account for technologies such as carotid stents that increase costs. For this reason, we did not propose a change to the current DRG assignment for these cases.

Comment: More than a dozen commenters addressed this topic. State hospital associations, in particular, were unanimous in their recommendation that all carotid stenting cases should immediately be assigned only to DRG 533, bypassing DRG 534 entirely. The commenters suggested this solution to increase payments to hospitals in order that the higher costs associated with carotid stents are recognized within the existing DRG system.

Response: We are opposed to this suggestion. The DRGs comprise a native structure of the types of patients within each DRG category. Further, this structure is based on an organizing principle. For example, cases in DRGs 533 and 534 are organized on the principle of surgical approach (extracranial procedures) as well as the presence or absence of CCs. To ignore the structure of the DRG solely for the purpose of increasing payment would set an unwelcome precedent for defining all of the other DRGs in the system.

Comment: Several commenters mentioned that, while CMS suggested that the higher average charges and lower lengths of stay for cases involving carotid artery stents are likely accounted for by the cost of the device, CMS provided no evidence to support this assertion.

Response: The average length of stay for patients in DRGs 533 and 534 with the placement of carotid stent(s) are 19.4 and 11.6 percent shorter than the other patients assigned to DRGs 533 and 534, respectively. Therefore, a long length of stay is not the reason for the higher average charges. We based our assertion on the contribution of the cost of the device to the total cost of the patients in these DRGs compared to other cases in the DRG with longer lengths of stay. We note that the next comment suggests that our analysis is correct that the higher charges for the carotid artery stent cases relative to other cases in the DRG are, in part, associated with higher supply costs.

Comment: One commenter suggested that CMS create a new pair of DRGs with and without MCVs until the adequacy of payment under the severity adjustment methodology is fully assessed. This commenter noted that, while length of stay and operating room costs are lower for carotid stenting, supply and radiology charges associated with the stent and the angiography are higher, resulting in higher overall costs for carotid stenting.

Response: While we recognize the creativity of this approach, we note that the MCVs are applicable to cases in MDC 5 (Diseases and Disorders of the Circulatory System), while DRGs 533 and 534 are in MDC 1 (Diseases and Disorders of the Nervous System). Such an approach for MDC 1 might have merit, but we would want to evaluate the entire MDC thoroughly before creating such a list of complicating diagnoses. We will further consider this concept as we evaluate severity DRG systems for adoption in FY 2008.

Comment: One commenter, while urging CMS to reconsider our decision not to assign all carotid cases to DRG 533, noted that the current National Coverage Determination on CAS [Carotid Artery Stenting] very clearly states that only those patients who are at high risk for [open] surgery due to the presence of a detailed list of complications or comorbidities are eligible for carotid artery stenting. Therefore, by CMS’ own characterization, all patients undergoing carotid artery stenting have complications and comorbidities and should be assigned to DRG 533.

Response: This assumption is theoretically correct. However, the detailed list of comorbidities or anatomical risk factors that are required to support the surgeon's decision to perform carotid stenting instead of a carotid endarterectomy is not the same as the CMS list of CCs. For example, amaurosis fugax, code 362.34 (Transient arterial occlusion) is recognized as a risk factor which would justify carotid stenting, but is not recognized by the CMS GROUPER as a diagnosis defined as a CC.

Comment: Several commenters suggested that CMS create two new DRGs for the carotid stent cases.

Response: We note that the number of procedures has increased from the data reported in the FY 2006 IPPS final rule (70 FR 47300), thus indicating acceptance of this procedure by the medical community as a main-stream surgical alternative. In FY 2006, as the specific codes for carotid stenting had only been in use since October 1, 2004, we used the existing codes 39.50 (Angioplasty or atherectomy of other noncoronary vessel(s)) and 39.90 (Insertion of non-drug-eluting peripheral vessel stent(s)), in combination with principal diagnosis code 433.10 (Occlusion and stenosis of carotid artery, without mention of cerebral infarction) as a proxy for the number of cases involved in clinical trials. In DRG 533, we had 1,586 cases with the proxy codes reported, and in DRG 534, there were 1,397 cases. In FY 2005, the patients represented 3.5 percent and 3.3 percent of all cases in DRGs 533 and 534, respectively. That figure has now climbed to 2,400 cases and 2,056 cases, and 5.5 percent and 5.1 percent, respectively.

In addition, the difference in the average charges are 26 percent higher for carotid artery stent cases in DRG 533 than for the average charges in all cases in that DRG, and 45 percent higher using the same parameters for DRG 534. We believe these data are compelling enough to warrant creation of a new DRG.

Accordingly, we are creating DRG 583 (Carotid Artery Stent Procedure). This DRG will be located in MDC 1, and will be hierarchically ordered above DRGs 533 and 534. DRG 583 will contain two procedure codes. Code 00.61 will determine the DRG, and will be combined with code 00.63. Both codes must be reported in order for cases to be assigned to this DRG.

We are not splitting this DRG based on the presence or absence of a CC as suggested by the commenters. One criterion for splitting a DRG based on the presence or the absence of a CC is that it must have an impact of at least $40 million. In this situation, the overall average of the charges for all cases in DRGs 533 and 534 is $30,193. We then subtracted the actual average charges for only the carotid stent cases in both DRGs 533 and 534, and multiplied that figure by the actual number of cases. For DRG 533 and DRG 534, we estimate an impact of approximately $10 million each. Added together, the total impact would be $20 million, falling short of our threshold of a $40 million impact to create a CC/non-CC split. Therefore, we are not creating a CC/non-CC split in the DRG for carotid artery stenting at this time.

We reiterate that coverage of the carotid artery stent procedure is limited to patients at risk of developing a stroke due to narrowing or stenosis of the carotid artery. Diagnosis code 433.10 (Occlusion and stenosis of carotid artery without mention of cerebral infarction) should be used to identify the site of the procedure in the carotid artery. If it is necessary to identify bilateral occlusion or stenosis, diagnosis code 433.30 (Occlusion and stenosis of multiple and bilateral arteries without mention of cerebral infarction) may also be used. These codes should be used together, as code 433.30 contains arterial sites that are not currently covered for Medicare patients. Reporting of code 433.30 alone will cause the case to fail the editing system at the fiscal intermediary, and the case could be denied.

Inclusion of the fifth digit of “1” (with cerebral infarction) with either 433.1x or 433.3x will cause the claim to be rejected.

3. MDC 5 (Diseases and Disorders of the Circulatory System)

a. Insertion of Epicardial Leads for Defibrillator Devices

As we indicated in the FY 2007 IPPS proposed rule (71 FR 24033), we received a comment indicating that a change in coding advice for the insertion of epicardial leads for CRT-D defibrillator devices affects DRG assignment. The commenter noted that the Third Quarter 2005 issue of the American Hospital Association's publication Coding Clinic for ICD-9-CM instructs coders to assign code 37.74 (Insertion or replacement of epicardial lead [electrode] into atrium) for pacemaker or defibrillator leads inserted through use of a thoracotomy into the epicardium. While the use of code 37.74 is standard coding practice for pacemakers, the advice is new for defibrillators. This coding advice was discussed at the ICD-9-CM Coordination and Maintenance Committee meeting held on September 29 and 30, 2005. Participants at the Committee meeting proposed modifications for the code category 37.7 (insertion, revision, replacement, and removal of pacemaker leads; insertion of temporary pacemaker system; and revision of cardiac device pocket). These modifications involved expanding the category so that the codes for leads would no longer be restricted to pacemakers. This change would guide coders to use code 37.74 for the insertion of epicardial leads for both defibrillators and pacemakers for the ICD-9-CM and will become effective on October 1, 2006.

The commenter indicated that this coding advice would restrict some defibrillator cases from being assigned to the defibrillator DRGs. Specifically, the commenter expressed concerns about the DRG logic for the following DRGs:

• DRG 515 (Cardiac Defibrillator Implant without Cardiac Catheter)
• DRG 535 (Cardiac Defibrillator Implant with Cardiac Catheter with AMI/Heart Failure/Shock)
• DRG 536 (Cardiac Defibrillator Implant with Cardiac Catheter without AMI/Heart Failure/Shock)

Cases are assigned to one of these three DRGs when a total defibrillator system, including both the device and one or more leads, is implanted. The implant could be represented by the ICD-9-CM codes for the total system, that is, code 00.51 (Implantation of cardiac resynchronization defibrillator, total system [CRT-D]) or code 37.94 (Implantation or replacement of automatic cardioverter/defibrillator, total system [AICD]). Cases can also be assigned to DRGs 515, 535, and 536 when a combination of a device and a lead code is reported. The following combinations of defibrillator device and lead codes are present in the current DRG logic:

• 00.52 (Implantation or replacement of transvenous lead [electrode] into left ventricular coronary venous system) and 00.54 (Implantation or replacement of cardiac resynchronization defibrillator, pulse generator device only [CRT-D])
• 37.95 (Implantation of automatic cardioverter/defibrillator lead(s) only) and 00.54 (Implantation or replacement of cardiac resynchronization defibrillator, pulse generator device only [CRT-D])
• 37.95 (Implantation of automatic cardioverter/defibrillator lead(s) only) and 37.96 (Implantation of automatic cardioverter/defibrillator pulse generator only)
• 37.97 (Replacement of automatic cardioverter/defibrillator lead(s) only) and 00.54 (Implantation or replacement of cardiac resynchronization defibrillator, pulse generator device only [CRT-D])
• 37.97 (Replacement of automatic cardioverter/defibrillator lead(s) only) and 37.98 (Replacement of automatic cardioverter/defibrillator pulse generator only)

A DRG logic issue has arisen concerning the instruction to use code 37.74 for epicardial leads inserted with CRT-D defibrillators. The new combination of a defibrillator device with an epicardial lead (code 37.74) is not included in DRGs 515, 535, and 536. The commenter recommended that the following combinations be added to DRGs 515, 535, and 536 so that all types of defibrillator device and lead combinations would be included: code 37.74 and code 00.54; code 37.74 and code 37.96; and code 37.74 and code 37.98.

We agree that these three combinations should be added to the list of combination codes included in DRGs 515, 535, and 536. This change would result in all combinations of defibrillator devices and leads being assigned to one of the defibrillator DRGs. Therefore, in the FY 2007 IPPS proposed rule, we proposed to add these three combinations to the list of procedure combinations under DRGs 515, 535, and 536.

Comment: A number of commenters supported adding the new combinations of defibrillator devices with the epicardial leads to DRGs 515, 535, and 536. One commenter stated that this change would bring the DRGs into alignment with the change in coding advice to assign code 37.74 in conjunction with implantation of CRT-D defibrillators.

Response: We appreciate the support of commenters and agree that this change would bring the DRGs into alignment with the change in coding advice.

In this final rule, we are adding the following combinations of device and lead codes to DRGs 515, 535, and 536: code 37.74 and code 00.54; code 37.74 and code 37.96; and code 37.74 and code 37.98.

b. Application of Major Cardiovascular Diagnoses (MCVs) List to Defibrillator DRGs

In the FY 2006 IPPS final rule (70 FR 47289 and 47474 through 47479), we addressed a comment we had received in response to the FY 2006 proposed rule which noted that section 507(c) of Pub. L. 108-173 required MedPAC to conduct a study to determine how the DRG system should be updated to better reflect the cost of delivering care in a hospital setting. The commenter noted that MedPAC reported that the “cardiac surgery DRGs have high relative profitability ratios.” While the commenter acknowledged that it may take time to conduct and complete a thorough evaluation of the MedPAC payment recommendations for all DRGs, the commenter strongly encouraged CMS to revise the cardiac DRGs through patient severity refinement as part of the IPPS final rule effective for FY 2006.

In response to this comment, we performed an extensive review of the cardiovascular DRGs in MDC 5, particularly those DRGs that were commonly billed by specialty hospitals. We observed that there was some overlap between the lists of cardiovascular complications and complex diagnoses and that these lists were already used to segregate patients into DRGs that used greater resources. Because the hospital industry already was familiar with the major complication and complex diagnosis lists used within the cardiovascular DRGs, we began our analysis with these two overlapping lists.

The two lists were originally developed for the current DRG system because they contained conditions that could have an impact on the resources needed to treat a patient with cardiovascular complications. Many of the conditions were cardiovascular diagnoses and, therefore, would be classified to MDC 5. However, we determined that some of the diagnoses were not cardiovascular, but would still have an impact on a patient with cardiovascular complications. The conditions that were not cardiovascular diagnoses were not assigned to MDC 5 if they were the principal diagnosis.

We reviewed the conditions on the two overlapping lists and identified conditions that we believed would lead to a more complicated patient stay requiring greater resource use. We referred to these conditions as “major cardiovascular conditions (MCVs).” The MCVs could be present as either a principal diagnosis or a secondary diagnosis and lead to greater resource consumption. The complete list of MCVs was published in the FY 2006 IPPS final rule (70 FR 47477 and 47478).

In the FY 2006 IPPS final rule, we also adopted new DRGs 547 through 558, effective October 1, 2005 (70 FR 47475 and 47476). However, we emphasized that the refinements to the DRGs were being taken as an interim step to better recognize severity in the DRG system for FY 2006 until we could complete a more comprehensive analysis of the APR DRG system and the CC list as part of a complete analysis of the MedPAC recommendations that we planned to perform for FY 2007 (and which was addressed in section II.C. of the preamble of the FY 2007 proposed rule).

Since publication of the FY 2006 IPPS final rule, we have received a question from a commenter as to why we did not apply the MCV list to the following defibrillator DRGs: 515, 535, and 536. The commenter noted that the pacemaker DRGs were revised using the MCV list, but the defibrillator DRGs were not.

As noted above, for FY 2006, we created new DRGs 546 through 558 to identify cases with more costly and severely ill patients as an interim step to evaluating severity DRGs. We analyzed for the first time last year data on cases within MDC 5 and presented data that showed significant difference for patients in certain DRGs based on the presence or absence of an MCV. This split did not work for the defibrillator DRGs, as we could not identify groups with significantly different resource use. For instance, splitting DRG 515 based on the presence of an MCV would lead to two groups with differences in charges of only $3,430 ($89,341 for those with an MCV and $85,911 for those without an MCV). In the data we displayed in the FY 2006 IPPS final rule, the differences for DRGs selected for an MCV split ranged from $10,319 to $21,035. Splitting DRG 515 based on an MCV would produce a difference in charges of only 10.1 percent as compared to differences of 28.7 to 47.7 percent for DRGs 547 through 558. Therefore, the data did not support including DRG 515 among those split based on the presence or absence of an MCV. Similar results were found when DRG 536 was split by an MCV. There was only an 8.1 percent difference in charges between the two groups. We also identified other problems with splitting DRG 535 based on the presence or absence of an MCV. Some of the codes a claim must include for the case to be grouped to DRG 535 under our current system are also codes on the MCV list. Therefore, applying the MCV list to DRG 535 would result in all cases being assigned to the DRG with an MCV and none to the DRG without an MCV. For these reasons, we did not subdivide DRGs 515, 535, and 536 based on the presence or absence of an MCV.

In the FY 2007 IPPS proposed rule, we indicated that we had decided not to propose additional refinements of the DRGs based on MCVs for FY 2007 because of our efforts to propose a broader refinement of the DRG system, as discussed in detail in section II.C. of the proposed rule. However, as discussed further in section II.C. of the preamble of the proposed rule, we solicited comments on whether it would be appropriate in FY 2007 to apply a clinical severity concept to an expanded set of DRGs, similar to the approach we used in FY 2006 to refine cardiac DRGs based on the presence or absence of an MCV.

Comment: Commenters agreed with the recommendation that we not subdivide DRGs 515, 535, and 536 based on MCV. However, one commenter expressed concerns about how the current DRGs were achieving their goal of identifying patients with greater severity of illness. Other commenters opposed the proposal to delay refining defibrillator DRGs based on MCVs. These commenters believed it was appropriate for CMS to apply a clinical severity concept similar to the approach used in FY 2006 to refine cardiac DRGs to an expanded set of DRGS (for example, defibrillator DRGs) based on the presence or absence of an MCV.

Response: We agree with the commenters who suggested that our goal should be to reform the Medicare DRG system to develop a better means of capturing severity of illness and complexity. As discussed in section II.C. of the preamble of the proposed rule, we solicited comments on whether it would be appropriate in FY 2007 to apply a clinical severity concept to an expanded set of DRGs, similar to the approach we used in FY 2006 to refine cardiac DRGs based on the presence or absence of an MCV. As discussed in section II.C.7., we are implementing revisions to the current DRGs to better recognize severity of illness. However, the analysis we have performed to this point does not support splitting defibrillator DRGs based on the presence or absence of an MCV. As stated earlier, simply applying the MCVs to the defibrillator DRGs in DRGs 515, 535, and 536 would not lead to significant improvements for DRG 515. Applying the MCV list to DRG 535 would result in all cases being assigned to the DRG with an MCV and none to the DRG without an MCV. For these reasons, we did not subdivide DRGs 515, 535, and 536 based on the presence or absence of an MCV.

While we did not find additional severity improvements for defibrillator cases, we will continue to study this area and look for further improvements.

4. MDC 8 (Diseases and Disorders of the Musculoskeletal System and Connective Tissue)

a. Hip and Knee Replacements

In the FY 2006 final rule (70 FR 47303), we deleted DRG 209 (Major Joint and Limb Reattachment Procedures of Lower Extremity) and created new DRGs 544 (Major Joint Replacement or Reattachment of Lower Extremity) and 545 (Revision of Hip or Knee Replacement) to help resolve payment issues for hospitals that perform revisions of joint replacements because we found revisions of joint replacements to be significantly more resource intensive than original hip and knee replacements. DRG 544 includes the following code assignments:

• 81.51, Total hip replacement
• 81.52, Partial hip replacement
• 81.54, Total knee replacement
• 81.56, Total ankle replacement
• 84.26, Foot reattachment
• 84.27, Lower leg or ankle reattachment
• 84.28, Thigh reattachment

DRG 545 includes the following procedure code assignments:

• 00.70, Revision of hip replacement, both acetabular and femoral components
• 00.71, Revision of hip replacement, acetabular component
• 00.72, Revision of hip replacement, femoral component
• 00.73, Revision of hip replacement, acetabular liner and/or femoral head only
• 00.80, Revision of knee replacement, total (all components)
• 00.81, Revision of knee replacement, tibial component
• 00.82, Revision of knee replacement, femoral component
• 00.83, Revision of knee replacement, patellar component
• 00.84, Revision of knee replacement, tibial insert (liner)
• 81.53, Revision of hip replacement, not otherwise specified
• 81.55, Revision of knee replacement, not otherwise specified

In the FY 2006 IPPS final rule (70 FR 47305), we indicated that the American Association of Orthopaedic Surgeons had requested that, once we receive claims data using the two DRG procedure code assignments, we closely examine data from the use of the codes under the two DRGs to determine if future additional DRG modifications are needed.

After publication of the FY 2006 IPPS final rule, a number of hospitals and coding personnel advised us that the DRG logic for DRG 471 (Bilateral or Multiple Major Joint Procedures of Lower Extremity), which utilizes the new and revised hip and knee procedure codes under DRGs 544 and 545, also includes codes that describe procedures that are not bilateral or that do not involve multiple major joints. DRG 471 was developed to include cases where major joint procedures such as revisions or replacements were performed either bilaterally or on two joints of one lower extremity. We changed the logic for DRG 471 last year for the first time when we added the new and revised codes. The commenters indicated that, by adding the more detailed codes that do not include total revisions or replacements to the list of major joint procedures to DRG 471, we are assigning cases to DRG 471 that do not have bilateral or multiple joint procedures. For example, when a hospital reports a code for revision of the tibial component (code 00.81) and patellar component of the right knee (code 00.83), the current DRG logic assigns the case to DRG 471. The commenters indicated that this code assignment is incorrect because only one joint has undergone surgery, but two components were used. One commenter indicated that ICD-9-CM does not identify left/right laterality. Therefore, it is difficult to use the current coding structure to determine if procedures are performed on the same leg or on both legs. The commenters raised a concern about whether CMS intended to pay hospitals using DRG 471 for procedures performed on one joint. The commenters indicated that the DRG assignments for these codes would also make future data analysis misleading. The commenters recommended removing codes from DRG 471 that do not specifically identify bilateral or multiple joint procedures.

We agree that the new and revised joint procedure codes should not be assigned to DRG 471 unless they include bilateral and multiple joints. Therefore, in the FY 2007 IPPS proposed rule (71 FR 24035), we proposed to remove the following codes from DRG 471:

• 00.71, Revision of hip replacement, acetabular component
• 00.72, Revision of hip replacement, femoral component
• 00.73, Revision of hip replacement, acetabular liner and/or femoral head only
• 00.81, Revision of knee replacement, tibial component
• 00.82, Revision of knee replacement, femoral component
• 00.83, Revision of knee replacement, patellar component
• 00.84, Revision of total knee replacement, tibial insert (liner)
• 81.53, Revision of hip replacement, not otherwise specified
• 81.55, Revision of knee replacement, not otherwise specified

The proposed revised DRG 471 would then contain only the following codes:

• 00.70, Revision of hip replacement, both acetabular and femoral components
• 00.80, Revision of knee replacement, total (all components)
• 81.51, Total hip replacement
• 81.52, Partial hip replacement
• 81.54, Total knee replacement
• 81.56, Total ankle replacement

We proposed to assign the codes removed from DRG 471 (codes 00.71, 00.72, 00.73, 00.81, 00.82, 00.83, 00.84, 81.53, and 81.55) to DRG 545 when used either alone or in combination. This list of codes removed from DRG 471 and added to DRG 545 includes partial revisions of the knee and hip as well as unspecified joint procedures such as code 81.55 where it is not clear if the revision is total or partial.

Comment: Several comments supported our proposals to remove codes 00.71, 00.72, 00.73, 00.81, 00.82, 00.83, 00.84, 81.53, and 81.55 from the combinations assigned to DRG 471 and assign cases with these codes to DRG 545. The commenters agreed that these codes should be removed from DRG 471 because they do not represent bilateral and multiple joint revisions or replacements.

Response: We appreciate the commenters support to remove codes 00.71, 00.72, 00.73, 00.81, 00.82, 00.83, 00.84, 81.53, and 81.55 from the combinations assigned to DRG 471. These cases will be assigned to DRG 545.

We are finalizing the changes to DRG 471 and DRG 545 that we proposed. Further, as we indicated in the proposed rule, we plan to perform extensive data analysis on the new and revised joint procedure codes as we receive billing data to determine if future refinements of these DRGs are needed. In addition, as indicated in section II.C. of the preamble of the proposed rule, we are planning in the future to adopt a revised DRG system for the IPPS that addresses severity of illness. We encouraged commenters to evaluate how the new and revised joint procedures should be addressed in such a revised system. We received comments indicating that the CS DRGs that we proposed do not distinguish between patients receiving an original joint replacement from a revision. As we indicate elsewhere in this final rule, we will evaluate these issues as we develop our plans for adopting a revised DRG system that addresses severity of illness.

b. Spinal Fusion

In the FY 2006 IPPS final rule (70 FR 47307), we created new DRG 546 (Spinal Fusions Except Cervical with Curvature of the Spine or Malignancy). DRG 546 is composed of all noncervical spinal fusions previously assigned to DRGs 497 (Spinal Fusion Except Cervical with CC) and 498 (Spinal Fusion Except Cervical without CC) that have a principal or secondary diagnosis of curvature of the spine or a principal diagnosis of a malignancy. The principal diagnosis codes that lead to DRG 546 assignment are the following:

• 170.2, Malignant neoplasm of vertebral column, excluding sacrum and coccyx
• 198.5, Secondary malignant neoplasm of bone and bone marrow
• 213.2, Benign neoplasm of bone and articular cartilage; vertebral column, excluding sacrum and coccyx
• 238.0, Neoplasm of uncertain behavior of other and unspecified sites and tissues; Bone and articular cartilage
• 239.2, Neoplasms of unspecified nature; bone, soft tissue, and skin
• 732.0, Juvenile osteochondrosis of spine
• 733.13, Pathologic fracture of vertebrae
• 737.0, Adolescent postural kyphosis
• 737.10, Kyphosis (acquired) (postural)
• 737.11, Kyphosis due to radiation
• 737.12, Kyphosis, postlaminectomy
• 737.19, Kyphosis (acquired), other
• 737.20, Lordosis (acquired) (postural)
• 737.21, Lordosis, postlaminectomy
• 737.22, Other postsurgical lordosis
• 737.29, Lordosis (acquired), other
• 737.30, Scoliosis [and kyphoscoliosis], idiopathic
• 737.31, Resolving infantile idiopathic scoliosis
• 737.32, Progressive infantile idiopathic scoliosis
• 737.33, Scoliosis due to radiation
• 737.34, Thoracogenic scoliosis
• 737.39, Other kyphoscoliosis and scoliosis
• 737.8, Other curvatures of spine
• 737.9, Unspecified curvature of spine
• 754.2, Congenital scoliosis
• 756.51, Osteogenesis imperfecta

The secondary diagnoses that will lead to DRG 546 assignment are:

• 737.40, Curvature of spine, unspecified
• 737.41, Curvature of spine associated with other conditions, kyphosis
• 737.42, Curvature of spine associated with other conditions, lordosis
• 737.43, Curvature of spine associated with other conditions, scoliosis

After publication of the FY 2006 IPPS final rule, we received a comment stating that creating new DRG 546 was insufficient to address clinical severity and resource differences among spinal fusion cases that involve fusing multiple levels of the spine. Specifically, the commenter suggested that the spinal fusion DRGs be further modified to incorporate Bone Morphogenic Protein (BMP), code 84.52 (Insertion of recombinant bone morphogenetic protein). The commenter also suggested that CMS apply a clinical severity concept to all back and spine surgical cases similar to the approach that we used for the MCVs to refine the cardiac DRGs in the final rule for FY 2006. The commenter recommended recognizing additional conditions that reflect higher resource needs, regardless of whether they are principal or secondary diagnoses. The commenter also suggested that the spine DRGs be further subdivided based on the use of specific spinal devices such as artificial discs. These changes would entail the creation of 10 new spine DRGs in addition to other changes requested.

Response: We agree that it is important to recognize severity when classifying patients into specific DRGs. In response to recommendations made by MedPAC last year that are discussed in section II.C. of this final rule, we are conducting a comprehensive analysis of the entire DRG system to determine whether to undertake significant reform to better recognize severity of illness. At this time, we believe it is premature to develop a severity adjustment for spine surgeries while we are considering a more systematic approach to capturing severity of illness across all DRGs. We also believe it would be premature to make revisions to DRG 546 because this DRG was created on October 1, 2005, and we do not yet have data to analyze its impact. Given the number of innovations occurring in spinal surgery over the last several years (for example, artificial spinal disc prostheses, kyphoplasty, and vertebroplasty), we agree that additional analysis of the spine DRGs would be warranted if we were to continue with the current DRG system and not adopt CS DRGs. However, as discussed above, in the FY 2007 IPPS proposed rule, we proposed to develop a severity-adjusted DRG system. For this reason, we are not further researching this issue for FY 2007. However, in the proposed rule, we encouraged commenters to examine the proposed CS DRG system described in section II.C. of the preamble of the proposed rule to determine whether there is a better recognition of severity of illness and resource use in that system.

Comment: One commenter stated that it was premature to consider splitting the spinal fusion DRGs into potentially up to 10 new DRGs at this time. The commenter stated there is a need for additional data analysis prior to recommending new DRGs.

Response: We agree with the commenter that it is premature to consider splitting the spinal fusion DRGs into as many as 10 new DRGs. We will continue to study this area. In the meantime, we will not modify the spinal fusion DRGs for October 1, 2006.

c. CHARITE^TM Spinal Disc Replacement Device

CHARITE^TM is a prosthetic intervertebral disc. On October 26, 2004, the FDA approved the CHARITE^TM Artificial Disc for single level spinal arthroplasty in skeletally mature patients with degenerative disc disease between L4 and S1. On October 1, 2004, we created new procedure codes for the insertion of spinal disc prostheses (codes 84.60 through 84.69). We provided the DRG assignments for these new codes in Table 6B of the FY 2005 IPPS proposed rule (69 FR 28673). We received comments on the FY 2005 proposed rule recommending that we change the assignments for these codes from DRG 499 (Back and Neck Procedures Except Spinal Fusion With CC) and DRG 500 (Back and Neck Procedures Except Spinal Fusion Without CC) to the DRGs for spinal fusion, DRG 497 (Spinal Fusion Except Cervical With CC) and DRG 498 (Spinal Fusion Except Cervical Without CC) for procedures on the lumbar spine and to DRGs 519 and 520 for procedures on the cervical spine. In the FY 2005 IPPS final rule (69 FR 48938, August 11, 2004), we indicated that DRGs 497 and 498 are limited to spinal fusion procedures. Because the surgery involving the CHARITE^TM Artificial Disc is not a spinal fusion, we decided not to include this procedure in these DRGs. However, we stated that we would continue to analyze this issue and solicited further public comments on the DRG assignment for spinal disc prostheses.

In the FY 2006 final rule (70 FR 47353, August 12, 2005), we noted that, if a product meets all of the criteria for Medicare to pay for the product as a new technology under section 1886(d)(5)(K) of the Act, there is a clear preference expressed in the statute for us to assign the technology to a DRG based on similar clinical or anatomical characteristics or costs. However, for FY 2006, we did not find that the CHARITE^TM Artificial Disc met the substantial clinical improvement criterion and, thus, did not qualify as a new technology. Consequently, we did not address the DRG classification request made under the authority of this provision of the Act.

However, we did evaluate whether to reassign the CHARITE^TM Artificial Disc to different DRGs using the Secretary's authority under section 1886(d)(4) of the Act (70 FR 47308, August 12, 2005). We indicated that we did not have Medicare charge information to evaluate DRG changes for cases involving an implant of a prosthetic intervertebral disc like the CHARITE^TM and did not make a change in its DRG assignments. We stated that we would consider whether changes to the DRG assignments for the CHARITE^TM Artificial Disc were warranted for FY 2007, once we had information from Medicare's data system that would assist us in evaluating the costs of these patients.

As we discussed in the FY 2007 IPPS proposed rule (71 FR 24036), we received correspondence regarding the DRG assignments for the CHARITE^TM Artificial Disc, code 84.65 (Insertion of total spinal disc prosthesis, lumbosacral). The commenter had previously submitted an application for the CHARITE^TM Artificial Disc for new technology add-on payments for FY 2006 and had requested a reassignment of cases involving CHARITE^TM implantation to DRGs 497 and 498. The commenter asked that we examine claims data for FY 2005 and reassign procedure code 84.65 from DRGs 499 and 500 into DRGs 497 and 498. The commenter again stated the view that cases with the CHARITE^TM Artificial Disc reflect comparable resource use and similar clinical indications as do those in DRGs 497 and 498. If CMS were to reject reassignment of the CHARITE^TM Artificial Disc to DRGs 497 and 498, the commenter suggested creating two separate DRGs for lumbar disc replacements.

On February 15, 2006, we posted a proposed national coverage determination (NCD) on the CMS Web site seeking public comment on our proposed finding that the evidence is not adequate to conclude that lumbar artificial disc replacement with the CHARITE^TM Artificial Disc is reasonable and necessary. The proposed NCD stated that lumbar artificial disc replacement with the CHARITE^TM Artificial Disc is generally not indicated in patients over 60 years old. Further, it stated that there is insufficient evidence among either the aged or disabled Medicare population to make a reasonable and necessary determination for coverage. With an NCD pending to make spinal arthroplasty with the CHARITE^TM Artificial Disc noncovered, we indicated in the FY 2007 IPPS proposed rule that we did not believe it was appropriate at that time to reassign procedure code 84.65 from DRGs 499 and 500 to DRGs 497 and 498.

After considering the public comments and additional evidence received, we made a final NCD on May 16, 2006, that Medicare would not cover the CHARITE^TM Artificial Disc for the Medicare population over 60 years of age. For Medicare beneficiaries 60 years of age and under, local Medicare contractors have the discretion to determine coverage for lumbar artificial disc replacement procedures involving the CHARITE^TM Artificial Disc. The final NCD can be found at: http://www.cms.hhs.gov/mcd/viewncd.asp:ncd_;id-150.10&ncd_version1&basket=ncd%3A150%2E10%3A1%3ALumbar+Artificial+Disc+Replacement%280ADR%29.

Comment: Some commenters agreed with our proposed decision not to reassign CHARITE^TM Artificial Disc at this time to the spinal fusion DRGs. Other commenters disagreed with our proposal not to move code 84.65 (CHARITE^TM) from DRGs 499 and 500 to DRGs 497 and 498. One commenter noted that the national noncoverage determination for the CHARITE^TM Artificial Disc only applies to patients over 60 years of age. The commenter further noted that local Medicare carriers have the discretion to make coverage decisions for Medicare beneficiaries who are under 60 years of age. The commenter stated that patients who receive the CHARITE^TM Artificial Disc are candidates for a fusion procedure involving an anterior surgical approach. The commenter goes on to state that the CHARITE^TM Artificial Disc is an alternative therapy to spinal fusion for patients with similar diagnoses. The commenter supplied data from FY 2005 MedPAR file in support of its request for a DRG change. These data included 54 cases that were assigned to DRGs 499 and 500. The 23 cases in DRG 499 had mean charges of $61,750, while the 31 cases assigned to DRG 500 had mean charges of $53,802. These data compare to mean charges of $26,974 for all cases in DRG 499 and $17,731 for all cases in DRG 500. The commenter reported mean charges of $71,581 for DRG 497 and $55,489 for DRG 498. The commenter stated that the 54 CHARITE^TM cases are more similar in average charges to all cases in DRGs 497 and 498 than to DRGs 499 and 500.

Response: We agree with the commenter that it is not appropriate to consider a DRG revision at this time for the CHARITE^TM Artificial Disc, given the recent decision to limit coverage for surgical procedures involving this device. Although we have reviewed the Medicare charge data, we are concerned that there are a very small number of cases for patients under 60 years of age who have received the CHARITE^TM Artificial Disc. We believe it appropriate to base the decision on a DRG change on charge data only on the population for which the procedure is covered. We have an extremely small number of cases for patients under 60 on which to base such a decision. For this reason, we do not believe it is appropriate to modify the DRGs at this time for CHARITE^TM cases.

5. MDC 18 (Infectious and Parasitic Diseases (Systemic or Unspecified Sites)): Severe Sepsis

In FYs 2005 and 2006, we considered requests for the creation of a separate DRG for the diagnosis of severe sepsis. Severe sepsis is described by ICD-9-CM code 995.92 (Systemic inflammatory response syndrome due to infection with organ dysfunction). Patients admitted with sepsis as a principal diagnosis currently are assigned to DRG 416 (Septicemia Age > 17) and DRG 417 (Septicemia Age 0-17) in MDC 18 (Infectious and Parasitic Diseases (Systemic or Unspecified Sites)). The commenter requested that all cases in which severe sepsis is present on admission, as well as those cases in which it develops after admission (which are currently classified elsewhere), be included in this new DRG. In both FY 2005 and FY 2006 (69 FR 48975 and 70 FR 47309), we did not believe the current clinical definition of severe sepsis was specific enough to identify a meaningful cohort of patients in terms of clinical coherence and resource utilization to warrant a separate DRG. Sepsis is found across hundreds of medical and surgical DRGs, and the term “organ dysfunction” implicates numerous currently existing diagnosis codes. While we recognize that Medicare beneficiaries with severe sepsis are quite ill and require extensive hospital resources, in the past we have not found that they can be identified adequately to justify removing them from all of the other DRGs in which they appear. For this reason, we did not create a new DRG for severe sepsis for FY 2005 or FY 2006. We indicated that we would continue to work with National Center for Health Statistics (NCHS) to improve the codes so that our data on these patients improve. We also indicated that we would continue to examine data on these patients as we consider future modifications.

For the FY 2007 IPPS proposed rule, we again received a request to consider creating a separate DRG for patients diagnosed with severe sepsis (71 FR 24037). The information and data available to us from hospital bills with respect to identifying patients with severe sepsis have not changed since last year. However, the NCHS discussed modifications to the current ICD-9-CM diagnosis codes for systemic inflammatory response syndrome (SIRS), codes 995.91 through 995.94 (which include severe sepsis) at the September 29-30, 2005 ICD-9-CM Coordination and Maintenance Committee meeting. During the meeting, it became clear that there is still confusion surrounding the use of these codes. As a result of the meeting and the comments received, the Committee made modifications to the set of SIRS codes. These modifications are reflected in Table 6E, Revised Diagnosis Code Titles, of the Addendum to this final rule.

We believe that implementation of the modified SIRS diagnosis codes and the updated coding guidelines over the next year could begin the process of improving data for this group of patients. The desired outcome is to be able to better evaluate Medicare beneficiaries with severe sepsis with regard to their clinical coherence, resource utilization, and charges. Therefore, in the FY 2007 IPPS proposed rule, we did not propose to create a new DRG for severe sepsis for FY 2007.

Comment: Numerous commenters asked for changes to the current sepsis classification. The commenters agreed that coding of systemic inflammatory response syndrome (SIRS), sepsis, septicemia, severe sepsis, and septic shock has been confusing to the provider community in the last few years. Specifically, one commenter stated coding guidelines have been revised based on clinical definitions, which in turn has affected the DRG classification for sepsis. Another commenter referenced the ICD-9-CM Code Book tabular section and the American Hospital Association's (AHA) fourth quarter (4Q) 2003 Coding Clinic, “for patients with severe sepsis, the code for the systemic infection (038.x) or trauma should be sequenced first, followed by either code 995.92 (Systemic Inflammatory Response Syndrome due to infectious process with organ dysfunction) or code 995.94 (Systemic inflammatory response syndrome due to noninfectious process with organ dysfunction). Codes for the specific organ dysfunction should also be assigned.” The commenter stated that as a result of this coding guideline, respiratory failure cannot be sequenced as the principal diagnosis because it is considered an organ dysfunction of the patient's sepsis. However, reverting sequencing instructions would be confusing and again disrupt the data according to some of the commenters. As a result, many commenters stated that a new DRG for severe sepsis is not appropriate due to the inconsistent data.

Response: We agree that there has been a great deal of confusion in the coding and sequencing of cases with severe sepsis and SIRS. The commenters are correct that the coding directives lead cases with severe sepsis that are on mechanical ventilation for respiratory failure to be assigned to DRG 416 (Septicemia Age >17) and DRG 417 (Septicemia Age 0 >17) instead of DRG 475 (Respiratory System Diagnosis with Ventilator Support). As stated in the proposed rule, we have continued to work with NCHS to improve the codes so that our data on these patients improve. We believe that implementation of the modified SIRS diagnosis codes and the updated coding guidelines over the next year will further improve the coding of this subset of patients.

Comment: One commenter presented its analysis of the MedPAR data and again requested the creation of two new DRGS for severe sepsis, one medical and one surgical. The other option suggested by the commenter was to split DRGs 415 and 416 into DRGs with and without severe sepsis cases. The commenter expressed concern that, while there has been some confusion over the use of the SIRS family of codes (995.90-995.94) over the past three years, the confusion has been mainly associated with the other codes and not the severe sepsis code (995.92). The commenter provided information concerning the definition of severe sepsis and its adoption following a 1992 consensus panel of the American College of Chest Physicians and the Society of Critical Care Medicine. According to the commenter, the panel defined severe sepsis as a systemic inflammatory response to infection that leads to acute organ dysfunction. The commenter noted this definition has been used successfully to identify thousands of patients with severe sepsis and in more than 30 large-scale clinical trials. The commenter also stated severe sepsis cases are clinically coherent with a common underlying problem (SIRS) leading to complications (acute organ dysfunction) and are managed similarly, receiving advanced life support in intensive care units. The commenter also provided examples to demonstrate how clinical coherence leads to resource use coherence.

Response: We appreciate the commenter's analysis of the data. As stated above, there has been significant confusion over the use of the sepsis codes. While the definition may be well understood among the individuals involved with the clinical trials, there has been uncertainty in the application of the codes as evidenced by repeated discussions at the ICD-9-CM Coordination and Maintenance Committee meetings and comments received in response to the proposed rule. We note that the National Center for Health Statistics has revised the sepsis and systemic inflammatory response syndrome codes in response to suggestions made at the Committee meetings. These revisions are shown in Table 6E of the Addendum to this final rule and will go into effect on October 1, 2006 (codes 995.91 through 995.94). We did not propose a new DRG for severe sepsis for FY 2007 in the proposed rule due to the data inconsistencies and difficulty expressed with properly assigning the sepsis codes, among other reasons cited previously.

In the FY 2007 IPPS proposed rule, we also solicited comments on the proposal we were considering to adopt a CS DRG system. We noted it is possible that the proposed system would better recognize the extensive resources that hospitals use to treat patients with severe sepsis. We encouraged commenters to examine the proposed system and provide comments. The comments and responses on this proposal are discussed in section II.C of this final rule.

Therefore, in this FY 2007 final rule we are not creating new DRGs for medical or surgical severe sepsis cases as requested by the commenter.

6. Medicare Code Editor (MCE) Changes

As explained under section II.B.1. of this preamble, the Medicare Code Editor (MCE) is a software program that detects and reports errors in the coding of Medicare claims data. Patient diagnoses, procedure(s), discharge status, and demographic information go into the Medicare claims processing systems and are subjected to a series of automated screens. The MCE screens are designed to identify cases that require further review before classification into a DRG.

For FY 2007, we proposed to make several changes to the MCE edits (71 FR 24038 and 24039). We received one comment on this topic. As a result of new and modified codes approved after the annual spring ICD-9-CM Coordination and Maintenance meeting, we make changes to the MCE. In the past, in both the IPPS proposed and final rules, we only provided the list of changes to the MCE in the IPPS that were brought to our attention after the prior year's final rule. We historically have not listed the changes we have made to the MCE as a result of the new and modified codes approved after the annual spring ICD-9-CM Coordination and Maintenance meeting. These changes are approved too late in the rulemaking schedule for inclusion in the proposed rule. Furthermore, although our MCE policies have been described in our proposed and final rules, we have not provided the detail of each new or modified diagnosis and procedure code edit in the final rule. However, in response to a public comment and in the interest of making the IPPS more transparent, we are including in this final rule a comprehensive list of all the changes to the MCE edits for the next fiscal year as a result of coding changes.

a. Edit: Newborn Diagnoses

We proposed to add code 780.92 (Excessive crying of infant (baby)) to the “Newborn Diagnoses” edit in the MCE. This edit is structured for patients with an age of “0”. In the Tabular portion of the ICD-9-CM diagnosis codes, the “excludes” note at code 780.92 states that this code “excludes excessive crying of child, adolescent or adult” and sends the coder to code 780.95 (Other excessive crying. (The new title of this code, shown on Table 6E of the Addendum to this final rule is “Excessive crying of child, adolescent, or adult”.) To make a conforming change, we also proposed that code 780.92 be removed from the “Pediatric Diagnoses—Age 0 Through 17” edit.

We did not receive any public comments on the proposed edit and, therefore, are adopting it as final.

In addition, there were diagnosis codes discussed at the March 2006 ICD-9-CM Coordination and Maintenance meeting that were approved too late in the rulemaking schedule for inclusion in the proposed rule. Therefore, the following ICD-9-CM diagnosis codes are added to the “Newborn Diagnosis” MCE edit for FY 2007:

• 768.7, Hypoxic-ischemic encephalopathy (HIE)
• 770.87, Respiratory arrest of newborn
• 770.88, Hypoxemia of newborn
• 775.81, Other acidosis of newborn
• 775.89, Other neonatal endocrine and metabolic disturbances
• 779.85, Cardiac arrest of newborn

Because diagnosis code 775.8 (Other transitory neonatal endocrine and metabolic disturbances) was expanded to the fifth-digit level, this code is being deleted from the Newborn Diagnosis edit.

b. Edit: Diagnoses for Pediatric—Age 0-17 Years Old

We are adding the following new diagnosis codes to the edit for diagnosis for pediatrics—age 0-17 years old:

• V85.51, Body Mass Index, pediatric, less than 5th percentile for age
• V85.52, Body Mass Index, pediatric, 5th percentile to less than 85th percentile for age
• V85.53, Body Mass Index, pediatric, 85th percentile to less than 95th percentile for age
• V85.54, Body Mass Index, pediatric, greater than or equal to 95th percentile for age

c. Edit: Maternity Diagnoses—Age 12 through 55

We are adding the following new codes to the edit for maternity diagnoses—age 12 through 55:

• 649.00, Tobacco use disorder complicating pregnancy, childbirth, or the puerperium, unspecified as to episode of care or not applicable
• 649.01, Tobacco use disorder complicating pregnancy, childbirth, or the puerperium, delivered, with or without mention of antepartum condition
• 649.02, Tobacco use disorder complicating pregnancy, childbirth, or the puerperium, delivered, with mention of postpartum complication
• 649.03, Tobacco use disorder complicating pregnancy, childbirth, or the puerperium, antepartum condition or complication
• 649.04, Tobacco use disorder complicating pregnancy, childbirth, or the puerperium, postpartum condition or complication
• 649.10, Obesity complicating pregnancy, childbirth, or the puerperium, unspecified as to episode of care or not applicable
• 649.11, Obesity complicating pregnancy, childbirth, or the puerperium, delivered, with or without mention of antepartum condition
• 649.12, Obesity complicating pregnancy, childbirth, or the puerperium, delivered, with mention of postpartum complication
• 649.13, Obesity complicating pregnancy, childbirth, or the puerperium, antepartum condition or complication
• 649.14, Obesity complicating pregnancy, childbirth, or the puerperium, postpartum condition or complication
• 649.20, Bariatric surgery status complicating pregnancy, childbirth, or the puerperium, unspecified as to episode of care or not applicable
• 649.21, Bariatric surgery status complicating pregnancy, childbirth, or the puerperium, delivered, with or without mention of antepartum condition
• 649.22, Bariatric surgery status complicating pregnancy, childbirth, or the puerperium, delivered, with mention of postpartum complication
• 649.23, Bariatric surgery status complicating pregnancy, childbirth, or the puerperium, antepartum condition or complication
• 649.24, Bariatric surgery status complicating pregnancy, childbirth, or the puerperium, postpartum condition or complication
• 649.30, Coagulation defects complicating pregnancy, childbirth, or the puerperium, unspecified as to episode of care or not applicable
• 649.31, Coagulation defects complicating pregnancy, childbirth, or the puerperium, delivered, with or without mention of antepartum condition
• 649.32, Coagulation defects complicating pregnancy, childbirth, or the puerperium, delivered, with mention of postpartum complication
• 649.33, Coagulation defects complicating pregnancy, childbirth, or the puerperium, antepartum condition or complication
• 649.34, Coagulation defects complicating pregnancy, childbirth, or the puerperium, postpartum condition or complication
• 649.40, Epilepsy complicating pregnancy, childbirth, or the puerperium, unspecified as to episode of care or not applicable
• 649.41, Epilepsy complicating pregnancy, childbirth, or the puerperium, delivered, with or without mention of antepartum condition
• 649.42, Epilepsy complicating pregnancy, childbirth, or the puerperium, delivered, with mention of postpartum complication
• 649.43, Epilepsy complicating pregnancy, childbirth, or the puerperium, antepartum condition or complication
• 649.44, Epilepsy complicating pregnancy, childbirth, or the puerperium, postpartum condition or complication
• 649.50, Spotting complicating pregnancy unspecified as to episode of care or not applicable
• 649.51, Spotting complicating pregnancy delivered, with or without mention of antepartum condition
• 649.53, Spotting complicating pregnancy antepartum condition or complication
• 649.60, Uterine size date discrepancy, unspecified as to episode of care or not applicable
• 649.61, Uterine size date discrepancy, delivered, with or without mention of antepartum condition
• 649.62, Uterine size date discrepancy, delivered, with mention of postpartum complication
• 649.63, Uterine size date discrepancy, antepartum condition or complication
• 649.64, Uterine size date discrepancy, postpartum condition or complication

d. Edit: Diagnoses Allowed for Females Only

The following codes are now invalid codes, as shown in Table 6C of the Addendum to the FY 2007 IPPS proposed rule and this final rule. In the FY 2007 IPPS proposed rule, we proposed to remove them from the “Diagnosis Allowed for Females Only” edit in the MCE.

• 616.8, Other specified inflammatory diseases of cervix, vagina, and vulva
• 629.8, Other specified disorders of female genital organs

Codes 616.8 and 629.8 have been expanded to the fifth-digit level. Therefore, we proposed to place the following expanded codes in the “Diagnoses Allowed for Females Only” edit.

• 616.81, Mucositis (ulcerative) of cervix, vagina, and vulva
• 616.89, Other inflammatory disease of cervix, vagina, and vulva
• 629.81, Habitual aborter without current pregnancy
• 629.89, Other specified disorders of female genital organs

The following two codes have revised descriptions (as shown in Table 6E of the Addendum to this final rule) which specify gender. Therefore, we proposed to add them to “Diagnoses Allowed for Females Only” edit.

• V26.31, Testing of female for genetic disease carrier status
• V26.32, Other genetic testing of female

We did not receive any public comments on the proposed changes to this edit. Therefore, we are adopting the changes as final.

In addition, we are adding the following new ICD-9-CM codes to this edit:

• 618.84, Cervical stump prolapse
• 629.29, Other female genital mutilation status
• 649.00, Tobacco use disorder complicating pregnancy, childbirth, or the puerperium, unspecified as to episode of care or not applicable
• 649.01, Tobacco use disorder complicating pregnancy, childbirth, or the puerperium, delivered, with or without mention of antepartum condition
• 649.02, Tobacco use disorder complicating pregnancy, childbirth, or the puerperium, delivered, with mention of postpartum complication
• 649.03, Tobacco use disorder complicating pregnancy, childbirth, or the puerperium, antepartum condition or complication
• 649.04, Tobacco use disorder complicating pregnancy, childbirth, or the puerperium, postpartum condition or complication
• 649.10, Obesity complicating pregnancy, childbirth, or the puerperium, unspecified as to episode of care or not applicable
• 649.11, Obesity complicating pregnancy, childbirth, or the puerperium, delivered, with or without mention of antepartum condition
• 649.12, Obesity complicating pregnancy, childbirth, or the puerperium, delivered, with mention of postpartum complication
• 649.13, Obesity complicating pregnancy, childbirth, or the puerperium, antepartum condition or complication
• 649.14, Obesity complicating pregnancy, childbirth, or the puerperium, postpartum condition or complication
• 649.20, Bariatric surgery status complicating pregnancy, childbirth, or the puerperium, unspecified as to episode of care or not applicable
• 649.21, Bariatric surgery status complicating pregnancy, childbirth, or the puerperium, delivered, with or without mention of antepartum condition
• 649.22, Bariatric surgery status complicating pregnancy, childbirth, or the puerperium, delivered, with mention of postpartum complication
• 649.23, Bariatric surgery status complicating pregnancy, childbirth, or the puerperium, antepartum condition or complication
• 649.24, Bariatric surgery status complicating pregnancy, childbirth, or the puerperium, postpartum condition or complication
• 649.30, Coagulation defects complicating pregnancy, childbirth, or the puerperium, unspecified as to episode of care or not applicable
• 649.31, Coagulation defects complicating pregnancy, childbirth, or the puerperium, delivered, with or without mention of antepartum condition
• 649.32, Coagulation defects complicating pregnancy, childbirth, or the puerperium, delivered, with mention of postpartum complication
• 649.33, Coagulation defects complicating pregnancy, childbirth, or the puerperium, antepartum condition or complication
• 649.34, Coagulation defects complicating pregnancy, childbirth, or the puerperium, postpartum condition or complication
• 649.40, Epilepsy complicating pregnancy, childbirth, or the puerperium, unspecified as to episode of care or not applicable
• 649.41, Epilepsy complicating pregnancy, childbirth, or the puerperium, delivered, with or without mention of antepartum condition
• 649.42, Epilepsy complicating pregnancy, childbirth, or the puerperium, delivered, with mention of postpartum complication
• 649.43, Epilepsy complicating pregnancy, childbirth, or the puerperium, antepartum condition or complication
• 649.44, Epilepsy complicating pregnancy, childbirth, or the puerperium, postpartum condition or complication
• 649.50, Spotting complicating pregnancy unspecified as to episode of care or not applicable
• 649.51, Spotting complicating pregnancy delivered, with or without mention of antepartum condition
• 649.53, Spotting complicating pregnancy antepartum condition or complication
• 649.60, Uterine size date discrepancy, unspecified as to episode of care or not applicable
• 649.61, Uterine size date discrepancy, delivered, with or without mention of antepartum condition
• 649.62, Uterine size date discrepancy, delivered, with mention of postpartum complication
• 649.63, Uterine size date discrepancy, antepartum condition or complication
• 649.64, Uterine size date discrepancy, postpartum condition or complication
• 795.06, Papanicolaou smear of cervix with cytologic evidence of malignancy
• 795.82, Elevated cancer antigen 125 [CA 125]

e. Edit: Diagnoses Allowed for Males Only

Code 608.2 (Torsion of testis) is now an invalid code (as shown in Table 6C of the Addendum to the proposed rule and this final rule). Therefore, we proposed to remove it from the “Diagnoses Allowed for Males Only” edit. This code has been expanded to the fifth-digit level. We proposed to place the following expanded codes in the “Diagnoses Allowed for Males Only” edit:

• 608.20, Torsion of testis, unspecified
• 608.21, Extravaginal torsion of spermatic cord
• 608.22 Intravaginal torsion of spermatic cord
• 608.23, Torsion of appendix testis
• 608.24, Torsion of appendix epididymis

The following codes have been created effective for FY 2007 and are gender specific. Therefore, we proposed to add them to the “Diagnosis Allowed for Males Only” edit.

• V26.34, Testing of male for genetic disease carrier status
• V26.35, Encounter for testing of male partner of habitual aborter
• V26.39, Other genetic testing of male

We did not receive any public comments on our proposed changes to this edit. Therefore, we are adopting the changes as final.

f. Edit: Procedures Allowed for Females Only

The following new codes are added to the list of female procedures:

• 68.41, Laparoscopic total abdominal hysterectomy
• 68.49, Other and unspecified total abdominal hysterectomy
• 68.61, Laparoscopic radical abdominal hysterectomy
• 68.69, Other and unspecified radical abdominal hysterectomy
• 68.71, Laparoscopic radical vaginal hysterectomy [LRVH]
• 68.79, Other and unspecified radical vaginal hysterectomy

In addition, the following codes were expanded to the fourth digit and, therefore, are removed from this edit:

• 68.4, Total abdominal hysterectomy
• 68.6, Radical abdominal hysterectomy
• 68.7, Radical vaginal hysterectomy

g. Edit: Manifestations Not Allowed as Principal Diagnosis

We proposed to add the following codes to the “Manifestations Not Allowed as Principal Diagnosis” edit in the MCE:

• 362.03, Nonproliferative diabetic retinopathy, NOS
• 362.04, Mild nonproliferative diabetic retinopathy
• 362.05, Moderate nonproliferative diabetic retinopathy
• 362.06, Severe nonproliferative diabetic retinopathy
• 362.07, Diabetic macular edema.

We did not receive any public comments concerning this proposed change. Therefore, we are adopting the above proposed changes as final.

In addition, we are adding the following new codes to this edit:

• 284.2, Myelophthisis
• 289.83, Myelofibrosis
• 323.01, Encephalitis and encephalomyelitis in viral diseases classified elsewhere
• 323.02, Myelitis in viral diseases classified elsewhere
• 323.41, Other encephalitis and encephalomyelitis due to infection classified elsewhere
• 323.42, Other myelitis due to infection classified elsewhere
• 323.61, Infectious acute disseminated encephalomyelitis (ADEM)
• 323.62, Other postinfectious encephalitis and encephalomyelitis
• 323.63, Postinfectious myelitis
• 323.71, Toxic encephalitis and encephalomyelitis
• 323.72, Toxic myelitis
• 341.21, Acute (transverse) myelitis in conditions classified elsewhere

The following codes have been expanded to the fifth-digit level of specificity, which results in making the four-digit code invalid. Therefore, these codes are removed from the manifestation edit:

• 323.0, Encephalitis in viral diseases classified elsewhere
• 323.4, Other encephalitis due to infection classified elsewhere
• 323.6, Postinfectious encephalitis
• 323.7, Toxic encephalitis

In the proposed rule, we had suggested we would remove code 525.10 (Acquired absence of teeth, unspecified) from this edit in the MCE. However, all codes in subcategory 525.1 (Loss of teeth due to trauma, extraction, or periodontal disease) are considered manifestation codes. Therefore, we are retracting this proposal, and are leaving code 525.10 in this edit.

h. Edit: Nonspecific Principal Diagnosis

We proposed to add the following codes to the “Nonspecific Principal Diagnosis” edit in the MCE:

• 255.10, Hyperaldosteronism, unspecified
• 323.9, Unspecified causes of encephalitis, myelitis, and encephalomyelitis
• 770.10, Fetal and newborn aspiration, unspecified.
• 780.31, Febrile convulsions (simple), unspecified

Codes 255.10, 323.9, and 780.31 appear on Table 6E, Revised Diagnosis Codes, and are being included in this edit because of their revised descriptions. Code 770.10 was inadvertently left off this list for FY 2006 when the code was created.

We did not receive any public comments on the proposed changes to this edit. Therefore, we are adopting the proposed changes as final. In addition, we are adding the following codes to this edit:

• 238.75, Myelodysplastic syndrome, unspecified
• 276.50, Volume depletion NOS
• 277.30, Amyloidosis, unspecified
• 288.00, Neutropenia, unspecified
• 288.50, Leukocytopenia, unspecified
• 288.60, Leukocytosis, unspecified
• 341.20, Acute (transverse) myelitis NOS
• 379.60, Inflammation (infection) of postprocedural bleb, unspecified
• 523.30, Aggressive periodontitis, unspecified
• 523.40, Chronic periodontitis, unspecified
• 525.60, Unspecified unsatisfactory restoration of tooth
• 528.00, Stomatitis and mucositis, unspecified
• 608.20, Torsion of testis, unspecified
• 649.00, Tobacco use disorder complicating pregnancy, childbirth, or the puerperium, unspecified as to episode of care or not applicable
• 649.10, Obesity complicating pregnancy, childbirth, or the puerperium, unspecified as to episode of care or not applicable
• 649.20, Bariatric surgery status complicating pregnancy, childbirth, or the puerperium, unspecified as to episode of care or not applicable
• 649.30, Coagulation defects complicating pregnancy, childbirth, or the puerperium, unspecified as to episode of care or not applicable
• 649.40, Epilepsy complicating pregnancy, childbirth, or the puerperium, unspecified as to episode of care or not applicable
• 649.50, Spotting complicating pregnancy, childbirth, or the puerperium, unspecified as to episode of care or not applicable
• 649.60, Uterine size date discrepancy, unspecified as to episode of care or not applicable
• 958.90, Compartment syndrome, unspecified
• 995.20, Unspecified adverse effect of unspecified drug, medicinal and biological substance
• 995.22, Unspecified adverse effect of anesthesia
• 995.23, Unspecified adverse effect of insulin
• 995.29, Unspecified adverse effect of other drug, medicinal and biological substance

We are removing the following codes from this edit:

• 362.03, Nonproliferative diabetic retinopathy NOS
• 525.10, Acquired absence of teeth, unspecified
• 793.9, Other nonspecific abnormal findings on radiological and other examinations of body structure

Comment: Two commenters suggested that the expanded code (793.99, Other nonspecific abnormal findings on radiological and other examinations of body structure) be added back into this edit.

Response: We will not act on those suggestions at this time, as we believe that code 739.9 should not originally have been in the edit as it is more like an “other” code than a “nonspecific” code.

i. Edit: Unacceptable Principal Diagnosis

Most V-codes describe an individual's health status, but these codes are not usually a current illness or injury. Therefore, most V-codes are included in the “Unacceptable Principal Diagnosis” edit. The following codes became invalid (as shown in Table 6C of the Addendum to the proposed rule and this final rule) for FY 2007, and we proposed to remove them from this edit:

• V18.5, Family history, digestive disorders
• V58.3, Attention to surgical dressings and sutures
• V72.1, Examination of ears and hearing

The following V-codes represent either fifth-digit extensions of the above codes, or new codes that were created effective October 1, 2006 (Table 6A of the Addendum to the proposed rule and this final rule). Therefore, we proposed to add the following codes to the “Unacceptable Principal Diagnosis” edit:

• V18.51, Family history, colonic polyps
• V18.59, Family history, other digestive disorders
• V26.34, Testing of male for genetic disease carrier status
• V26.35, Encounter for testing of male partner of habitual aborter
• V26.39, Other genetic testing of male
• V45.86, Bariatric surgery status
• V58.30, Encounter for change or removal of nonsurgical wound dressing
• V58.31, Encounter for change or removal of surgical wound dressing
• V58.32, Encounter for removal of sutures
• V72.11, Encounter for hearing examination following failed hearing screening
• V72.19, Other examination of ears and hearing
• V82.71, Screening for genetic disease carrier status
• V82.79, Other genetic screening
• V85.51, Body mass index, pediatric, less than 5th percentile for age
• V85.52, Body mass index, pediatric, 5th percentile to less than 85th percentile for age
• V85.53, Body mass index, pediatric, 85th percentile to less than 95th percentile for age
• V85.54, Body mass index, pediatric, greater than or equal to 95th percentile for age
• V86.0, Estrogen receptor positive status [ER+]
• V86.1, Estrogen receptor negative status [ER−]

We did not receive any public comments on these proposed edits. Therefore, we are adopting the proposed changes as final.

j. Edit: Nonspecific O.R. Procedures

We proposed to remove code 00.29 (Intravascular imaging unspecified vessel(s)) from the “Nonspecific O.R. Procedure” edit in the MCE. This code was erroneously placed in this edit; it is not considered an O.R. procedure.

We did not receive any public comments on these proposed edits. Therefore, we are adopting the proposed changes as final.

In addition, we are removing code 68.39 (Other subtotal abdominal hysterectomy) from this edit. Code 68.39 is not a nonspecific code, it is considered other, and was originally included in this edit in error.

k. Edit: Noncovered Procedures

Under the proposed changes to DRG 513 (Pancreas Transplant) under the Pre-MDCs described in section II.D.1. of the preamble of the FY 2007 IPPS proposed rule, a patient must have a history of medically uncontrollable, insulin-dependent diabetes mellitus, that is, Type I diabetes mellitus. Therefore, to conform the “Noncovered Procedures” Edit in the MCE to these proposed changes, we proposed to revise Diagnosis List 1 in this edit to include only the following codes:

• 250.01, Diabetes mellitus without mention of complication, type I [juvenile type], not stated as uncontrolled
• 250.03, Diabetes mellitus without mention of complication, type I [juvenile type], uncontrolled
• 250.11, Diabetes with ketoacidosis, type I [juvenile type], not stated as uncontrolled
• 250.13, Diabetes with ketoacidosis, type I [juvenile type], uncontrolled
• 250.21, Diabetes with hyperosmolarity, type I [juvenile type], not stated as uncontrolled
• 250.23, Diabetes with hyperosmolarity, type I [juvenile type], uncontrolled
• 250.31, Diabetes with other coma, type I [juvenile type], not stated as uncontrolled
• 250.33, Diabetes with other coma, type I [juvenile type], uncontrolled
• 250.41, Diabetes with renal manifestations, type I [juvenile type], not stated as uncontrolled
• 250.43, Diabetes with renal manifestations, type I [juvenile type], uncontrolled
• 250.51, Diabetes with ophthalmic manifestations, type I [juvenile type], not stated as uncontrolled
• 250.53, Diabetes with ophthalmic manifestations, type I [juvenile type], uncontrolled
• 250.61, Diabetes with neurological manifestations, type I [juvenile type], not stated as uncontrolled
• 250.63, Diabetes with neurological manifestations, type I [juvenile type], uncontrolled
• 250.71, Diabetes with peripheral circulatory disorders, type I [juvenile type], not stated as uncontrolled
• 250.73, Diabetes with peripheral circulatory disorders, type I [juvenile type], uncontrolled
• 250.81, Diabetes with other specified manifestations, type I [juvenile type], not stated as uncontrolled
• 250.83, Diabetes with other specified manifestations, type I [juvenile type], uncontrolled
• 250.91, Diabetes with unspecified complication, type I [juvenile type], not stated as uncontrolled
• 250.93, Diabetes with unspecified complication, type I [juvenile type], uncontrolled

In addition, we proposed to remove Diagnosis List 2 from the “Noncovered Procedures” edit, which is comprised of the following codes:

• 403.01, Hypertensive kidney disease, malignant, with chronic kidney disease
• 403.11, Hypertensive kidney disease, benign, with chronic kidney disease
• 403.91, Hypertensive kidney disease, unspecified, with chronic kidney disease
• 404.02, Hypertensive heart and kidney disease, malignant, with chronic kidney disease
• 404.03, Hypertensive heart and kidney disease, malignant, with heart failure and chronic kidney disease
• 404.12, Hypertensive heart and kidney disease, benign, with chronic kidney disease
• 404.13, Hypertensive heart and kidney disease, benign, with heart failure and chronic kidney disease
• 404.92, Hypertensive heart and kidney disease, unspecified, with chronic kidney disease
• 404.93, Hypertensive heart and kidney disease, unspecified, with heart failure and chronic kidney disease
• 585.1, Chronic kidney disease, Stage I
• 585.2, Chronic kidney disease, Stage II (mild)
• 585.3, Chronic kidney disease, Stage III (moderate)
• 585.4, Chronic kidney disease, Stage IV (severe)
• 585.5, Chronic kidney disease, Stage V
• 585.6, End stage renal disease
• 585.9, Chronic kidney disease, unspecified
• V42.0, Organ or tissue replaced by transplant, kidney
• V43.89, Organ or tissue replaced by other means, other organ or tissue, other

All of the comments we received regarding this proposal were favorable. Therefore, we are adopting the above changes as final.

Lumbar Artificial Disc: CMS has found that lumbar artificial disc replacement (LADR) with the Charite^TM lumbar artificial disc is not reasonable and necessary for the Medicare population over 60 years of age. Therefore, we issued a national noncoverage determination for LADR with the Charite^TM lumbar artificial disc for Medicare patients over 60 years of age. For Medicare beneficiaries 60 years of age and under, there is no national coverage determination, leaving such determinations to be made on a local basis. The coverage decision memo can be viewed on the CMS Web site at http://www.cms.hhs.gov/mcd/viewdecisionmemo.asp?id=170.

To conform to this decision, procedure code 84.65 (Insertion of total spinal disc prosthesis, lumbosacral) is put on the Non-Covered Procedure edit except when the patient is 60 years of age or less. The logic will be as follows:

84.65, Insertion of total spinal disc prosthesis, lumbosacral

AND

Age <=61

l. Edit: Bilateral Procedure

We proposed to remove the following codes from the Bilateral Procedure edit, as these are adjunct codes. They are not O.R. codes recognized by the GROUPER as procedures, and the edit was created in error last year.

• 00.74, Hip replacement bearing surface, metal-on-polyethylene
• 00.75, Hip replacement bearing surface, metal-on-metal
• 00.76, Hip replacement bearing surface, ceramic-on-ceramic

We did not receive any public comments on these proposed edits. Therefore, we are adopting the proposed changes as final.

In addition, we are deleting the following joint revision codes from this edit, as they should not have been added last year.

• 00.71, Revision of hip replacement, acetabular component
• 00.72, Revision of hip replacement, femoral component
• 00.73, Revision of hip replacement, acetabular liner and/or femoral head only
• 00.81, Revision of knee replacement, tibial component
• 00.82, Revision of knee replacement, femoral component
• 00.83, Revision of knee replacement, patellar component
• 00.84, Revision of total knee replacement, tibial insert (liner)
• 81.53, Revision of hip replacement not otherwise specified
• 81.55, Revision of knee replacement not otherwise specified

7. Surgical Hierarchies

Some inpatient stays entail multiple surgical procedures, each one of which, occurring by itself, could result in assignment of the case to a different DRG within the MDC to which the principal diagnosis is assigned. Therefore, it is necessary to have a decision rule within the GROUPER by which these cases are assigned to a single DRG. The surgical hierarchy, an ordering of surgical classes from most resource-intensive to least resource-intensive, performs that function. Application of this hierarchy ensures that cases involving multiple surgical procedures are assigned to the DRG associated with the most resource-intensive surgical class.

Because the relative resource intensity of surgical classes can shift as a function of DRG reclassification and recalibrations, we reviewed the surgical hierarchy of each MDC, as we have for previous reclassifications and recalibrations, to determine if the ordering of classes coincides with the intensity of resource utilization.

A surgical class can be composed of one or more DRGs. For example, in MDC 11, the surgical class “kidney transplant” consists of a single DRG (DRG 302) and the class “kidney, ureter and major bladder procedures” consists of three DRGs (DRGs 303, 304, and 305). Consequently, in many cases, the surgical hierarchy has an impact on more than one DRG. The methodology for determining the most resource-intensive surgical class involves weighting the average resources for each DRG by frequency to determine the weighted average resources for each surgical class. For example, assume surgical class A includes DRGs 1 and 2 and surgical class B includes DRGs 3, 4, and 5. Assume also that the average charge of DRG 1 is higher than that of DRG 3, but the average charges of DRGs 4 and 5 are higher than the average charge of DRG 2. To determine whether surgical class A should be higher or lower than surgical class B in the surgical hierarchy, we would weight the average charge of each DRG in the class by frequency (that is, by the number of cases in the DRG) to determine average resource consumption for the surgical class. The surgical classes would then be ordered from the class with the highest average resource utilization to that with the lowest, with the exception of “other O.R. procedures” as discussed below.

This methodology may occasionally result in assignment of a case involving multiple procedures to the lower-weighted DRG (in the highest, most resource-intensive surgical class) of the available alternatives. However, given that the logic underlying the surgical hierarchy provides that the GROUPER search for the procedure in the most resource-intensive surgical class, in cases involving multiple procedures, this result is sometimes unavoidable.

We note that, notwithstanding the foregoing discussion, there are a few instances when a surgical class with a lower average charge is ordered above a surgical class with a higher average charge. For example, the “other O.R. procedures” surgical class is uniformly ordered last in the surgical hierarchy of each MDC in which it occurs, regardless of the fact that the average charge for the DRG or DRGs in that surgical class may be higher than that for other surgical classes in the MDC. The “other O.R. procedures” class is a group of procedures that are only infrequently related to the diagnoses in the MDC, but are still occasionally performed on patients in the MDC with these diagnoses. Therefore, assignment to these surgical classes should only occur if no other surgical class more closely related to the diagnoses in the MDC is appropriate.

A second example occurs when the difference between the average charges for two surgical classes is very small. We have found that small differences generally do not warrant reordering of the hierarchy because, as a result of reassigning cases on the basis of the hierarchy change, the average charges are likely to shift such that the higher-ordered surgical class has a lower average charge than the class ordered below it.

Based on the preliminary recalibration of the DRGs, in the FY 2007 IPPS proposed rule (71 FR 24039), we proposed to revise the surgical hierarchy for Pre-MDCs, MDC 1 (Diseases and Disorders of the Nervous System), MDC 2 (Diseases and Disorders of the Eye), MDC 3 (Diseases and Disorders of the Ear, Nose, Mouth, and Throat), MDC 8 (Diseases and Disorders of the Musculoskeletal System and Connective Tissue), MDC 10 (Endocrine, Nutritional and Metabolic Diseases and Disorders), and MDC 13 (Diseases and Disorders of the Female Reproductive System) as follows:

In Pre-MDCs, we proposed to reorder DRG 481 (Bone Marrow Transplant) above DRG 513 (Pancreas Transplant).

In MDC 1, we proposed to reorder DRGs 531-532 (Spinal Procedures, With CC and Without CC, respectively) above DRGs 529-530 (Ventricular Shunt Procedures, With CC and Without CC, respectively).

In MDC 2, we proposed to reorder DRG 42 (Intraocular Procedures Except Retina, Iris and Lens) above DRG 36 (Retinal Procedures).

In MDC 3, we proposed to reorder DRGs 168-169 (Mouth Procedures, With CC and Without CC, respectively) above DRG 57 (T&A Procedures, Except Tonsillectomy and/or Adenoidectomy Only, Age > 17) and DRG 58 (T&A Procedures, Except Tonsillectomy and/or Adenoidectomy Only, Age 0-17).

In MDC 8, we proposed to reorder DRG 213 (Amputation for Musculoskeletal System and Connective Tissue Disorders) above DRG 216 (Biopsies of Musculoskeletal System and Connective Tissue).

In MDC 10, we proposed to reorder DRG 285 (Amputation of Lower Limb for Endocrine, Nutritional and Metabolic Diseases and Disorders) above DRG 288 (O.R. Procedures for Obesity).

In MDC 13, we proposed to reorder DRG 363 (D&C, Conization and Radio-Implant, for Malignancy) and DRG 364 (D&C, Conization and Radio-Implant, Except for Malignancy) above DRG 360 (Vagina, Cervix, and Vulva Procedures).

We did not receive any public comments on the proposed changes to the surgical hierarchy described above. Based on a test of the proposed revisions using the March 2006 update of the FY 2005 MedPAR file and the revised GROUPER software, we found that the revisions are still supported by the data. Therefore, we are incorporating these proposed revisions to the surgical hierarchy as final for FY 2007. In addition, because, in this final rule, we are deleting 8 DRGs and creating 20 new DRGs as discussed under section II.D.7. of this preamble, we are reordering the following DRGs in MDC 1 (Diseases and disorders of the Nervous System), MDC 6 (Diseases and Disorders of the Digestive System), MDC 11 (Diseases and Disorders of the Kidney and urinary Tract), and MDC 18 (Infectious and Parasitic Diseases (Systemic or Unspecified Sites)):

• In MDC 1, we are reordering DRG 577 (Carotid Artery Stent Procedure) above DRG 533 (Extracranial Procedures With CC).
• In MDC 6, we are reordering DRGs 567 and 568 (Stomach, Esophageal and Duodenal Procedures Age >17 With CC With and Without Major GI Diagnoses, respectively) above DRG 155 (Stomach, Esophageal and Duodenal Procedures Age >17 Without CC);
• In MDC 6, we are reordering DRGs 569-570 (Major Small and Large Bowel Procedures With CC With and Without Major GI Diagnoses, respectively) above DRG 149 (Major Small and Large Bowel Procedures Without CC).
• In MDC 11, we are reordering DRG 573 (Major Bladder Procedures) above DRG 303 (Kidney, Ureter and Major Bladder Procedures for Neoplasm).
• In MDC 18, we are reordering DRG 578 (Infections and Parasite Diseases With O.R. Procedure) above DRG 579 (Postoperative or Post-Traumatic Infections With O.R. Procedure).

8. Refinement of Complications and Comorbidities (CC) List

a. Background

As indicated earlier in this preamble, under the IPPS DRG classification system, we have developed a standard list of diagnoses that are considered complications or comorbidities (CCs). Historically, we developed this list using physician panels that classified each diagnosis code based on whether the diagnosis, when present as a secondary condition, would be considered a substantial complication or comorbidity. A substantial complication or comorbidity was defined as a condition that, because of its presence with a specific principal diagnosis, would cause an increase in the length of stay by at least 1 day in at least 75 percent of the patients.

b. Comprehensive Review of the CC List

In previous years, we have made changes to the standard list of CCs, either by adding new CCs or deleting CCs already on the list, but we have never conducted a comprehensive review of the list. Given the long period of time that had elapsed since the original CC list was developed, the incremental nature of changes to it, and changes in the way inpatient care is delivered, and in partial response to recommendations in MedPAC's March 2005 Report to Congress on Physician-Owned Specialty Hospitals, for the FY 2006 IPPS final rule, we reviewed the 121-paired DRGs that were split on the presence or absence of a CC among the 3,285 diagnosis codes on the CC list. We presented the results of that review and summarized public comments that we received in the FY 2006 proposed rule on the review results in the FY 2006 IPPS final rule (70 FR 47313 through 47315). Further analysis of the CC list and refinement to recognize the effects of differences in severity of illness among patients is discussed in section II.C. of the preamble of the proposed rule as part of our efforts to develop a CSDRG system for use in the IPPS.

During this past winter, CMS began a comprehensive review of over 13,000 diagnosis codes to determine whether they should be classified as CCs when present as a secondary diagnosis. Although we did not complete this review because of the work we did to develop the CS DRGs, we are considering whether to continue our analysis of the CC list as part of an effort to develop and adopt a severity DRG system that is in the public domain for FY 2008. As we explained in more detail above, we may update the work we did to develop a severity DRG system in the mid-1990s that classified patients into a base DRG that was further subdivided based on three levels of severity depending upon whether the patient had no CC, a CC, or a major CC in conjunction with continuing our review of the CC list.

c. CC Exclusions List for FY 2007

In the September 1, 1987 final notice (52 FR 33143) concerning changes to the DRG classification system, we modified the GROUPER logic so that certain diagnoses included on the standard list of CCs would not be considered valid CCs in combination with a particular principal diagnosis. We created the CC Exclusions List for the following reasons: (1) To preclude coding of CCs for closely related conditions; (2) to preclude duplicative or inconsistent coding from being treated as CCs; and (3) to ensure that cases are appropriately classified between the complicated and uncomplicated DRGs in a pair. As we indicated above, we developed a list of diagnoses, using physician panels, to include those diagnoses that, when present as a secondary condition, would be considered a substantial complication or comorbidity. In previous years, we have made changes to the list of CCs, either by adding new CCs or deleting CCs already on the list. We did not propose to delete any of the diagnosis codes on the CC list for FY 2007.

In the May 19, 1987 proposed notice (52 FR 18877) and the September 1, 1987 final notice (52 FR 33154), we explained that the excluded secondary diagnoses were established using the following five principles:

• Chronic and acute manifestations of the same condition should not be considered CCs for one another.
• Specific and nonspecific (that is, not otherwise specified (NOS)) diagnosis codes for the same condition should not be considered CCs for one another.
• Codes for the same condition that cannot coexist, such as partial/total, unilateral/bilateral, obstructed/unobstructed, and benign/malignant, should not be considered CCs for one another.
• Codes for the same condition in anatomically proximal sites should not be considered CCs for one another.
• Closely related conditions should not be considered CCs for one another.

The creation of the CC Exclusions List was a major project involving hundreds of codes. We have continued to review the remaining CCs to identify additional exclusions and to remove diagnoses from the master list that have been shown not to meet the definition of a CC.¹⁷

As we proposed, we are making limited revisions to the CC Exclusions List to take into account the changes that will be made in the ICD-9-CM diagnosis coding system effective October 1, 2006. (See section II.D.10. of this preamble for a discussion of ICD-9-CM changes.) We are making these changes in accordance with the principles established when we created the CC Exclusions List in 1987.

Tables 6G and 6H in the Addendum to this final rule contain the revisions to the CC Exclusions List that will be effective for discharges occurring on or after October 1, 2006. Each table shows the principal diagnoses with changes to the excluded CCs. Each of these principal diagnoses is shown with an asterisk, and the additions or deletions to the CC Exclusions List are provided in an indented column immediately following the affected principal diagnosis.

CCs that are added to the list are in Table 6G—Additions to the CC Exclusions List. Beginning with discharges on or after October 1, 2006, the indented diagnoses will not be recognized by the GROUPER as valid CCs for the asterisked principal diagnosis.

CCs that are deleted from the list are in Table 6H—Deletions from the CC Exclusions List. Beginning with discharges on or after October 1, 2006, the indented diagnoses will be recognized by the GROUPER as valid CCs for the asterisked principal diagnosis.

Copies of the original CC Exclusions List applicable to FY 1988 can be obtained from the National Technical Information Service (NTIS) of the Department of Commerce. It is available in hard copy for $152.50 plus shipping and handling. A request for the FY 1988 CC Exclusions List (which should include the identification accession number (PB) 88-133970) should be made to the following address: National Technical Information Service, United States Department of Commerce, 5285 Port Royal Road, Springfield, VA 22161; or by calling (800) 553-6847.

Users should be aware of the fact that all revisions to the CC Exclusions List (FYs 1989, 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2001, 2002, 2003, 2004, 2005, and 2006) and those in Tables 6G and 6H of this final rule for FY 2007 must be incorporated into the list purchased from NTIS in order to obtain the CC Exclusions List applicable for discharges occurring on or after October 1, 2006. (Note: There was no CC Exclusions List in FY 2000 because we did not make changes to the ICD-9-CM codes for FY 2000.)

Alternatively, the complete documentation of the GROUPER logic, including the current CC Exclusions List, is available from 3M/Health Information Systems (HIS), which, under contract with CMS, is responsible for updating and maintaining the GROUPER program. The current DRG Definitions Manual, Version 23.0, is available for $225.00, which includes $15.00 for shipping and handling. Version 24.0 of this manual, which will include the final FY 2007 DRG changes, will be available in hard copy for $250.00. Version 24.0 of the manual is also available on a CD for $200.00; a combination hard copy and CD is available for $400.00. These manuals may be obtained by writing 3M/HIS at the following address: 100 Barnes Road, Wallingford, CT 06492; or by calling (203) 949-0303. Please specify the revision or revisions requested.

9. Review of Procedure Codes in DRGs 468, 476, and 477

Each year, we review cases assigned to DRG 468 (Extensive O.R. Procedure Unrelated to Principal Diagnosis), DRG 476 (Prostatic O.R. Procedure Unrelated to Principal Diagnosis), and DRG 477 (Nonextensive O.R. Procedure Unrelated to Principal Diagnosis) to determine whether it would be appropriate to change the procedures assigned among these DRGs.

DRGs 468, 476, and 477 are reserved for those cases in which none of the O.R. procedures performed are related to the principal diagnosis. These DRGs are intended to capture atypical cases, that is, those cases not occurring with sufficient frequency to represent a distinct, recognizable clinical group. DRG 476 is assigned to those discharges in which one or more of the following prostatic procedures are performed and are unrelated to the principal diagnosis:

• 60.0, Incision of prostate
• 60.12, Open biopsy of prostate
• 60.15, Biopsy of periprostatic tissue
• 60.18, Other diagnostic procedures on prostate and periprostatic tissue
• 60.21,Transurethral prostatectomy
• 60.29, Other transurethral prostatectomy
• 60.61, Local excision of lesion of prostate
• 60.69, Prostatectomy, not elsewhere classified
• 60.81, Incision of periprostatic tissue
• 60.82, Excision of periprostatic tissue
• 60.93, Repair of prostate
• 60.94, Control of (postoperative) hemorrhage of prostate
• 60.95, Transurethral balloon dilation of the prostatic urethra
• 60.96, Transurethral destruction of prostate tissue by microwave thermotherapy
• 60.97, Other transurethral destruction of prostate tissue by other thermotherapy
• 60.99, Other operations on prostate

All remaining O.R. procedures are assigned to DRGs 468 and 477, with DRG 477 assigned to those discharges in which the only procedures performed are nonextensive procedures that are unrelated to the principal diagnosis.¹⁸

For FY 2007, we did not propose to change the procedures assigned among these DRGs. We did not receive any comments on our proposal and, therefore, are adopting it as final.

a. Moving Procedure Codes From DRG 468 or DRG 477 to MDCs

We annually conduct a review of procedures producing assignment to DRG 468 or DRG 477 on the basis of volume, by procedure, to see if it would be appropriate to move procedure codes out of these DRGs into one of the surgical DRGs for the MDC into which the principal diagnosis falls. The data are arrayed in two ways for comparison purposes. We look at a frequency count of each major operative procedure code. We also compare procedures across MDCs by volume of procedure codes within each MDC.

We identify those procedures occurring in conjunction with certain principal diagnoses with sufficient frequency to justify adding them to one of the surgical DRGs for the MDC in which the diagnosis falls. Based on this year's review, as proposed, we are not removing any procedures from DRG 477 with assignment to one of the surgical DRGs. We did not receive any comments on our proposal, and, therefore, there will be no change to DRG 477.

However, we did receive a comment regarding DRG 468 after the publication of the proposed rule. The comment addressed advances in treatment technology for hypertension and noted that two procedure codes cause cases to be assigned to DRG 468 instead of more appropriately to DRGs in MDC 5. Therefore, we are moving the following two codes into MDC 5, DRG 479 (Other Vascular Procedures without CC), and paired DRGs 553 and 554 (Other Vascular Procedures with CC with and without Major CV Diagnosis, respectively):

• 04.92, Implantation or replacement of peripheral neurostimulator lead(s)
• 86.96, Insertion or replacement of other neurostimulator pulse generator

b. Reassignment of Procedures Among DRGs 468, 476, and 477

We also annually review the list of ICD-9-CM procedures that, when in combination with their principal diagnosis code, result in assignment to DRGs 468, 476, and 477, to ascertain if any of those procedures should be reassigned from one of these three DRGs to another of the three DRGs based on average charges and the length of stay. We look at the data for trends such as shifts in treatment practice or reporting practice that would make the resulting DRG assignment illogical. If we find these shifts, we would propose to move cases to keep the DRGs clinically similar or to provide payment for the cases in a similar manner. Generally, we move only those procedures for which we have an adequate number of discharges to analyze the data.

We did not propose to move any procedure codes from DRG 476 to DRGs 468 or 477, or from DRG 477 to DRGs 468 or 476 for FY 2007. We did not receive any public comments on our proposal and; therefore, are adopting it as final.

c. Adding Diagnosis or Procedure Codes to MDCs

Based on our review this year, as we proposed, we are not adding any diagnosis codes to MDCs for FY 2007. We did not receive any public comments on our proposal and, therefore, are adopting it as final.

10. Changes to the ICD-9-CM Coding System

As described in section II.B.1. of this preamble, the ICD-9-CM is a coding system used for the reporting of diagnoses and procedures performed on a patient. In September 1985, the ICD-9-CM Coordination and Maintenance Committee was formed. This is a Federal interdepartmental committee, co-chaired by the National Center for Health Statistics (NCHS), the Centers for Disease Control and Prevention, and CMS, charged with maintaining and updating the ICD-9-CM system. The Committee is jointly responsible for approving coding changes, and developing errata, addenda, and other modifications to the ICD-9-CM to reflect newly developed procedures and technologies and newly identified diseases. The Committee is also responsible for promoting the use of Federal and non-Federal educational programs and other communication techniques with a view toward standardizing coding applications and upgrading the quality of the classification system.

The Official Version of the ICD-9-CM contains the list of valid diagnosis and procedure codes. (The Official Version of the ICD-9-CM is available from the Government Printing Office on CD-ROM for $25.00 by calling (202) 512-1800.) The Official Version of the ICD-9-CM is no longer available in printed manual form from the Federal Government; it is only available on CD-ROM. Users who need a paper version are referred to one of the many products available from publishing houses.

The NCHS has lead responsibility for the ICD-9-CM diagnosis codes included in the Tabular List and Alphabetic Index for Diseases, while CMS has lead responsibility for the ICD-9-CM procedure codes included in the Tabular List and Alphabetic Index for Procedures.

The Committee encourages participation in the above process by health-related organizations. In this regard, the Committee holds public meetings for discussion of educational issues and proposed coding changes. These meetings provide an opportunity for representatives of recognized organizations in the coding field, such as the American Health Information Management Association (AHIMA), the American Hospital Association (AHA), and various physician specialty groups, as well as individual physicians, health information management professionals, and other members of the public, to contribute ideas on coding matters. After considering the opinions expressed at the public meetings and in writing, the Committee formulates recommendations, which then must be approved by the agencies.

The Committee presented proposals for coding changes for implementation in FY 2007 at a public meeting held on September 29-30, 2005, and finalized the coding changes after consideration of comments received at the meetings and in writing by December 2, 2005. Those coding changes were announced in the FY 2007 IPPS proposed rule and are listed in Tables 6A through 6F in the Addendum to this final rule. The Committee held its 2006 meeting on March 23-24, 2006. Proposed new codes for which there was a consensus of public support and for which complete tabular and indexing changes can be made by May 2006 will be included in the October 1, 2006 update to ICD-9-CM. Code revisions that were discussed at the March 23-24, 2006 Committee meeting could not be finalized in time to include them in the FY 2007 IPPS proposed rule. These additional codes are included in Tables 6A through 6F of this final rule and are marked with an asterisk (*).

Copies of the minutes of the procedure codes discussions at the Committee's September 29-30, 2005 meeting can be obtained from the CMS Web site: http://cms.hhs.gov/ICD9ProviderDiagnosticCodes/03_meetings.asp. The minutes of the diagnosis codes discussions at the September 29-30, 2005 meeting are found at: http://www.cdc.gov/nchs/icd9.htm. Paper copies of these minutes are no longer available and the mailing list has been discontinued. These Web sites also provide detailed information about the Committee, including information on requesting a new code, attending a Committee meeting, and timeline requirements and meeting dates.

We encourage commenters to address suggestions on coding issues involving diagnosis codes to: Donna Pickett, Co-Chairperson, ICD-9-CM Coordination and Maintenance Committee, NCHS, Room 2402, 3311 Toledo Road, Hyattsville, MD 20782. Comments may be sent by E-mail to: dfp4@cdc.gov.

Questions and comments concerning the procedure codes should be addressed to: Patricia E. Brooks, Co-Chairperson, ICD-9-CM Coordination and Maintenance Committee, CMS, Center for Medicare Management, Hospital and Ambulatory Policy Group, Division of Acute Care, C4-08-06, 7500 Security Boulevard, Baltimore, MD 21244-1850. Comments may be sent by E-mail to: patricia.brooks2@cms.hhs.gov.

The ICD-9-CM code changes that have been approved will become effective October 1, 2006. The new ICD-9-CM codes are listed, along with their DRG classifications, in Tables 6A and 6B (New Diagnosis Codes and New Procedure Codes, respectively) in the Addendum to this proposed rule. As we stated above, the code numbers and their titles were presented for public comment at the ICD-9-CM Coordination and Maintenance Committee meetings. Both oral and written comments were considered before the codes were approved. In the FY 2007 IPPS proposed rule, we only solicited comments on the proposed classification of these new codes.

Comment: One commenter expressed concern about the DRG assignment for codes 629.81 (Habitual aborter without current pregnancy) and 629.89 (Other specified disorders of female genital organs). The commenter indicated that CMS proposed to assign both codes to DRG 368 (Infections, Female Reproductive System) within MDC-18. The commenter posited that CMS may have erred in listing the DRG assignment as DRG 368 and instead intended to assign the code to DRG 369 (Menstrual and Other Female Reproductive System Disorders) since these conditions are not infections.

Response: We agree with the commenter that codes 629.81 and 629.89 do not represent infections and should not be assigned to DRG 368 within MDC 18. They should instead be assigned to DRG 369 as the commenter suggested. Therefore, we are changing the DRG assignment for codes 629.81 and 629.89 from DRG 368 to DRG 369. This change is shown in Table 6A of the Addendum to this final rule.

Comment: One commenter asked whether the footnotes for codes 995.20 through 995.29 in Table 6A of the Addendum to the proposed rule was in error. The commenter stated that the predecessor code, 995.2 (Unspecified adverse effect of drug, medicinal and biological substance) is considered a secondary diagnosis of a “major problem” diagnosis that will assign a patient to DRGs 387 (Prematurity with Major Problems) and DRG 389 (Full-Term Neonate with Major Problems) when present only as a secondary diagnosis. However, the commenter added, the footnote on the expanded codes 995.20 through 995.29 lists them as principal or secondary diagnoses that will assign a patient to DRGs 387 and 389 for neonates with major problems. The specific codes are as follows:

• 995.20 (Unspecified adverse effect of unspecified drug, medicinal and biological substance)
• 995.21 (Arthus phenomenon)
• 995.22 (Unspecified adverse effect of anesthesia)
• 995.23 (Unspecified adverse effect of insulin)
• 995.27 (Other drug allergy)
• 995.29 (Unspecified adverse effect of other drug, medicinal and biological substance)

Response: The commenter is correct that we made an error in the footnote. The predecessor code 995.2 when present as a secondary diagnosis, will be a major problem that assigns the patient to DRGs 387 and 389. The footnote should have indicated codes 995.20 through 995.29 will only assign patients DRGs 387 and 389 when present as a secondary diagnosis. We have corrected the footnote in Table 6A of the Addendum to this final rule.

For codes that have been replaced by new or expanded codes, the corresponding new or expanded diagnosis codes are included in Table 6A. New procedure codes are shown in Table 6B. Diagnosis codes that have been replaced by expanded codes or other codes or have been deleted are in Table 6C (Invalid Diagnosis Codes). These invalid diagnosis codes will not be recognized by the GROUPER beginning with discharges occurring on or after October 1, 2006. Table 6D contains invalid procedure codes. These invalid procedure codes will not be recognized by the GROUPER beginning with discharges occurring on or after October 1, 2006. Revisions to diagnosis code titles are in Table 6E (Revised Diagnosis Code Titles), which also includes the DRG assignments for these revised codes. Table 6F includes revised procedure code titles for FY-2007.

In the September 7, 2001 final rule implementing the IPPS new technology add-on payments (66 FR 46906), we indicated we would attempt to include proposals for procedure codes that would describe new technology discussed and approved at the April meeting as part of the code revisions effective the following October. As stated previously, ICD-9-CM codes discussed at the March 23-24, 2006 Committee meeting that received consensus and that were finalized by May 2006, are included in Tables 6A through 6F of the Addendum to this final rule.

Section 503(a) of Pub. L. 108-173 included a requirement for updating ICD-9-CM codes twice a year instead of a single update on October 1 of each year. This requirement was included as part of the amendments to the Act relating to recognition of new technology under the IPPS. Section 503(a) amended section 1886(d)(5)(K) of the Act by adding a clause (vii) which states that the “Secretary shall provide for the addition of new diagnosis and procedure codes in April 1 of each year, but the addition of such codes shall not require the Secretary to adjust the payment (or diagnosis-related group classification) * * * until the fiscal year that begins after such date.” This requirement improves the recognition of new technologies under the IPPS system by providing information on these new technologies at an earlier date. Data will be available 6 months earlier than would be possible with updates occurring only once a year on October 1.

While section 1886(d)(5)(K)(vii) of the Act states that the addition of new diagnosis and procedure codes on April 1 of each year shall not require the Secretary to adjust the payment, or DRG classification, under section 1886(d) of the Act until the fiscal year that begins after such date, we have to update the DRG software and other systems in order to recognize and accept the new codes. We also publicize the code changes and the need for a mid-year systems update by providers to capture the new codes. Hospitals also have to obtain the new code books and encoder updates, and make other system changes in order to capture and report the new codes.

The ICD-9-CM Coordination and Maintenance Committee holds its meetings in the spring and fall in order to update the codes and the applicable payment and reporting systems by October 1 of each year. Items are placed on the agenda for the ICD-9-CM Coordination and Maintenance Committee meeting if the request is received at least 2 months prior to the meeting. This requirement allows time for staff to review and research the coding issues and prepare material for discussion at the meeting. It also allows time for the topic to be publicized in meeting announcements in the Federal Register as well as on the CMS Web site. The public decides whether or not to attend the meeting based on the topics listed on the agenda. Final decisions on code title revisions are currently made by March 1 so that these titles can be included in the IPPS proposed rule. A complete addendum describing details of all changes to ICD-9-CM, both tabular and index, is publicized on CMS and NCHS Web pages in May of each year. Publishers of coding books and software use this information to modify their products that are used by health care providers. This 5-month time period has proved to be necessary for hospitals and other providers to update their systems.

A discussion of this timeline and the need for changes are included in the December 4-5, 2005 ICD-9-CM Coordination and Maintenance Committee minutes. The public agreed that there was a need to hold the fall meetings earlier, in September or October, in order to meet the new implementation dates. The public provided comment that additional time would be needed to update hospital systems and obtain new code books and coding software. There was considerable concern expressed about the impact this new April update would have on providers.

In the FY 2005 IPPS final rule, we implemented section 1886(d)(5)(K)(vii) of the Act, as added by section 503(a) of Pub. L. 108-173, by developing a mechanism for approving, in time for the April update, diagnosis and procedure code revisions needed to describe new technologies and medical services for purposes of the new technology add-on payment process. We also established the following process for making these determinations. Topics considered during the Fall ICD-9-CM Coordination and Maintenance Committee meeting are considered for an April 1 update if a strong and convincing case is made by the requester at the Committee's public meeting. The request must identify the reason why a new code is needed in April for purposes of the new technology process. The participants at the meeting and those reviewing the Committee meeting summary report are provided the opportunity to comment on this expedited request. All other topics are considered for the October 1 update. Participants at the Committee meeting are encouraged to comment on all such requests. There were no requests for an expedited April 1, 2006 implementation of an ICD-9-CM code at the September 29-30, 2005 Committee meeting. Therefore, there were no new ICD-9-CM codes implemented on April 1, 2006.

We believe that this process captures the intent of section 1886(d)(5)(K)(vii) of the Act. This requirement was included in the provision revising the standards and process for recognizing new technology under the IPPS. In addition, the need for approval of new codes outside the existing cycle (October 1) arises most frequently and most acutely where the new codes will capture new technologies that are (or will be) under consideration for new technology add-on payments. Thus, we believe this provision was intended to expedite data collection through the assignment of new ICD-9-CM codes for new technologies seeking higher payments.

Current addendum and code title information is published on the CMS Web page at: www.cms.hhs.gov/icd9ProviderDiagnosticCodes/01_overview.asp#TopofPage. Information on ICD-9-CM diagnosis codes, along with the Official ICD-9-CM Coding Guidelines, can be found on the Web page at: www.cdc.gov/nchs/icd9.htm. Information on new, revised, and deleted ICD-9-CM codes is also provided to the AHA for publication in the Coding Clinic for ICD-9-CM. AHA also distributes information to publishers and software vendors.

CMS also sends copies of all ICD-9-CM coding changes to its contractors for use in updating their systems and providing education to providers.

These same means of disseminating information on new, revised, and deleted ICD-9-CM codes will be used to notify providers, publishers, software vendors, contractors, and others of any changes to the ICD-9-CM codes that are implemented in April. The code titles are adopted as part of the ICD-9-CM Coordination and Maintenance Committee process. Thus, although we publish the code titles in the IPPS proposed and final rules, they are not subject to comment in the proposed or final rules. We will continue to publish the October code updates in this manner within the IPPS proposed and final rules. For codes that are implemented in April, we will assign the new procedure code to the same DRG in which its predecessor code was assigned so there will be no DRG impact as far as DRG assignment. This mapping was specified by section 1886(d)(5)(K)(vii) of the Act as added by section 503(a) of Pub. L. 108-173. Any midyear coding updates will be available through the Web sites indicated above and through the Coding Clinic for ICD-9-CM. Publishers and software vendors currently obtain code changes through these sources in order to update their code books and software systems. We will strive to have the April 1 updates available through these Web sites 5 months prior to implementation (that is, early November of the previous year), as is the case for the October 1 updates.

Comment: Many commenters recommended that the Secretary use the regulatory process to replace ICD-9-CM with ICD-10-CM and ICD-10-PCS expeditiously. Several commenters indicated that the April 2005 ICD-9-CM Coordination and Maintenance Committee meeting included discussions of limiting the creation of new procedure codes in order to allow the classification system to last at least 2 more years. ICD-9-CM procedure code categories 00 and 17 were created to identify a diverse group of procedures and interventions affecting all body systems. The commenters stressed that the establishment of these code categories represented a deviation from the normal structure of ICD-9-CM and was a stopgap measure to accommodate new technology when there are no other codes available in the corresponding body system chapters (for example, musculoskeletal system and circulatory system). The commenters indicated that category 00 is now full, and the ICD-9-CM Coordination and Maintenance Committee is considering proposals for codes in category 17. The commenters stated that at the April Coordination and Maintenance meeting a proposal was presented that would in effect leave only 80 codes available in the new category 17. The commenters stated that in recent years, as many as 50 new procedure codes have been created in a single year. The commenters strongly recommended that the Secretary use the regulatory process to implement ICD-10-CM and ICD-10-PCS in place of ICD-9-CM expeditiously.

Several commenters indicated that limitations with ICD-9-CM make data collected with these codes less precise. The commenters stated that systems such as the CS DRGs could make use of the more detailed information in ICD-10-CM and ICD-10-PCS to group claims more accurately and better identify differences in severity and complexity. Similar comments were received from a number of other individuals.

Response: We agree that it is important to have an accurate and precise coding system. The Department will continue to study whether to adopt ICD-10-CM. In the interim, we continue to update both ICD-9-CM and ICD-10-PCS.

Comment: A number of commenters expressed concern that only nine diagnosis codes and six procedure codes are processed by Medicare. The commenters recommended that CMS modify its systems so that the number of diagnoses codes processed would increase from 9 to 25 and the number of procedures processed would increase from 6 to 25. The commenters stated that hospitals submit claims to CMS in electronic format, and that the HIPAA compliant electronic transaction standard, HIPAA 837i, allows up to 25 diagnoses and 25 procedures. The commenters stated that CMS does not require its fiscal intermediaries to process codes beyond the first nine diagnosis codes and six procedure codes. The commenters indicated that complex classification systems such as the proposed CS DRGs could make use of the information in these additional codes to better classify the patients.

One commenter stated that an incremental step in working towards a refined DRG system is to have CMS systems process 25 diagnosis and procedure codes.

Response: The commenters are correct that CMS does not process codes submitted electronically on the 837i electronic format beyond the first 9 diagnosis codes and first 6 procedure codes. While HIPAA requires CMS to accept up to 25 ICD-9-CM diagnosis and procedure codes on the HIPAA 837i electronic format, it does not require that CMS process that many diagnosis and procedure codes.

As suggested by the commenters, there is value in retaining additional data on patient conditions that would result from expanding Medicare's data system so it can accommodate additional diagnosis and procedure codes. We have been considering this issue while we contemplated refinements to our DRG system to better recognize patient severity of illness. However, extensive lead time is required to allow for modifications to our internal and contractors' electronic systems in order to process and store this additional information. We are unable to move forward with this recommendation without carefully evaluating implementation issues. We will continue to carefully evaluate this request to expand the process capacity of our systems.

Comment: One commenter expressed concern about the process involved with updating the ICD-9-CM Coding Guidelines. The guidelines are updated by the cooperating parties of ICD-9-CM, including representatives from the Centers for Disease and Prevention Control (CDC), CMS, the AHA, and the AHIMA. The commenter complimented CMS staff for becoming more “provider friendly” and using such tools as the open door forum to involve providers in policy discussions. The commenter requested that some of the coding guideline discussions be held in an open meeting so that providers could give input.

Response: We agree with the commenter that it is important to involve the provider community in activities involving the updating of ICD-9-CM codes and guidelines. The Department utilizes the ICD-9-CM Coordination and Maintenance Committee to discuss proposed changes to the coding system. At times, this Committee also addresses coding guidelines that affect code selection. The current process of approving new and revised coding guidelines involves approval by all four cooperating parties. It is our understanding that AHA and AHIMA actively seek input from their members on coding issues. AHA and AHIMA use this input when they are voting on coding issue to be published in the AHA's Coding Clinic for ICD-9-CM and on coding guidelines. We will refer these concerns to the cooperating parties so that they may discuss improvements which could be made in obtaining providers' input into coding guidelines. We will also welcome recommendations on specific coding guideline issues that providers wish to be included in future agendas of the ICD-9-CM Coordination and Maintenance Committee. The Committee recently discussed coding guidelines for septicemia. We will continue to work with the provider community to offer a public forum for discussion of ICD-9-CM code revisions and guidelines.

11. Other Issues

a. Chronic Kidney Disease

Comment: Two commenters expressed concern regarding the revised diagnosis codes for chronic kidney disease and their DRG assignments which appeared in Table 6E of the Addendum to the proposed rule. The following codes were identified as being classified to DRGs 331, 332, and 333 (Other kidney and urinary tract diagnoses with and without CC, and age 0-17, respectively) in MDC 11 (Diseases and Disorders of the Kidney and Urinary Tract):

• 403.00 (Hypertensive chronic kidney disease, malignant, with chronic kidney disease stage I through stage IV, or unspecified)
• 403.10 (Hypertensive chronic kidney disease, benign, with chronic kidney disease stage I through stage IV, or unspecified)
• 403.90 (Hypertensive chronic kidney disease, unspecified, with chronic kidney disease stage I through stage IV, or unspecified)

The commenters stated that revisions made to these three codes will go into effect October 1. These changes would add the concept of chronic kidney disease to the three codes. Therefore, these three codes should be assigned to the same DRGs as other codes for chronic kidney disease. The codes with chronic kidney disease are assigned to DRGs 315 (Other kidney and urinary tract procedures) and 316 (Renal failure) and not to DRGs 331 through 333 where they were proposed.

Response: The commenters are correct. The three codes listed above were modified to include the concept of chronic kidney disease. As such, they should be assigned to DRG 315 (Other Kidney and Urinary Tract Procedures) and DRG 316 (Renal Failure) (and not to DRGs 331 through 333. We have made these changes in Table 6E of the Addendum to this final rule. Therefore, we will assign codes 403.00, 403.10, and 403.90 to DRG 315-316.

b. Bronchial Valve

Comment: Two commenters that manufacture minimally invasive surgical therapies for patients with chronic obstructive pulmonary disease addressed the establishment of a new code for the insertion of a bronchial valve. This topic was discussed at the March 23-24, 2006 meeting of the ICD-9-CM Coordination and Maintenance Committee. (A complete summary report of the meeting including handouts can be found at: http://www.cms.hhs.gov/ICD9ProviderDiagnosticCodes/03_meetings.asp#TopofPage.) CMS created a new code for endoscopic insertion of a bronchial valve: code 33.71 (Endoscopic insertion or replacement of bronchial valve(s)). The new code is listed in Table 6B of the Addendum to this final rule. The predecessor codes that are currently used for this procedure are:

• 33.22, Fiber-optic bronchoscopy
• 96.05, Other intubation of respiratory tract

The commenters expressed support for the creation of the new code, but requested that the code not be assigned to the same DRG as its predecessor codes. The predecessor codes are assigned to a medical DRG if the patient is admitted with a respiratory diagnosis. If the patient is admitted with a history of malignancy, the patient would be assigned to DRG 412 (History of Malignancy with Endoscopy). The commenters requested that code 33.71 be assigned to DRG 75 (Major Chest Procedure). Although the commenters acknowledged that CMS has no data on which to evaluate this request, they recommended that CMS use a combination of the diagnosis of air leaks and treatment with scarification as a proxy for cases that receive a bronchial valve. The commenters stated that these patients are clinically similar and can be expected to have similar resource intensity to patients that would receive an endobronchial insertion or replacement of bronchial valves.

The commenters undertook their own data analysis using the FY 2005 MedPAR file. They used the following diagnosis procedure codes to identify the proxy patients:

• 512.0, Spontaneous tension pneumothorax
• 512.8, Other spontaneous pneumothorax
• 34.6, Scarification of the pleura

Using these codes, the companies identified 490 patients which were assigned to DRG 75. These patients had average charges of $56,711 as compared to $49,698 for all patients in DRG 75. The commenters stated that, although the resource utilization for scarification (and by inference, valve implantation) appears to be higher than the average for DRG 75, they believed it would still be reasonable to initially assign code 33.71 to DRG 75 until actual cost data can be gathered using the new procedure code.

Response: We do not agree that the endoscopic insertion of a bronchial valve is clinically similar to scarification of the pleura. The commenters themselves indicate that insertion of the bronchial valve is a minimally invasive procedure. Scarification of the pleura is a significantly invasive procedure. Furthermore, the bronchial valves are inserted into patients admitted with chronic obstructive pulmonary disease, not spontaneous pneumothorax. Therefore, we do not agree with using the pneumothorax diagnoses as a proxy for patients who will receive the bronchial valve.

The bronchial valve code 33.71 will go into effect on October 1, 2006. At this time, we have no information that suggests we should assign this new code to a DRG that is different than the predecessor codes. For this reason, we are classifying code 33.71, Endoscopic insertion or replacement of bronchial valve(s) as a nonoperating room procedure that will be assigned to DRG 412. This classification is listed in Table 6B of the Addendum to this final rule. Once we receive data using the new code, we will evaluate this issue further.

c. Female Reproductive System Reconstruction Procedures

Comment: One commenter recommended that CMS consider revising the current procedure code assignments for DRG 356 (Female Reproductive System Reconstructive Procedures) under MDC 13 (Diseases and Disorders of the Female Reproduction System) to better reflect the clinical coherence of those procedures that are specific to maintaining reproductive health. The proposal suggested by the commenter would distinguish procedures that are intended to ensure the reproductive function of a woman from urinary conditions that cause discomfort and emptying the bladder. The commenter suggested revising DRG 356 to limit it to procedures that are specific to maintaining reproductive health while creating four new DRGs that would be clinically similar for procedures performed to repair pelvic floor defects which cause urinary incontinence. The commenter stated these new DRGs would be timely with the procedure code proposal they are planning to present at the September 28-29, 2006 ICD-9-CM Coordination and Maintenance Committee meeting.

Response: We appreciate the commenter's recommendation to create four new DRGs in order to recognize the clinical coherence of procedures specific to maintaining reproductive health. There are two aspects to the commenter's proposal. The first part of the proposal would limit DRG 356 to procedures that are intended to maintain reproductive health. The second part of the commenter's proposal would create four new DRGs for repairing pelvic floor defects that create urinary incontinence. These four new DRGs would consist of two new DRG pairs (each split based on whether or not the patient has a CC) that would separate patients based on whether or not they had a graft procedure.

The commenter provided no data to support its proposal. Further, two of the four new DRGs being requested by the commenter would be based on new and revised procedure codes that have not yet been proposed or created. Therefore, we are unable to evaluate the request at this time. We may consider this proposal further in the future if the ICD-9-CM Coordination and Maintenance Committee creates the new codes being requested by the commenter and further data are made available for review.

d. Devices That Are Replaced Without Cost or Where Credit for a Replaced Device Is Furnished to the Hospital

In recent years, there have been several field actions and recalls with regard to failure of implantable cardiac defibrillators (ICDs) and pacemakers. In many of these cases, the manufacturers have offered replacement devices without cost to the hospital or credit for the device being replaced if the patient required a more expensive device. In some circumstances, manufacturers have also offered, through a warranty package, to pay specified amounts for unreimbursed expenses to persons who had replacement devices implanted. In addition, we believe that incidental device failures that are covered by manufacturer warranties occur routinely. While we understand that some device malfunctions may be inevitable as medical technology grows increasingly sophisticated, we believe that early recognition of problems would reduce the number of people with the potential to be adversely affected by these device problems. The medical community needs heightened and early awareness of patterns of device failures, voluntary field actions, and recalls so that it can take appropriate action to care for Medicare beneficiaries. Systematic efforts must be undertaken by all interested and involved parties, including manufacturers, insurers, and the medical community, to ensure that device problems are recognized and addressed as early as possible so that people's health is protected and high quality medical care is provided. We are taking several steps to assist in the early recognition and analysis of patterns of device problems to minimize the potential for harmful device-related effects on the health of Medicare patients and the public in general.

In recent years, CMS has recognized the importance of data collection as a condition of Medicare coverage for selected services. In 2005, CMS issued a National Coverage Determination (NCD) that expanded coverage of ICDs and also required registry participation when the devices were implanted for certain clinical indications. The NCD included this requirement in order to ensure that the care received by Medicare beneficiaries was reasonable and necessary and, therefore, appropriately reimbursed. Presently, the American College of Cardiology—National Cardiovascular Data Registry (ACC-NCDR) collects these data and maintains the registry.

In addition to ensuring appropriate payment of claims, collection, and ongoing analysis of ICD implantation, data can facilitate public health response in the event of future device recalls. The systematic recording of device manufacturer and model number can enhance patient and provider notification. Analysis of registry data may uncover patterns in complication rates (for example, device malfunction, device related infection, or early battery depletion) associated with particular devices that signify the need for a more specific investigation. Patterns found in registry data may identify problems earlier than the currently available mechanisms, which do not systematically collect such detailed information surrounding procedures.

We encourage the medical community to work to develop additional registries for implantable devices, so that timely and comprehensive information is available regarding devices, recipients of those devices, and their health status and outcomes. While participation in an ICD registry is required as a condition of coverage for ICD implantation for certain clinical conditions, we believe that the potential benefits of registries extend well beyond their application in Medicare's specific NCDs. As medical technology continues to advance swiftly, data collection regarding the short and long term outcomes of new technologies, and especially concerning implanted devices that may remain in the bodies of patients for their lifetimes, will be essential to the timely recognition of any specific problems and patterns of complications. This information will facilitate early interventions to mitigate harm and improve the quality and efficiency of health care services.

Moreover, data from registries may help further the development of high quality, evidence-based clinical practice guidelines for the care of patients who may receive device-intensive procedures. In turn, widespread use of evidence-based guidelines may reduce variation in medical practice, leading to improved personal and public health. Registry information may also contribute to the development of more comprehensive and refined quality metrics that may be used to systematically assess and then improve the safety and quality of health care. Such improvements in the quality of care that result in better personal health will require the sustained commitment of industry, payers, health care providers, and others towards that goal, along with excellent and open communication and rapid systemwide responses in a comprehensive effort to protect and enhance the health of the public. We look forward to further discussions with the public about new strategies to recognize device problems early and how to definitively address them, in order to minimize both the harmful health effects and increased health care costs that may result.

In addition, we believe that the routine identification of Medicare claims for certain device implantation procedures in situations where a payment adjustment is appropriate may enhance the medical community's recognition of device problems, potentially leading to more timely improvements in device technologies. This systematic approach, where hospitals identify and then appropriately report selected services when devices are replaced without cost to the hospital or with full credit to the hospital for the cost of the replaced device, should provide comprehensive information regarding the hospital experiences of Medicare patients with certain devices that are being replaced. Because Medicare patients are common recipients of implanted devices, this claims information may be particularly helpful in identifying patterns of device problems early in their natural history so that appropriate strategies to reduce future problems may be developed.

In addition to our concern for the public health, we also have a fiduciary responsibility to the Medicare Trust Fund to ensure that Medicare pays only for covered services. Therefore, we believe that we need to consider whether it is appropriate to reduce the Medicare payment in cases in which an implanted device is replaced without cost to the hospital or with full credit for the removed device. Such a proposal could cover certain devices for which credit for the replaced device is given or which are replaced as a result of or pursuant to a warranty, field action, voluntary recall, involuntary recall, and certain devices which are provided free of charge. It could provide for a reduction in the IPPS payment when we determine that the device is replaced without cost to the provider or beneficiary or when the provider receives full credit for the cost of a replaced device. We will need to develop a methodology to determine the amount of the reduction to the otherwise payable IPPS payment. We believe that this is appropriate because in these cases the full cost of the replaced device is not incurred and, therefore, we believe that an adjustment to the payment is necessary to remove the cost of the device.

E. Recalibration of DRG Weights

In the FY 2007 IPPS proposed rule, we proposed to change the DRG recalibration process methodology for FY 2007 to move to an HSRV weighting method as discussed in section II.C.2. of the preamble to the proposed rule (71 FR 24044). For FY 2006 and years prior, we have recalibrated the DRG weights based on charge data for Medicare discharges using the most current charge information available (for example, the FY 2005 MedPAR file would have been used for FY 2007). Our thorough analysis of the March 2005 MedPAC recommendations regarding refinement of the DRG system used for the IPPS (see discussion of the MedPAC recommendations in section II.C.2. of this preamble) has shown that using gross charges as a basis for setting the DRG weights has introduced bias into the weighting process. Specifically, hospitals that are systematically more expensive than others (that is, teaching hospitals and specialty hospitals) tend to treat certain cases more commonly than others, causing the weights for these cases to be artificially high. In addition, hospitals may mark up their charges for routine days, intensive care days, and various ancillary services by different percentages. This practice of differential markups among hospital cost centers may also introduce bias into the weights. For instance, we have observed that ancillary service cost centers generally have higher charge markups than routine services. Thus, the charge-based relative weight methodology may result in higher weights for DRGs that use more ancillary services relative to DRGs that use more routine services than would occur under a system where the weights are based on costs.

As discussed in section II.C.2. of the preamble of the proposed rule, based on our study of the MedPAC recommendations, we developed an alternative methodology for recalibrating the DRG weights. This proposed method is discussed in detail beginning on 71 FR 24044. The proposed method involved applying the HSRV methodology at the cost center level (HSRVcc) to remove the bias introduced by hospital characteristics (that is, teaching, disproportionate share, location, and size, among others) and then scaling the weights to costs using national cost center CCRs derived from cost report data. However, in response to comments discussed in section II.C.2 of this final rule, we have postponed the implementation of the HSRV methodology in order to further study its effects and have subsequently revised the methodology for setting relative weights based on cost. Further, we are adopting the cost relative weights under a 3-year transition period such that in FY 2007, year one of the transition, the relative weights will be a blend of 33 percent of the relative cost weight and 67 percent of the relative charge weight. In year two, the relative weights will be based on 67 percent of the relative cost weight and 33 percent of the relative charge weight and in year three, the relative weights will be 100 percent cost based.

In developing the final system of weights, we used two data sources: Claims data and cost report data. As in previous years, the claims data source is the MedPAR file. This file is based on fully coded diagnostic and procedure data for all Medicare inpatient hospital bills. The FY 2005 MedPAR data used in this proposed rule include discharges occurring on October 1, 2004, through September 30, 2005, based on bills received by CMS through March 31, 2006, from all hospitals subject to the IPPS and short-term acute care hospitals in Maryland (which are under a waiver from the IPPS under section 1814(b)(3) of the Act). The FY 2005 MedPAR file used in calculating the relative weights includes data for approximately 12,238,146 Medicare discharges. Discharges for Medicare beneficiaries enrolled in a Medicare+Choice managed care plan are excluded from this analysis. The data exclude CAHs, including hospitals that subsequently became CAHs after the period from which the data were taken. The second data source used in the cost relative weight methodology are the FY 2004 Medicare cost report data files from HCRIS, which represents the most recent full set of cost report data available. We used the March 31, 2006 update of the HCRIS cost report files for FY 2004 in setting the final relative cost based weights.

Because we are implementing the relative weights on a transitional basis it is necessary to calculate both charge based and cost based relative weights. The charge-based methodology used to calculate the DRG relative weights from the MedPAR data is the same methodology that was in place for FY 2006 and was applied as follows:

• To the extent possible, all the claims were regrouped using the FY 2007 DRG classification revisions discussed in section II.D. of this preamble.
• The transplant cases that were used to establish the relative weight for heart and heart-lung, liver and/or intestinal, and lung transplants (DRGs 103, 480, and 495) were limited to those Medicare-approved transplant centers that have cases in the FY 2005 MedPAR file. (Medicare coverage for heart, heart-lung, liver and/or intestinal, and lung transplants is limited to those facilities that have received approval from CMS as transplant centers.)
• Organ acquisition costs for kidney, heart, heart-lung, liver, lung, pancreas, and intestinal (or multivisceral organs) transplants continue to be paid on a reasonable cost basis. Because these acquisition costs are paid separately from the prospective payment rate, it was necessary to subtract the acquisition charges from the total charges on each transplant bill that showed acquisition charges before computing the average charge for the DRG and before eliminating statistical outliers.
• Total charges were standardized to remove the effects of differences in area wage levels, indirect medical education and disproportionate share payments, and, for hospitals in Alaska and Hawaii, the applicable cost-of-living adjustment.
• The average standardized charge per DRG was calculated by summing the standardized total charges for all cases in the DRG and dividing that amount by the number of cases classified in the DRG. A transfer case was counted as a fraction of a case based on the ratio of its transfer payment under the per diem payment methodology to the full DRG payment for non-transfer cases. That is, a transfer case receiving payment under the transfer methodology equal to half of what the case would receive as a non-transfer would be counted as 0.5 of a total case.
• Statistical outliers were eliminated by removing all cases that were beyond 3.0 standard deviations from the mean of the log distribution of both the charges per case and the charges per day for each DRG.
• The average charge for each DRG was then recomputed (excluding the statistical outliers) and divided by the national average standardized charge per case to determine the relative weight.

The new charge-based weights were then normalized by an adjustment factor of 1.49338 so that the average case weight after recalibration was equal to the average case weight before recalibration. This normalization adjustment is intended to ensure that recalibration by itself neither increases nor decreases total payments under the IPPS as required by section 1886(d)(4)(C)(iii) of the Act. We note that due to the decision in Bellevue Hosp. Center v. Leavitt, in which the Court of Appeals for the Second Circuit (the Court) ordered CMS to apply the occupational mix adjustment to 100 percent of the wage index effective for FY 2007 (see section III.C. of this final rule for more details of this Court decision), we are unable to finalize the FY 2007 wage index data at this time. Since we are relying on the wage index data as one of the standardizing factors that we use in calculating both the charge-based and the cost-based relative weights that are blended to set the FY 2007 transitional relative weights, we will recalculate the FY 2007 relative weights when the wage data becomes available and will publish these recalculated relative weights in a subsequent Federal Register notice prior to October 1, 2006.

The methodology we used to calculate the DRG cost-based weights from the FY 2005 MedPAR claims data and FY 2004 Medicare cost report data is as follows:

• To the extent possible, all the claims were regrouped using the FY 2007 DRG classification revisions discussed in section II.D. of this preamble.
• The transplant cases that were used to establish the relative weight for heart and heart-lung, liver and/or intestinal, and lung transplants (DRGs 103, 480, and 495) were limited to those Medicare-approved transplant centers that have cases in the FY 2005 MedPAR file. (Medicare coverage for heart, heart-lung, liver and/or intestinal, and lung transplants is limited to those facilities that have received approval from CMS as transplant centers.)
• Organ acquisition costs for kidney, heart, heart-lung, liver, lung, pancreas, and intestinal (or multivisceral organs) transplants continue to be paid on a reasonable cost basis. Because these acquisition costs are paid separately from the prospective payment rate, it is necessary to subtract the acquisition charges from the total charges on each transplant bill that showed acquisition charges before computing the average cost for each DRG and before eliminating statistical outliers.
• Claims with total charges or total length of stay less than or equal to zero were dropped. Claims that had an amount in the total charge field that differed by more than $10.00 from the sum of the routine day charges, intensive care charges, pharmacy charges, special equipment charges, therapy services charges, operating room charges, cardiology charges, laboratory charges, radiology charges, other service charges, labor and delivery charges, inhalation therapy charges and anesthesia charges were also dropped. At least 94 percent of the providers in the MedPAR file had charges for 10 of the 13 cost centers. Claims for providers that did not have charges greater than zero for at least 10 of the 13 cost centers were dropped.
• Statistical outliers were eliminated by removing all cases that were beyond 3.0 standard deviations from the mean of the log distribution of both the total charges per case and the total charges per day for each DRG.

Once the MedPAR data were trimmed and the statistical outliers were removed, the charges for each of the 13 cost groups for each claim were standardized to remove the effects of differences in area wage levels, indirect medical education and disproportionate share payments, and for hospitals in Alaska and Hawaii, the applicable cost-of-living adjustment. Charges were then summed by DRG for each of the 13 cost groups such that each DRG had 13 standardized charge totals. These charges were then adjusted to cost by applying the national average CCRs developed from the FY 2004 cost report data.

The 13 cost centers that we used in the DRG cost calculation are shown in the following table. In addition, the table shows the lines on the cost report that we used to create the national cost center CCRs that we used to adjust the DRG charges to cost:

We developed the national average CCRs as follows:

Taking the FY 2004 cost report data, we removed CAHs, Indian Health Service hospitals, all-inclusive rate hospitals, and cost reports that represented time periods of less than 1 year (365 days). We included hospitals located in Maryland as we are including their charges in our claims database. We then created CCRs for each provider for each cost center (see prior table for line items used in the calculations) and removed any CCRs that were greater than 10 or less than .01. In response to a comment from MedPAC discussed in section II.C.1. of this preamble, we normalized the departmental CCRs by dividing the CCR for each department by the total CCR for the hospital for the purpose of trimming the data. We then took the logs of all of the normalized cost center CCRs and removed any cost center CCRs where the log of the cost center CCR was greater or less than the mean log plus/minus 3 times the standard deviation for the log of that cost center CCR. In the proposed rule we had used a trim of 1.96 times the standard deviation. However, in response to comments as discussed in section II C. of this preamble, we have subsequently revised our trim to 3 standard deviations as commenters stated that this less stringent trim appropriately retains more providers in the database. Once the cost report data were trimmed, we calculated a Medicare specific CCR, again in response to a comment from MedPAC as discussed in section II.C. of this preamble. The Medicare specific CCR was determined by taking the Medicare charges for each line item from worksheet D Part 4 and deriving the Medicare specific costs by applying the hospital-specific departmental CCRs to the Medicare specific charges for each line item from worksheet D Part 4. Once each hospital's Medicare specific costs were established, we summed the total Medicare specific costs and divided by the sum of the total Medicare specific charges to produce national average, charge weighted CCRs. In the proposed rule, we used hospital-specific CCRs, but in response to comments as discussed in section II C. of this preamble, we have revised our methodology to use charge-weighted CCRs in establishing the national average CCRs.

After we multiplied the total charges for each DRG in each of the 13 cost centers by the corresponding national average CCR, we summed the 13 “costs” across each DRG to produce a total standardized cost for the DRG. The average standardized cost for each DRG was then computed as the total standardized cost for the DRG divided by the transfer adjusted case count for the DRG. The average cost for each DRG was then divided by the national average standardized cost per case to determine the relative weight.

The new cost-based weights were then normalized by an adjustment factor of 1.49338 so that the average case weight after recalibration was equal to the average case weight before recalibration. This normalization adjustment is intended to ensure that recalibration by itself neither increases nor decreases total payments under the IPPS as required by section 1886(d)(4)(C)(iii) of the Act.

When we recalibrated the DRG weights for previous years, we set a threshold of 10 cases as the minimum number of cases required to compute a reasonable weight. We used that same case threshold in recalibrating the DRG weights for FY 2007. Using the FY 2005 MedPAR data set, there are 40 DRGs that contain fewer than 10 cases. In FY 2006, we computed weights for low volume DRGs by adjusting the FY 2005 weights of these low volume DRGs by the percentage change in the average weight of the cases in other DRGs. Because we believe that we do not have sufficient MedPAR data to set accurate and stable HSRVcc weights for these low-volume DRGs, we proposed to assign them the weights of similar DRGs for which we have more complete data and solicited comment on this proposal. The crosswalk table we proposed is shown in the FY 2007 IPPS proposed rule (71 FR 24048).

Comment: One commenter stated that we should not assign weights based on other DRGs but should instead supplement our current data the data from other sources so that we can set weights for these DRGs based on actual cases.

Response: Because we are implementing cost based weights in a transition phase and because we intend to study the DRGs and relative weight methodologies during the coming year we have reconsidered our proposal to assign low volume DRGs the weights of other DRGS for FY 2007 and are reverting to our previous method of updating the prior year's weight for these DRGs by the percentage change in the average weight of the cases in the other DRGs. We may consider supplementing our MedPAR data with additional claims data in the future.

Section 1886(d)(4)(C)(iii) of the Act requires that, beginning with FY 1991, reclassification and recalibration changes be made in a manner that assures that the aggregate payments are neither greater than nor less than the aggregate payments that would have been made without the changes. Although normalization is intended to achieve this effect, equating the average case weight after recalibration to the average case weight before recalibration does not necessarily achieve budget neutrality with respect to aggregate payments to hospitals because payments to hospitals are affected by factors other than average case weight. Therefore, as we have done in past years, and as discussed in section II.A.4.a. of the Addendum to this final rule, we are making a budget neutrality adjustment to ensure that the requirement of section 1886(d)(4)(C)(iii) of the Act is met.

F. LTC-DRG Reclassifications and Relative Weights for LTCHs for FY 2007

1. Background

In the June 6, 2003 LTCH PPS final rule (68 FR 34122), we changed the LTCH PPS annual payment rate update cycle to be effective July 1 through June 30 instead of October 1 through September 30. In addition, because the patient classification system utilized under the LTCH PPS uses the same DRGs as those currently used under the IPPS for acute care hospitals, in that same final rule, we explained that the annual update of the long-term care diagnosis-related group (LTC-DRG) classifications and relative weights will continue to remain linked to the annual reclassification and recalibration of the DRGs used under the IPPS. In that same final rule, we specified that we will continue to update the LTC-DRG classifications and relative weights to be effective for discharges occurring on or after October 1 through September 30 each year. Furthermore, we stated that we will publish the annual update of the LTC-DRGs in the proposed and final rules for the IPPS.

In the past, the annual update to the IPPS DRGs has been based on the annual revisions to the ICD-9-CM codes and was effective each October 1. As discussed in the FY 2006 IPPS final rule (70 FR 47323 through 47341) and in the Rate Year (RY) 2007 LTCH PPS final rule (71 FR 27803 through 27809), with the implementation of section 503(a) of Pub. L. 108-173, there is the possibility that one feature of the GROUPER software program may be updated twice during a Federal fiscal year (October 1 and April 1) as required by the statute for the IPPS. Section 503(a) of Pub. L. 108-173 amended section 1886(d)(5)(K) of the Act by adding a new clause (vii) which states that “the Secretary shall provide for the addition of new diagnosis and procedure codes in [sic] April 1 of each year, but the addition of such codes shall not require the Secretary to adjust the payment (or diagnosis-related group classification) * * * until the fiscal year that begins after such date.” This requirement improves the recognition of new technologies under the IPPS by accounting for those ICD-9-CM codes in the MedPAR claims data at an earlier date. In implementing the statutory change, the agency has provided that ICD-9-CM diagnosis and procedure codes for new medical technology may be created and added to existing DRGs in the middle of the Federal fiscal year on April 1. However, this policy change will have no effect on the LTC-DRG relative weights, which will continue to be updated only once a year (October 1), nor will there be any impact on Medicare payments under the LTCH PPS. The use of the ICD-9-CM code set is also compliant with the current requirements of the Transactions and Code Sets Standards regulations at 45 CFR Parts 160 and 162, promulgated in accordance with the Health Insurance Portability and Accountability Act of 1996 (HIPAA), Pub. L. 104-191.

As we explained in the RY 2007 LTCH PPS final rule (71 FR 27805 through 27809), in the health care industry, historically annual changes to the ICD-9-CM codes were effective for discharges occurring on or after October 1 each year. Thus, the manual and electronic versions of the GROUPER software, which are based on the ICD-9-CM codes, were also revised annually and effective for discharges occurring on or after October 1 each year. As noted above, the patient classification system used under the LTCH PPS (LTC-DRGs) is based on the patient classification system used under the IPPS (CMS DRGs), which historically had been updated annually and is effective for discharges occurring on or after October 1 through September 30 each year. As also mentioned above, the ICD-9-CM coding update process was revised as a result of implementing section 503(a) of Pub. L. 108-173, which includes a requirement for updating diagnosis and procedure codes as often as twice a year instead of the current process of annual updates on October 1 of each year (as discussed in greater detail in section II.D.10. of the preamble of this final rule). The agency uses the ICD-9-CM codes as its code set for diagnoses and procedures. Therefore, the ICD-9-CM codes currently used under both the IPPS and LTCH PPS may be updated as often as twice a year. This requirement is included as part of the amendments to the Act relating to recognition of new medical technology under the IPPS.

Despite the fact that aspects of the GROUPER software may be updated to recognize any new technology ICD-9-CM codes, as discussed most recently in the RY 2007 LTCH PPS final rule (71 FR 27805 through 27808), there will be no impact on either LTC-DRG assignments or payments under the LTCH PPS at that time. That is, changes to the LTC-DRGs (such as the creation or deletion of LTC-DRGs) and the relative weights will continue to be updated in the manner and timing (October 1) as they are now. As noted above and as described in the RY 2007 LTCH PPS final rule (71 FR 27805 through 27809), updates to the GROUPER for both the IPPS and the LTCH PPS (with respect to relative weights and the creation or deletion of DRGs) are made in the annual IPPS proposed and final rules and are effective each October 1. We also explained that because we do not publish a midyear IPPS rule, any April 1 code updates will not be published in a midyear IPPS rule. Rather, we will assign any new diagnosis or procedure codes to the same DRG in which its predecessor code was assigned, so that there will be no impact on the DRG assignments (as also discussed in section II.D.10. of this preamble). Any coding updates will be available through the Web sites provided in section II.D.10. of this preamble and through the Coding Clinic for ICD-9-CM. Publishers and software vendors currently obtain code changes through these sources in order to update their code books and software system. If new codes are implemented on April 1, revised code books and software systems, including the GROUPER software program, will be necessary because we must use current ICD-9-CM codes. Therefore, for purposes of the LTCH PPS, because each ICD-9-CM code must be included in the GROUPER algorithm to classify each case into a LTC-DRG, the GROUPER software program used under the LTCH PPS would need to be revised to accommodate any new codes.

In implementing section 503(a) of Pub. L. 108-173, there will only be an April 1 update if new technology codes are requested and approved. We note that any new codes created for April 1 implementation will be limited to those diagnosis and procedure code revisions primarily needed to describe new technologies and medical services. However, we reiterate that the process of discussing updates to the ICD-9-CM has been an open process through the ICD-9-CM Coordination and Maintenance Committee since 1995. Requestors will be given the opportunity to present the merits for a new code and make a clear and convincing case for the need to update ICD-9-CM codes for purposes of the IPPS new technology add-on payment process through an April 1 update (as also discussed in section II.D.10. of this preamble).

However, as we discussed in the RY 2007 LTCH PPS final rule (71 FR 27805 through 27809), at the September 29-30, 2005 ICD-9-CM Coordination and Maintenance Committee meeting, there were no requests for an April 1, 2006 implementation of ICD-9-CM codes, and, therefore, the next update to the ICD-9-CM coding system would not occur until October 1, 2006 (FY 2007). Presently, as there were no coding changes suggested for an April 1, 2006 update, the ICD-9-CM coding set implemented on October 1, 2005, will continue through September 30, 2006 (FY 2006). The update to the ICD-9-CM coding system for FY 2007 is discussed above in section II.D.10. of this preamble. Accordingly, in this final rule, as discussed in greater detail below, we are revising the LTC-DRG classifications and relative weights, to be effective October 1,2006 through September 30, 2007 (FY 2007). Furthermore, we will notify LTCHs of any revisions to the GROUPER software used under the IPPS and the LTCH PPS that will be implemented April 1, 2007. The LTC-DRGs and relative weights for FY 2007 in this final rule are based on the IPPS DRGs (GROUPER Version 24.0) discussed in section II.B. of the preamble to this final rule.

Comment: Two commenters urged us to consolidate rulemaking for the LTCH PPS into one annual cycle rather than setting the payment rates and policy changes on a July 1 through June 30 rate year but making changes to the LTC-DRGs and relative weights based on the Federal fiscal year, October 1 through September 30. Both commenters noted that this situation has caused management and planning difficulty for some LTCHs. One of the commenters, whose LTCH has a June 1 through May 31 fiscal year, emphasizes the difficulties in “estimating the impact of changes in case weights as part of the final rule” associated with the hospital IPPS.

One commenter noted that other Medicare provider types only experience one routine annual adjustment to their respective PPSs and that it is not reasonable to expect the LTCH provider community to comment on the reasonableness of a proposed payment level in February when “that payment level is subject to change in a second rulemaking proposed in April or May of the same year.” This commenter suggested that, commencing with FY 2008, all LTCH PPS rulemaking should occur on the same schedule as it does under the IPPS, which would maintain the established cycle for the update of the LTC-DRGs and relative weights. The same commenter further suggested that, should CMS make this change in the rulemaking schedule, for the first year only, CMS should establish a 3-month (July through September) and 12-month (October through September) update factor to the Federal rate.

Response: In the LTCH RY 2004 final rule (68 FR 34122), we revised our regulations at § 412.535, which established a LTCH PPS rate year with a July 1 effective date for the annual update of the Federal payment rate and associated payment policies while also maintaining an October 1 implementation date for the update of the LTC-DRG patient classification system and associated weighting factors. In changing the effective date of the annual LTCH PPS rate year update and the resulting publication dates of the proposed and final regulations for the LTCH PPS, we stated that this shift in the schedule would promote “administrative feasibility and efficiency” by avoiding concurrent rulemaking and publication with the IPPS final rule. We also noted that although section 1886(e)(5)(A) of the Act required that, for the IPPS, the proposed rule be published in the Federal Register “not later than the April 1 before each fiscal year; and the final rule, not later than the August 1 before such fiscal year,” no similar requirement is imposed on the LTCH PPS and that we believed that this schedule change was well within the considerable discretion that Congress afforded the Secretary in the implementation of the LTCH PPS (68 FR 34125 through 34128). We maintained at that time, and we continue to believe, that this change to the LTCH rate year annual rulemaking schedule was not unduly burdensome for the LTCH industry because we had not added any requirements that LTCHs maintain payment systems or coding software in order to be paid under the LTCH PPS, although we understood that it was common for many hospitals, consultants, and industry associations to do so.

With regard to the commenter who described a LTCH with a fiscal year beginning on June 1, we would also reiterate what we stated in the FY 2004 final rule that “since the start of cost reporting periods for many LTCHs, as well as acute care hospitals, have not generally coincided with the October starting date of the Federal fiscal year, those hospitals that choose to have their own payment software are very familiar with the virtually seamless routine of inputting new numbers to their existing systems when a final rule is published” (68 FR 34127).

Therefore, we continue to believe that there is no significant administrative burden imposed on the LTCH industry by the establishment of the July 1 through June 30 rate year for the annual payment rate update under the LTCH PPS while still maintaining the October 1 through September 30 update of the LTC-DRGs and relative weights which are linked to the annual update of the diagnosis and procedure code set (ICD-9-CM) currently adopted by the DHHS and the IPPS DRGs and relative weights.

However, two commenters also stated that the separate rule-making cycles cause difficulty in “estimating the impact of changes in case weights,” which will be published in April or May, when commenting on the payment rates published in the LTCH PPS proposed rule in the preceding January or February. From the volume of correspondence that we receive from LTCH associations and their consultants, some of which include detailed analyses of CMS data, we do not believe that our annual publication in the IPPS proposed rule of the proposed updates of the LTC-DRGs and corresponding relative weights (which are derived solely from the best available LTCH MedPAR claims data) prohibits the public from assessing the impact such proposed changes would have if finalized. In fact, in their specific comments on the proposed FY 2007 LTC-DRG relative weights (discussed in greater detail below), several commenters presented analyses of the combined effect of the policy changes established in the RY 2007 LTCH PPS final rule, effective July 1, 2007 (for example, revisions to the short-stay outlier policy), and the proposed changes to the LTC-DRGs and relative weights for FY 2007. Furthermore, the comments received on the policies presented in the LTCH PPS RY 2007 proposed rule, a number of which contained detailed data evaluations, demonstrated the availability as well as the ability of the public to analyze the proposed policy changes using the most recent LTCH MedPAR claims data. Therefore, we do not believe that our present publication schedule deprives industry stakeholders of the opportunity to submit meaningful comments on proposed changes to payment levels when we are establishing the payment rates and associated policy under the LTCH PPS, even though changes to the LTC-DRG weights are proposed in a separate notice of proposed rulemaking.

Given the considerable discretion granted to the Secretary under the BBRA of 1999 and the BIPA of 2000 to develop the LTCH PPS, we may revisit the rulemaking schedule for the LTCH PPS in the future. If a revision to the schedule is proposed, the public will have the opportunity to submit comments on any proposed change to the schedule during the rulemaking process.

2. Changes in the LTC-DRG Classifications

a. Background

Section 123 of Pub. L. 106-113 specifically requires that the agency implement a PPS for LTCHs that is a per discharge system with a DRG-based patient classification system reflecting the differences in patient resources and costs in LTCHs. Section 307(b)(1) of Pub. L. 106-554 modified the requirements of section 123 of Pub. L. 106-113 by specifically requiring that the Secretary examine “the feasibility and the impact of basing payment under such a system [the LTCH PPS] on the use of existing (or refined) hospital diagnosis-related groups (DRGs) that have been modified to account for different resource use of long-term care hospital patients as well as the use of the most recently available hospital discharge data.”

In accordance with section 307(b)(1) of Pub. L. 106-554 and § 412.515 of our existing regulations, the LTCH PPS uses information from LTCH patient records to classify patient cases into distinct LTC-DRGs based on clinical characteristics and expected resource needs. The LTC-DRGs used as the patient classification component of the LTCH PPS correspond to the DRGs under the IPPS for acute care hospitals. Thus, in this final rule, we are using the IPPS GROUPER Version 24.0 for FY 2007 to process LTCH PPS claims for LTCH discharges occurring from October 1, 2006, through September 30, 2007. The changes to the CMS-DRG classification system used under the IPPS for FY 2007 (GROUPER Version 24.0) are discussed in section II.D. of the preamble to this final rule.

We note that, as we discuss in section II.C.6. of the preamble to this final rule, MedPAC, in its 2005 Report to Congress on Physician-Owned Specialty Hospitals, recommended that CMS, among other things, refine the current DRGs under the IPPS to more fully capture differences in severity of illness among patients. As we also discuss in that same section, in evaluating the MedPAC recommendation for the IPPS, we are evaluating the APR DRG GROUPER used by MedPAC in its analysis. Based on this analysis, we concur with MedPAC that the modified version of the APR DRGs would account more completely for differences in severity of illness and associated costs among hospitals. However, as we made clear in the proposed rule and reiterate in section II.C.6. of the preamble of this final rule, there are still further changes that are important to make to the CS DRG system before it is ready for adoption. At this time, we are not adopting a new severity-adjusted DRG system, such as the APR DRGs or a modified version of the APR DRGs, under the IPPS, as discussed in greater detail in section II.C.6. of the preamble of this final rule. However, we are refining the current CMS-DRG system by creating 20 new CMS DRGs and modifying 32 others across 13 different clinical areas involving 1,666,476 cases that would improve the CMS DRG system's recognition of severity of illness for FY 2007. We note that the LTCH PPS uses the same patient classification system (DRGs) as the IPPS. That is, the patient classification system used under the LTCH PPS (LTC DRGs) is based on the patient classification system used under the IPPS (CMS DRGs), which historically had been updated annually and is effective for discharges occurring on or after October 1 through September 30 each year. As such, the updates to the CMS DRG classification system used under the IPPS for FY 2007 (GROUPER Version 24.0), discussed in section II.D. of the preamble to this final rule, will also be updates that apply under the LTCH PPS.

Under the LTCH PPS, we determine relative weights for each of the DRGs to account for the difference in resource use by patients exhibiting the case complexity and multiple medical problems characteristic of LTCH patients. In a departure from the IPPS, as we discussed in the August 30, 2002 LTCH PPS final rule (67 FR 55985), which implemented the LTCH PPS, and the FY 2006 IPPS final rule (70 FR 47324), we use low-volume quintiles in determining the LTC-DRG relative weights for LTC-DRGs with less than 25 LTCH cases, because LTCHs do not typically treat the full range of diagnoses as do acute care hospitals. Specifically, we group those low-volume LTC-DRGs (that is, LTC-DRGs with fewer than 25 cases) into 5 quintiles based on average charge per discharge. (A listing of the composition of low-volume quintiles for the FY 2006 LTC-DRGs (based on FY 2004 MedPAR data) appears in section II.G.3. of the FY 2006 IPPS final rule (70 FR 47325 through 47332).) We also adjust for cases in which the stay at the LTCH is less than or equal to five-sixths of the geometric average length of stay; that is, short-stay outlier cases (§ 412.529), as discussed below in section II.F.4. of this preamble.

b. Patient Classifications Into DRGs

Generally, under the LTCH PPS, Medicare payment is made at a predetermined specific rate for each discharge; that is, payment varies by the LTC-DRG to which a beneficiary's stay is assigned. Just as cases are classified into DRGs for acute care hospitals under the IPPS (see section II.B. of this preamble), cases are classified into LTC-DRGs for payment under the LTCH PPS based on the principal diagnosis, up to eight additional diagnoses, and up to six procedures performed during the stay, as well as age, sex, and discharge status of the patient. The diagnosis and procedure information is reported by the hospital using the ICD-9-CM codes.

As discussed in section II.B. of this preamble, the CMS-DRGs are organized into 25 major diagnostic categories (MDCs), most of which are based on a particular organ system of the body; the remainder involve multiple organ systems (such as MDC 22, Burns). Accordingly, the principal diagnosis determines MDC assignment. Within most MDCs, cases are then divided into surgical DRGs and medical DRGs. Some surgical and medical DRGs are further differentiated based on the presence or absence of CCs. (See section II.B. of this preamble for further discussion of surgical DRGs and medical DRGs.)

Because the assignment of a case to a particular LTC-DRG will determine the amount that is paid for the case, it is important that the coding is accurate. As used under the IPPS, classifications and terminology used under the LTCH PPS are consistent with the ICD-9-CM and the Uniform Hospital Discharge Data Set (UHDDS), as recommended to the Secretary by the National Committee on Vital and Health Statistics (“Uniform Hospital Discharge Data: Minimum Data Set, National Center for Health Statistics, April 1980”) and as revised in 1984 by the Health Information Policy Council (HIPC) of the U.S. Department of Health and Human Services. We point out again that the ICD-9-CM coding terminology and the definitions of principal and other diagnoses of the UHDDS are consistent with the requirements of the Transactions and Code Sets Standards under HIPAA (45 CFR Parts 160 and 162).

The emphasis on the need for proper coding cannot be overstated. Inappropriate coding of cases can adversely affect the uniformity of cases in each LTC-DRG and produce inappropriate weighting factors at recalibration and result in inappropriate payments under the LTCH PPS. LTCHs are to follow the same coding guidelines used by acute care hospitals to ensure accuracy and consistency in coding practices. There will be only one LTC-DRG assigned per long-term care hospitalization; it will be assigned at the time of discharge of the patient. Therefore, it is mandatory that the coders continue to report the same principal diagnosis on all claims and include all diagnosis codes for conditions that coexist at the time of admission, for conditions that are subsequently developed, or for conditions that affect the treatment received. Similarly, all procedures performed in a LTCH, or paid for under arrangements by a LTCH, during that stay are to be reported on each claim.

Upon the discharge of the patient from a LTCH, the LTCH must assign appropriate diagnosis and procedure codes from the ICD-9-CM. Completed claim forms are to be submitted electronically to the LTCH's Medicare fiscal intermediary. Medicare fiscal intermediaries enter the clinical and demographic information into their claims processing systems and subject this information to a series of automated screening processes called the Medicare Code Editor (MCE). These screens are designed to identify cases that require further review before assignment into a LTC-DRG can be made.

After screening through the MCE, each LTCH claim will be classified into the appropriate LTC-DRG by the Medicare LTCH GROUPER. The LTCH GROUPER is specialized computer software and is the same GROUPER used under the IPPS. After the LTC-DRG is assigned, the Medicare fiscal intermediary determines the prospective payment by using the Medicare LTCH PPS PRICER program, which accounts for LTCH hospital-specific adjustments and payment rates. As provided for under the IPPS, we provide an opportunity for the LTCH to review the LTC-DRG assignments made by the fiscal intermediary and to submit additional information, if necessary, within a specified timeframe (§ 412.513(c)).

The LTCH GROUPER is used both to classify past cases in order to measure relative hospital resource consumption to establish the LTC-DRG weights and to classify current cases for purposes of determining payment. The records for all Medicare hospital inpatient discharges are maintained in the MedPAR file. The data in this file are used to evaluate possible DRG classification changes and to recalibrate the DRG weights during our annual update (as discussed in section II.E. of this preamble). The LTC-DRG relative weights are based on data for the population of LTCH discharges, reflecting the fact that LTCH patients represent a different patient-mix than patients in short-term acute care hospitals.

3. Development of the FY 2007 LTC-DRG Relative Weights

a. General Overview of Development of the LTC-DRG Relative Weights

As we stated in the August 30, 2002 LTCH PPS final rule (67 FR 55981), one of the primary goals for the implementation of the LTCH PPS is to pay each LTCH an appropriate amount for the efficient delivery of care to Medicare patients. The system must be able to account adequately for each LTCH's case-mix in order to ensure both fair distribution of Medicare payments and access to adequate care for those Medicare patients whose care is more costly. To accomplish these goals, we adjust the LTCH PPS standard Federal prospective payment system rate by the applicable LTC-DRG relative weight in determining payment to LTCHs for each case. Under the LTCH PPS, relative weights for each LTC-DRG are a primary element used to account for the variations in cost per discharge and resource utilization among the payment groups (§ 412.515). To ensure that Medicare patients classified to each LTC-DRG have access to an appropriate level of services and to encourage efficiency, we calculate a relative weight for each LTC-DRG that represents the resources needed by an average inpatient LTCH case in that LTC-DRG. For example, cases in a LTC-DRG with a relative weight of 2 will, on average, cost twice as much as cases in a LTC-DRG with a weight of 1.

b. Data

In the FY 2007 IPPS proposed rule (71 FR 24052), to calculate the proposed LTC-DRG relative weights for FY 2007, we obtained total Medicare allowable charges from FY 2005 Medicare LTCH bill data from the December 2005 update of the MedPAR file, which were the best available data at that time, and we used the proposed Version 24.0 of the CMS GROUPER used under the IPPS (as discussed in that same proposed rule) to classify cases. In that same proposed rule, we also proposed that if more recent data were available, we would use that data and the finalized Version 24.0 of the CMS GROUPER (used under the IPPS) to determine the final LTC-DRG relative weights for FY 2007. Accordingly, to calculate the final LTC-DRG relative weights for FY 2007 in this final rule, we obtained total Medicare allowable charges from FY 2005 Medicare hospital bill data from the March 2006 update of the MedPAR file (which are the most recent available data), and used the final Version 24.0 of the CMS GROUPER used under the IPPS (as discussed in section II.B. of this preamble) to classify cases.

We also stated in the FY 2007 IPPS proposed rule (71 FR 24052), as we discussed in the FY 2006 IPPS final rule (70 FR 47325), we have excluded the data from LTCHs that are all-inclusive rate providers and LTCHs that are reimbursed in accordance with demonstration projects authorized under section 402(a) of Pub. L. 90-248 as amended. Therefore, consistent with the proposed rule, in the development of the FY 2007 LTC-DRG relative weights in this final rule, we have excluded the data of the 19 all-inclusive rate providers and the 3 LTCHs that are paid in accordance with demonstration projects that had claims in the FY 2005 MedPAR file.

c. Hospital-Specific Relative Value Methodology

By nature, LTCHs often specialize in certain areas, such as ventilator-dependent patients and rehabilitation and wound care. Some case types (DRGs) may be treated, to a large extent, in hospitals that have, from a perspective of charges, relatively high (or low) charges. This nonarbitrary distribution of cases with relatively high (or low) charges in specific LTC-DRGs has the potential to inappropriately distort the measure of average charges. To account for the fact that cases may not be randomly distributed across LTCHs, we use a hospital-specific relative value (HSRV) method to calculate the LTC-DRG relative weights instead of the methodology used to determine the DRG relative weights under the IPPS described in section II.E. of this preamble. We believe this method will remove this hospital-specific source of bias in measuring LTCH average charges. Specifically, we reduce the impact of the variation in charges across providers on any particular LTC-DRG relative weight by converting each LTCH's charge for a case to a relative value based on that LTCH's average charge.

Under the HSRV method, we standardize charges for each LTCH by converting its charges for each case to hospital-specific relative charge values and then adjusting those values for the LTCH's case-mix. The adjustment for case-mix is needed to rescale the hospital-specific relative charge values (which, by definition, average 1.0 for each LTCH). The average relative weight for a LTCH is its case-mix, so it is reasonable to scale each LTCH's average relative charge value by its case-mix. In this way, each LTCH's relative charge value is adjusted by its case-mix to an average that reflects the complexity of the cases it treats relative to the complexity of the cases treated by all other LTCHs (the average case-mix of all LTCHs).

In accordance with the methodology established under § 412.523, as implemented in the August 30, 2002 LTCH PPS final rule (67 FR 55989 through 55991), we standardize charges for each case by first dividing the adjusted charge for the case (adjusted for short-stay outliers under § 412.529 as described in section II.F.4. (step 3) of this preamble) by the average adjusted charge for all cases at the LTCH in which the case was treated. Short-stay outliers under § 412.529 are cases with a length of stay that is less than or equal to five-sixths the average length of stay of the LTC-DRG. The average adjusted charge reflects the average intensity of the health care services delivered by a particular LTCH and the average cost level of that LTCH. The resulting ratio is multiplied by that LTCH's case-mix index to determine the standardized charge for the case.

Multiplying by the LTCH's case-mix index accounts for the fact that the same relative charges are given greater weight at a LTCH with higher average costs than they would at a LTCH with low average costs, which is needed to adjust each LTCH's relative charge value to reflect its case-mix relative to the average case-mix for all LTCHs. Because we standardize charges in this manner, we count charges for a Medicare patient at a LTCH with high average charges as less resource intensive than they would be at a LTCH with low average charges. For example, a $10,000 charge for a case at a LTCH with an average adjusted charge of $17,500 reflects a higher level of relative resource use than a $10,000 charge for a case at a LTCH with the same case-mix, but an average adjusted charge of $35,000. We believe that the adjusted charge of an individual case more accurately reflects actual resource use for an individual LTCH because the variation in charges due to systematic differences in the markup of charges among LTCHs is taken into account.

d. Low-Volume LTC-DRGs

In order to account for LTC-DRGs with low-volume (that is, with fewer than 25 LTCH cases), in accordance with the methodology established in the August 30, 2002 LTCH PPS final rule (67 FR 55984), we group those “low-volume LTC-DRGs” (that is, DRGs that contained between 1 and 24 cases annually) into one of five categories (quintiles) based on average charges, for the purposes of determining relative weights. Consistent with the FY 2007 IPPS proposed rule (71 FR 24052 and 24053), we will continue to employ this treatment of low-volume LTC-DRGs in determining the FY 2007 LTC-DRG relative weights using the best available LTCH data in this final rule. In that same proposed rule, using LTCH cases from the December 2005 update of the FY 2005 MedPAR file, we identified 173 LTC-DRGs that contained between 1 and 24 cases. As noted above, we also proposed that if more recent data were available, we would use that data and the finalized Version 24.0 of the CMS GROUPER (used under the IPPS) to determine the final LTC-DRG relative weights for FY 2007. Accordingly, for this final rule, using LTCH cases from the March 2006 update of the FY 2005 MedPAR file, we identified 180 LTC-DRGs that contained between 1 and 24 cases. This list of LTC-DRGs was then divided into one of the 5 low-volume quintiles, each containing 36 LTC-DRGs (180/5 = 36). In accordance with our established methodology, as we proposed, we then make an assignment to a specific low-volume quintile by sorting the low-volume LTC-DRGs in ascending order by average charge. For this final rule, this results in an assignment to a specific low-volume quintile of the sorted 180 low-volume LTC-DRGs by ascending order by average charge. For this final rule, based on LTCH claims data from the March 2006 update of the FY 2005 MedPAR file and the finalized Version 24.0 of the CMS GROUPER, the number of low-volume LTC-DRGs is evenly divisible by five (that is, the number of low-volume quintile used to determine the LTC-DRG relative weights). Consequently, for this final rule, it was not necessary to employ our established methodology to determine which low-volume quintile would receive the additional LTC-DRG(s) if the number of low-volume LTC-DRGs had not been evenly divisible by five. However, if the number of LTC-DRGs with less than 25 LTCH cases for this final rule had not evenly divisible by five, we would have employed our established methodology that compares the average charge of the low-volume LTC-DRGs, to determine which low-volume quintile would receive the additional LTC-DRG, as presented in greater detail in the FY 2007 IPPS proposed rule (71 FR 24053). Because, for this final rule, the number of LTC-DRGs with less than 25 LTCH cases was evenly divisible by five, to determine the composition of the low-volume quintiles, in accordance with our established methodology, as was proposed, we sorted the 180 low-volume LTC-DRGs in ascending order, and grouped the first fifth (1st through 36th) of low-volume LTC-DRGs (with the lowest average charge) into Quintile 1; the next fifth (37th through 72nd) of low-volume LTC-DRGs were into Quintile 2; and so on until the last fifth (145th through 180th) of low-volume LTC-DRGs (with the highest average charge) were grouped into Quintile 5.

In order to determine the relative weights for the LTC-DRGs with low volume for FY 2007, as was proposed, in accordance with the methodology established in the August 30, 2002 LTCH PPS final rule (67 FR 55984), in this final rule, we used the five low-volume quintiles described above. The composition of each of the five low-volume quintiles shown in the chart below was used in determining the LTC-DRG relative weights for FY 2007. As was proposed, for this final rule, we determined a relative weight and (geometric) average length of stay for each of the five low-volume quintiles using the formula that we apply to the regular LTC-DRGs (25 or more cases), as described below in section II.F.4. of this preamble. We assigned the same relative weight and average length of stay to each of the LTC-DRGs that make up that low-volume quintile. We note that, as this system is dynamic, it is possible that the number and specific type of LTC-DRGs with a low volume of LTCH cases will vary in the future. We use the best available claims data in the MedPAR file to identify low-volume LTC-DRGs and to calculate the relative weights based on our methodology.

We note that we will continue to monitor the volume (that is, the number of LTCH cases) in these low-volume quintiles to ensure that our quintile assignment results in appropriate payment for such cases and does not result in an unintended financial incentive for LTCHs to inappropriately admit these types of cases.

4. Steps for Determining the FY 2007 LTC-DRG Relative Weights

As we noted previously, as was proposed, the FY 2007 LTC-DRG relative weights in this final rule are determined in accordance with the methodology established in the August 30, 2002 LTCH PPS final rule (67 FR 55989 through 55991). In summary, LTCH cases must be grouped in the appropriate LTC-DRG, while taking into account the low-volume LTC-DRGs as described above, before the FY 2007 LTC-DRG relative weights can be determined. After grouping the cases in the appropriate LTC-DRG, we calculated the relative weights for FY 2007 in this final rule by first removing statistical outliers and cases with a length of stay of 7 days or less, as discussed in greater detail below. Next, we adjusted the number of cases in each LTC-DRG for the effect of short-stay outlier cases under § 412.529, as also discussed in greater detail below. The short-stay adjusted discharges and corresponding charges are used to calculate “relative adjusted weights” in each LTC-DRG using the HSRV method described above.

Below we discuss in detail the steps for calculating the FY 2007 LTC-DRG relative weights in this final rule. These steps are the same as the ones we presented in the FY 2007 IPPS proposed rule for calculating the proposed FY 2007 LTC-DRG relative weights. We note that, as we stated above in section II.F.3.b. of this preamble, we have excluded the data of all-inclusive rate LTCHs and LTCHs that are paid in accordance with demonstration projects that had claims in the FY 2005 MedPAR file.

Step 1—Remove statistical outliers.

The first step in the calculation of the FY 2007 LTC-DRG relative weights, as was proposed, is to remove statistical outlier cases. We define statistical outliers as cases that are outside of 3.0 standard deviations from the mean of the log distribution of both charges per case and the charges per day for each LTC-DRG. These statistical outliers are removed prior to calculating the relative weights. As noted above, we believe that they may represent aberrations in the data that distort the measure of average resource use. Including those LTCH cases in the calculation of the relative weights could result in an inaccurate relative weight that does not truly reflect relative resource use among the LTC-DRGs.

Step 2—Remove cases with a length of stay of 7 days or less.

The FY 2007 LTC-DRG relative weights reflect the average of resources used on representative cases of a specific type. Generally, cases with a length of stay of 7 days or less do not belong in a LTCH because these stays do not fully receive or benefit from treatment that is typical in a LTCH stay, and full resources are often not used in the earlier stages of admission to a LTCH. As explained above, if we were to include stays of 7 days or less in the computation of the FY 2007 LTC-DRG relative weights, the value of many relative weights would decrease and, therefore, payments would decrease to a level that may no longer be appropriate.

We do not believe that it would be appropriate to compromise the integrity of the payment determination for those LTCH cases that actually benefit from and receive a full course of treatment at a LTCH, in order to include data from these very short-stays. Thus, as explained above, in determining the FY 2007 LTC-DRG relative weights in this final rule, as was proposed, we remove LTCH cases with a length of stay of 7 days or less.

Step 3—Adjust charges for the effects of short-stay outliers.

After removing cases with a length of stay of 7 days or less, we are left with cases that have a length of stay of greater than or equal to 8 days. The next step in the calculation of the FY 2007 LTC-DRG relative weights is to adjust each LTCH's charges per discharge for those remaining cases for the effects of short-stay outliers as defined in § 412.529(a). (However, we note that even if a case was removed in Step 2 (that is, cases with a length of stay of 7 days or less), it was paid as a short-stay outlier if its length of stay was less than or equal to five-sixths of the average length of stay of the LTC-DRG, in accordance with § 412.529.)

We make this adjustment by counting a short-stay outlier as a fraction of a discharge based on the ratio of the length of stay of the case to the average length of stay for the LTC-DRG for nonshort-stay outlier cases. This has the effect of proportionately reducing the impact of the lower charges for the short-stay outlier cases in calculating the average charge for the LTC-DRG. This process produces the same result as if the actual charges per discharge of a short-stay outlier case were adjusted to what they would have been had the patient's length of stay been equal to the average length of stay of the LTC-DRG.

As we explained in the FY 2007 IPPS proposed rule (71 FR 24059), counting short-stay outlier cases as full discharges with no adjustment in determining the LTC-DRG relative weights would lower the LTC-DRG relative weight for affected LTC-DRGs because the relatively lower charges of the short-stay outlier cases would bring down the average charge for all cases within a proposed LTC-DRG. This would result in an “underpayment” for nonshort-stay outlier cases and an “overpayment” for short-stay outlier cases. Therefore, in this final rule, as was proposed, we adjust for short-stay outlier cases under § 412.529 in this manner because it results in more appropriate payments for all LTCH cases.

Step 4—Calculate the FY 2007 LTC-DRG relative weights on an iterative basis.

The process of calculating the LTC-DRG relative weights using the HSRV methodology is iterative. First, for each LTCH case, we calculate a hospital-specific relative charge value by dividing the short-stay outlier adjusted charge per discharge (see step 3) of the LTCH case (after removing the statistical outliers (see step 1)) and LTCH cases with a length of stay of 7 days or less (see step 2) by the average charge per discharge for the LTCH in which the case occurred. The resulting ratio is then multiplied by the LTCH's case-mix index to produce an adjusted hospital-specific relative charge value for the case. An initial case-mix index value of 1.0 is used for each LTCH.

For each LTC-DRG, the FY 2007 LTC-DRG relative weight is calculated by dividing the average of the adjusted hospital-specific relative charge values (from above) for the LTC-DRG by the overall average hospital-specific relative charge value across all cases for all LTCHs. Using these recalculated LTC-DRG relative weights, each LTCH's average relative weight for all of its cases (case-mix) is calculated by dividing the sum of all the LTCH's LTC-DRG relative weights by its total number of cases. The LTCHs' hospital-specific relative charge values above are multiplied by these hospital-specific case-mix indexes. These hospital-specific case-mix adjusted relative charge values are then used to calculate a new set of LTC-DRG relative weights across all LTCHs. In this final rule, as was proposed, this iterative process is continued until there is convergence between the weights produced at adjacent steps, for example, when the maximum difference is less than 0.0001.

Step 5—Adjust the FY 2007 LTC-DRG relative weights to account for nonmonotonically increasing relative weights.

As explained in section II.B. of this preamble, the FY 2007 CMS DRGs, on which the FY 2007 LTC-DRGs are based, contain “pairs” that are differentiated based on the presence or absence of CCs. The LTC-DRGs with CCs are defined by certain secondary diagnoses not related to or inherently a part of the disease process identified by the principal diagnosis, but the presence of additional diagnoses does not automatically generate a CC. As we discussed in the FY 2006 IPPS final rule (70 FR 47336), the value of monotonically increasing relative weights rises as the resource use increases (for example, from uncomplicated to more complicated). The presence of CCs in a LTC-DRG means that cases classified into a “without CC” LTC-DRG are expected to have lower resource use (and lower costs). In other words, resource use (and costs) are expected to decrease across “with CC/without CC” pairs of LTC-DRGs.

For a case to be assigned to a LTC-DRG with CCs, more coded information is called for (that is, at least one relevant secondary diagnosis), than for a case to be assigned to a LTC-DRG “without CCs” (which is based on only one principal diagnosis and no relevant secondary diagnoses). Currently, the LTCH claims data include both accurately coded cases without complications and cases that have complications (and cost more), but were not coded completely. Both types of cases are grouped to a LTC-DRG “without CCs” when only the principal diagnosis was coded. Since the LTCH PPS was only implemented for cost reporting periods beginning on or after October 1, 2002 (FY 2003), and LTCHs were previously paid under cost-based reimbursement, which is not based on patient diagnoses, coding by LTCHs for these cases may not have been as detailed as possible.

Thus, in developing the FY 2003 LTC-DRG relative weights for the LTCH PPS based on FY 2001 claims data, as we discussed in the August 30, 2002 LTCH PPS final rule (67 FR 55990), we found on occasion that the data suggested that cases classified to the LTC-DRG “with CCs” of a “with CC/without CC” pair had a lower average charge than the corresponding LTC-DRG “without CCs.” Similarly, as discussed in the FY 2006 IPPS final rule (70 FR 47336 through 47337), based on FY 2004 claims data, we also found on occasion that the data suggested that cases classified to the LTC-DRG “with CCs” of a “with CC/without CC” pair have a lower average charge than the corresponding LTC-DRG “without CCs” for the FY 2006 LTC-DRG relative weights.

We believe this anomaly may be due to coding that may not have fully reflected all comorbidities that were present. Specifically, LTCHs may have failed to code relevant secondary diagnoses, which resulted in cases that actually had CCs being classified into a “without CC” LTC-DRG. It would not be appropriate to pay a lower amount for the “with CC” LTC-DRG because, in general, cases classified into a “with CC” LTC-DRG are expected to have higher resource use (and higher cost) as discussed above. Therefore, previously when we determined the LTC-DRG relative weights in accordance with the methodology established in the August 30, 2002 LTCH PPS final rule (67 FR 55990) when we implanted the LTCH PPS, we grouped both the cases “with CCs” and “without CCs” together for the purpose of calculating the LTC-DRG relative weights. As we stated in that same final rule, we will continue to employ this methodology to account for nonmonotonically increasing relative weights until we have adequate data to calculate appropriate separate weights for these anomalous LTC-DRG pairs. We expect that, as was the case when we first implemented the IPPS, in most instances, this problem will be self-correcting, as LTCHs submit more completely coded data in the future.

There are three types of “with CC” and “without CC” pairs that could be nonmonotonic; that is, where the “without CC” LTC-DRG would have a higher average charge than the “with CC” LTC-DRG. For this final rule, using the LTCH cases in the March 2006 update of the FY 2005 MedPAR file (the most recent and complete data available at this time), we identified one of the three types of nonmonotonic LTC-DRG pairs. As we stated in the August 30, 2002 LTCH PPS final rule (67 FR 55990), we believe this anomaly may be due to coding inaccuracies and expect that, as was the case when we first implemented the acute care hospital IPPS, this problem will be self-correcting, as LTCHs submit more completely coded data in the future.

The first category of nonmonotonically increasing relative weights for LTC-DRG pairs “with and without CCs” contains one pair of LTC-DRGs in which both the LTC-DRG “with CCs” and the LTC-DRG “without CCs” had 25 or more LTCH cases and, therefore, did not fall into one of the 5 low-volume quintiles. For those nonmonotonic LTC-DRG pairs, based on our established methodology (67 FR 55983 through 55990), we combined the LTCH cases and computed a new relative weight based on the case-weighted average of the combined LTCH cases of the LTC-DRGs. The case-weighted average charge is determined by dividing the total charges for all LTCH cases by the total number of LTCH cases for the combined LTC-DRG. This new relative weight is then assigned to both of the LTC-DRGs in the pair. In this final rule, for FY 2007, there were no LTC-DRGs that fell into this category.

The second category of nonmonotonically increasing relative weights for LTC-DRG pairs “with and without CCs” consists of one pair of LTC-DRGs that has fewer than 25 cases, and each LTC-DRG is grouped to different low-volume quintiles in which the “without CC” LTC-DRG is in a higher-weighted low-volume quintile than the “with CC” LTC-DRG. For those pairs, based on our established methodology, we combine the LTCH cases and determine the case-weighted average charge for all LTCH cases. The case-weighted average charge is determined by dividing the total charges for all LTCH cases by the total number of LTCH cases for the combined LTC-DRG. Based on the case-weighted average LTCH charge, we determine within which low-volume quintile the “combined LTC-DRG” is grouped. Both LTC-DRGs in the pair are then grouped into the same low-volume quintile, thus having the same relative weight. In this final rule, for FY 2007, there were no LTC-DRGs that fell into this category.

The third category of nonmonotonically increasing relative weights for LTC-DRG pairs “with and without CCs” consists of one pair of LTC-DRGs where one of the LTC-DRGs has fewer than 25 LTCH cases and is grouped to a low-volume quintile and the other LTC-DRG has 25 or more LTCH cases and has its own LTC-DRG relative weight, and the LTC-DRG “without CCs” has the higher relative weight. Based on our established methodology, as proposed, we removed the low-volume LTC-DRG from the low-volume quintile and combined it with the other LTC-DRG for the computation of a new relative weight for each of these LTC-DRGs. This new relative weight is assigned to both LTC-DRGs, so they each have the same relative weight. In this final rule, for FY 2007, 5 “pairs” of LTC-DRGs fall into this category: LTC-DRGs 94 and 95; LTC-DRGs 96 and 97; LTC-DRGs 141 and 142; LTC-DRGs 205 and 206; and LTC-DRGs 292 and 293.

Step 6—Determine a FY 2007 LTC-DRG relative weight for LTC-DRGs with no LTCH cases.

As we stated above, in this final rule, as we proposed we determine the relative weight for each LTC-DRG using total Medicare allowable charges reported in the March 2006 update of the FY 2005 MedPAR file. Of the 538 LTC-DRGs for FY 2007, we identified 183 LTC-DRGs for which there were no LTCH cases in the database. That is, based on data from the FY 2005 MedPAR file used in this final rule, no patients who would have been classified to those LTC-DRGs were treated in LTCHs during FY 2005 and, therefore, no charge data were reported for those LTC-DRGs. Thus, in the process of determining the LTC-DRG relative weights, we are unable to determine weights for these 183 LTC-DRGs using the methodology described in Steps 1 through 5 above. However, because patients with a number of the diagnoses under these LTC-DRGs may be treated at LTCHs beginning in FY 2007, as was proposed, for this final rule, we assigned relative weights to each of the 183 “no volume” LTC-DRGs based on clinical similarity and relative costliness to one of the remaining 355 (538 - 183 = 355) LTC-DRGs for which we are able to determine relative weights, based on FY 2005 LTCH claims data. As there are currently no LTCH cases in these “no volume” LTC-DRGs, as proposed, we determined relative weights for the 183 LTC-DRGs with no LTCH cases in the FY 2005 MedPAR file used in this final rule by grouping them to the appropriate low-volume quintile. This methodology is consistent with our methodology used in determining relative weights to account for the low-volume LTC-DRGs described above.

As was proposed, for this final rule, our methodology for determining the relative weights for the “no volume” LTC-DRGs is as follows: We crosswalk the no volume LTC-DRGs by matching them to other similar LTC-DRGs for which there were LTCH cases in the FY 2005 MedPAR file based on clinical similarity and intensity of use of resources as determined by care provided during the period of time surrounding surgery, surgical approach (if applicable), length of time of surgical procedure, postoperative care, and length of stay. We assigned the relative weight for the applicable low-volume quintile to the no volume LTC-DRG if the LTC-DRG to which it is crosswalked is grouped to one of the low-volume quintiles. If the LTC-DRG to which the no volume LTC-DRG is crosswalked is not one of the LTC-DRGs to be grouped to one of the low-volume quintiles, we compared the relative weight of the LTC-DRG to which the no volume LTC-DRG is crosswalked to the relative weights of each of the five quintiles and we assigned the no volume LTC-DRG the relative weight of the low-volume quintile with the closest weight. For this final rule, a list of the no volume FY 2007 LTC-DRGs and the FY 2007 LTC-DRG to which it is crosswalked in order to determine the appropriate low-volume quintile for the assignment of a relative weight for FY 2007 is shown in the chart below.

To illustrate this methodology for determining the relative weights for the 183 LTC-DRGs with no LTCH cases, we are providing the following examples, which refer to the no volume LTC-DRGs crosswalk information for FY 2007 provided in the chart above.

Example 1: There were no cases in the FY 2005 MedPAR file used for this final rule for LTC-DRG 3 (Craniotomy Age 0-17). Since the procedure is similar in resource use and the length and complexity of the procedures and the length of stay are similar, we determined that LTC-DRG 1 (Craniotomy Age >17 with CC), which is assigned to low-volume Quintile 5 for the purpose of determining the FY 2007 relative weights, would display similar clinical and resource use. Therefore, we assigned the same relative weight of LTC-DRG 1 of 1.6835 (Quintile 5) for FY 2007 (Table 11 in the Addendum to this final rule) to LTC-DRG 3.

Example 2: There were no LTCH cases in the FY 2005 MedPAR file used in this final rule for LTC-DRG 91 (Simple Pneumonia and Pleurisy Age 0-17). Since the severity of illness in patients with pneumonia and pleurisy is similar in patients regardless of age, we determined that LTC-DRG 90 (Simple Pneumonia and Pleurisy Age >17 Without CC) would display similar clinical and resource use characteristics and have a similar length of stay to LTC-DRG 91. There were over 25 cases in LTC-DRG 90 in the FY 2005 MedPAR file data used determining the FY 2007 LTC-DRG relative weights in this final rule. Therefore, it would not be assigned to a low-volume quintile for the purpose of determining the LTC-DRG relative weights. However, under our established methodology, LTC-DRG 91, with no LTCH cases, would need to be grouped to a low-volume quintile. We determined that the low-volume quintile with the closest weight to LTC-DRG 90 (0.4958) (refer to Table 11 in the Addendum to this final rule) would be low-volume Quintile 2 (0.5594) (refer to Table 11 in the Addendum to this final rule). Therefore, we assigned LTC-DRG 91 a relative weight of 0.5694 for FY 2007. We note that we will continue to monitor the volume (that is, the number of LTCH cases) that have few or no LTCH cases to ensure that our no volume LTC-DRG crosswalking and relative weight assignment results in appropriate payments for such cases and does not result in an unintended financial incentive for LTCHs to inappropriately admit these types of cases.

Furthermore, as was proposed, we are establishing LTC-DRG relative weights of 0.0000 for heart, kidney, liver, lung, pancreas, and simultaneous pancreas/kidney transplants (LTC-DRGs 103, 302, 480, 495, 512, and 513, respectively) for FY 2007 in this final rule because Medicare will only cover these procedures if they are performed at a hospital that has been certified for the specific procedures by Medicare and presently no LTCH has been so certified. Based on our research, we found that most LTCHs only perform minor surgeries, such as minor small and large bowel procedures, to the extent any surgeries are performed at all. Given the extensive criteria that must be met to become certified as a transplant center for Medicare, we believe it is unlikely that any LTCHs would become certified as a transplant center. In fact, in the nearly 20 years since the implementation of the IPPS, there has never been a LTCH that even expressed an interest in becoming a transplant center.

However, if in the future a LTCH applies for certification as a Medicare-approved transplant center, we believe that the application and approval procedure would allow sufficient time for us to determine appropriate weights for the LTC-DRGs affected. At the present time, we would only include these six transplant LTC-DRGs in the GROUPER program for administrative purposes. Because we use the same GROUPER program for LTCHs as is used under the IPPS, removing these LTC-DRGs would be administratively burdensome.

Again, we note that, as this system is dynamic, it is entirely possible that the number of LTC-DRGs with a zero volume of LTCH cases based on the system will vary in the future. We used the best most recent available claims data in the MedPAR file to identify zero volume LTC-DRGs and to determine the relative weights in this final rule.

Table 11 in the Addendum to this final rule lists the LTC-DRGs and their respective relative weights, geometric mean length of stay, and five-sixths of the geometric mean length of stay (to assist in the determination of short-stay outlier payments under § 412.529) for FY 2007.

We also wish to point out that in section VI.A.5. of the preamble of this rule, we discuss our revision to the regulations for grandfathered HwHs, grandfathered hospital satellite facilities, and grandfathered satellite units at §§ 412.22(f), 412.22(h), and 412.25(e), respectively. In addition, in section VI.A.6. of the preamble of this final rule, we discuss our revision and clarification to the existing policies governing the determination of LTCHs' CCRs and the reconciliation of high-cost and short-stay outlier payments under the LTCH PPS based on the proposal presented in the FY 2007 IPPS proposed rule (71 FR 24126 through 24135).

5. Summary of Public Comments and Departmental Responses

Comment: Numerous commenters opposed the proposed changes in the LTC-DRG weights, which they noted would result in an approximately 1.4 percent decrease in estimate aggregate payments to LTCHs. Several of the commenters noted that LTCHs had been subject to a number of “significant Medicare payment reductions in recent years,” including an estimated 4.2 percent reduction as a result of the reweighting of the LTC-DRGs for FY 2006; a zero update (as opposed to a 3.4 percent market-basket increase) in the Federal rate for RY 2007; an estimated 3.7 percent decrease caused by the revised short-stay outlier payment policy for RY 2007; and, most recently, the estimated 1.4 percent reduction as a result of the proposed reclassification and reweighting of the LTC-DRGs for FY 2007. The commenters maintained that the cumulative effect of these established and proposed Medicare payment reductions is not sustainable for the LTCH industry and will cause much “volatility” for LTCH providers, and also restrict access to LTCHs for patients.

One commenter provided a chart that indicated that if CMS finalizes the proposed LTC-DRG relative weights, LTCH industry-wide margins would approximate 0 percent. Another commenter, an association that represents large LTCH chains, urged CMS to postpone implementation of the proposed FY 2007 reweighting of the LTC-DRGs until an analysis of the impact of this change on payment adequacy, as well as other payment changes established for RY 2007, is conducted.

Response: While we understand the commenters' concerns with the estimated decrease of 1.4 percent in LTCH PPS payments as a result of the proposed changes in the LTC-DRGs, and relative weights for FY 2007, we did not propose any changes in the methodology used to determine the proposed recalibration of the LTC-DRG relative weights for FY 2007. (We note that based on the final LTCH-DRG relative weights for FY 2007 the estimate is a 1.3 percent decrease.) The proposed update to the LTC-DRG relative weights for FY 2007 is based on the proposed Version 24.0 of the CMS GROUPER (including the proposed changes in the DRG classifications relative weights and geometric mean length of stay) and FY 2005 LTCH claims data. For this final rule, we used updated data as described previously. In the FY 2003 final rule for the LTCH PPS, which first implemented the payment system, we described in great detail, the methodology for the development of the LTC-DRG relative weights, and we have reiterated these steps in every subsequent rulemaking cycle. (When we revised our regulations at § 412.535, establishing the LTCH PPS rate year, while still publishing the LTC-DRG updates on the Federal fiscal year (October through September) cycle, we continued to include a brief write-up of our LTC-DRG update methodology in the annual LTCH PPS proposed and final rules and a comprehensive description of the policy in the annual IPPS proposed and final rules (67 FR 55984-55995; 68 FR 34131-34132; 69 FR 25681; 69 FR 48989-48999; 70 FR 24177-24178; 70 FR 37323-37341; and 71 FR 27808).) There has been no methodological change in the way in which the LTC-DRG relative weights are computed since the implementation of the LTCH PPS. The annual determination of the LTC-DRG relative weights is data-driven; that is, based on claims data in the most current MedPAR files which are derived from patient bills submitted by LTCHs.

We agree with the commenters who noted that the LTCH industry has indeed been impacted by significant changes since the start of the LTCH PPS for FY 2003. Since we first established the LTCH PPS, the unadjusted Federal payment rate, which began at $34,956.15, increased to $38,086.64 for RY 2006. (The zero percent update finalized in the RY 2007 LTCH PPS final rule (71 FR 27798) resulted in the stabilization of this amount for RY 2007.) From RY 2005 to RY 2006, there was a 5.7 percent increase in estimated aggregate LTCH PPS payments (70 FR 24217). The average Medicare payment per case for FY 2003 was reported at $26,751, while, for RY 2006, it was estimated to be $33,208, which is an increase of over 24 percent. Significantly, there was a 13.8 percent increase in estimated Medicare payments to LTCHs in RY 2005 alone. The results of the first 2 years of this “volatility” were aggregate industry margins estimated at 7.8 percent for FY 2003, and for FY 2004, preliminary cost report data revealed an estimated average Medicare margin of 12.7 percent, as stated in the RY 2007 LTCH PPS final rule (71 FR 27819).

The commenters noted the Medicare payment reductions in recent years, including the estimated 4.2 percent reduction for FY 2006 due to the recalibration of the LTC-DRG weights and the estimated 1.4 percent decrease in aggregate LTCH PPS payments due to the proposed update to the LTC-DRG relative weights for FY 2007. As noted above, the decrease in average case-mix based on the proposed LTC-DRG relative weights for FY 2007 as compared to FY 2006, as well as the decrease in average case-mix from FY 2006 as compared to FY 2005, which were estimated to result in an aggregate estimated decrease in LTCH PPS payments, were data driven. For this final rule they remain data driven as well. In the FY 2006 IPPS proposed rule (70 FR 23667), we noted that we continued to observe a significant increase of relatively lower charge cases being assigned to LTC-DRGs with higher relative weights in the prior year. The addition of these lower charge cases resulted in a decrease in many of the LTC-DRG relative weights from FY 2005 to FY 2006. This decrease in many of the LTC-DRG relative weights, in turn, resulted in an estimated decrease in LTCH PPS payments from FY 2005 to FY 2006. As we explained in that same rule, contributing to this increased number of relatively lower charge cases being assigned to LTC-DRGs with higher relative weights in the prior year were improvements in coding practices, which are typically found when moving from a reasonable cost-based payment system to a PPS.

Our analyses of data from the March 2005 update of the FY 2004 MedPAR files, which were used to calculate the FY 2006 LTC-DRG relative weights, and the most recent update of the FY 2005 MedPAR files which were used to determine the proposed and final FY 2007 LTC-DRG relative weights continue to show an increase of relatively lower charge cases being assigned to LTC-DRGs with higher relative weights in the prior year. As we explained in the FY 2006 IPPS final rule (70 FR 47335) and the FY 2007 IPPS proposed rule (71 FR 24413), the impact of including cases with relatively lower charges into LTC-DRGs that had a relatively higher relative weight in the previous fiscal year's GROUPER is a decrease in the average relative weight for those LTC-DRGs, which, in turn, may result in an estimated aggregate decrease in LTCH PPS payments.

The commenters also mentioned the zero update to the RY 2007 standard Federal rate as one of the “significant Medicare payment reductions in recent years.” In the RY 2007 LTCH PPS final rule (71 FR 27819 through 27827), we explained our rationale for establishing a zero percent update to the standard Federal rate for the 2007 LTCH PPS rate year, which was based on the most recent estimate in the Rehabilitation, Psychiatric and Long-Term Care (RPL) market basket offset by an adjustment for changes in coding practices that are unrelated to case mix, rather than solely using the most recent estimate of the RPL market basket to update the RY 2006 Federal rate. This market basket offset resulted from a number of factors that included our ongoing monitoring activities, which prompted us to examine the changes in LTCHs' patient case-mix index and margins since the inception of the LTCH PPS for FY 2003 (67 FR 56014).

First, we noted that there has been tremendous growth in the number of LTCHs reimbursed by Medicare. Specifically, the number of LTCHs almost doubled from approximately 200 LTCHs in FY 2003 to 378 LTCHs at the start of FY 2005. In addition, Medicare spending for LTCHs has also grown rapidly, as noted in MedPAC's June 2004 Report to Congress (page 122). Rapid increases in LTCH growth and Medicare spending under the LTCH PPS, in conjunction with the fact that over 98 percent of LTCHs are currently paid based fully on the Federal rate (rather than choosing to be paid under a blend of the reasonable cost-based (TEFRA) payment amount and the LTCH PPS Federal rate payment amount), prompted us to examine changes in LTCHs' patient case-mix index and margins under the LTCH PPS. We believed the zero percent update factor for RY 2007, which was based on the most recent estimate of the RPL market basket at that time, adjusted to account for coding changes, was supported by our findings regarding the case-mix index, Medicare margins, and patient census based on the most recent complete LTCH data.

As we explained in considerable detail in the RY 2007 final rule for the LTCH PPS (71 FR 27818 through 27824), a LTCH's case-mix index is defined as the case-weighted average LTC-DRG relative weight for all its discharges in a given period. Changes in the case-mix index consist of two components: “real” case-mix index changes and “apparent” case-mix index changes. Real case-mix index increase is defined as the increase in the average LTC-DRG relative weights resulting from the hospital's treatment of more resource intensive patients. Apparent case-mix index increase is defined as the increase in computed case-mix index that is due to changes in coding practices (including better documentation of the medical record by physicians and more complete coding of the medical record by coders). Observed case-mix index increase is defined as real case-mix index increase plus the apparent case-mix index increase.

If LTCH patients have more costly impairments, lower functional status, or increased comorbidities, and thus require more resources in the LTCH, we consider this a real change in case-mix. Conversely, if LTCH patients have the same impairments, functional status, and comorbidities but are coded differently resulting in higher payment, we consider this an apparent change in case-mix. We believe that changes in payment rates should accurately reflect changes in LTCHs' true cost of treating patients (real case-mix index increase), and should not be influenced by changes in coding practices (apparent case-mix index increase). Apparent case-mix index increase results in a case being grouped to a LTC-DRG with a higher weight than it would be without such changes in coding practices, which results in a higher payment to the LTCH that does not necessarily reflect the true cost of treating the patient. Therefore, in the RY 2007 LTCH PPS final rule (71 FR 27798) under the broad discretionary authority conferred upon the Secretary by section 123 of the BBRA as amended by section 307(b) of the BIPA to include appropriate adjustments, including updates, in the establishment of the LTCH PPS, we revised the annual update to the LTCH PPS standard Federal rate set forth at § 412.523(a)(2) for the 2007 LTCH PPS rate year to adjust the payment amount for LTCH inpatient hospital services to eliminate the effect of coding or classification changes that do not reflect real changes in LTCHs' case-mix.

Our determination to specifically provide a zero update resulted from data analysis by 3M Health Information Systems (3M) regarding changes in case-mix and coding since the implementation of the LTCH PPS, based on the most recently available data, which compared FY 2003 LTCH claims data from the first year of implementation of the PPS with FY 2004 LTCH claims data, and also looked at FY 2001 claims data (generated prior to the implementation of the LTCH PPS). (The FY 2001 data was the same LTCH claims data used to develop the LTCH PPS.) The analysis indicated, among other things, that the average annual case-mix index increase from FY 2001 to FY 2003 was 2.75 percent. Since coding of diagnoses was not a factor in determining payments under the former reasonable cost-based (TEFRA) payment system, and since payments were not directly tied to diagnosis codes, there was no incentive for LTCHs to attempt to influence payments through changes in coding practices. Therefore, it was reasonable to assume that the observed 2.75 percent change in case-mix in the years prior to the implementation of the LTCH PPS represent the value for the real case-mix index increase (that is, we assumed that the 2.75 percent increase in case-mix is due to treatment of more resource intensive patients, rather than to improvements in documentation or more complete coding of the medical record during this period). Using the average annual 2.75 percent observed case-mix index increase as a baseline, we separated the computed case-mix index increase between FY 2003 and FY 2004 into the real case-mix index increase, which is based on the treatment of more resource intensive patients, and the apparent case-mix index increase, due to improvements in documentation and coding practices.

As we stated in the RY 2007 LTCH PPS final rule (71 FR 27820), the calculated observed case-mix index increase between FY 2003 and FY 2004 was 6.75 percent. Assuming that the real case-mix index increase observed (on average) from FY 2001 to FY 2003 remained relatively constant into FY 2004, then the difference of 4.0 percent (6.75 percent minus 2.75 percent) represented the apparent case-mix index increase that was due to improvements in documentation and coding. This was considerably higher than the 0.34 percent behavioral offset originally estimated by the CMS Office of the Actuary, which was used in the development of the FY 2003 LTCH PPS standard Federal rate (67 FR 56033). Therefore, we believed that it was appropriate that the market basket be offset by an adjustment to account for changes in coding practices that do not reflect changes in real case mix. This adjustment was implemented to ensure that the LTCH PPS payment rates continue to reflect, as closely as possible, the true costs of treating LTCH patients. It was our intent that such an adjustment to the most recent estimate of the LTCH PPS market basket would eliminate the effect of coding or classification changes that did not reflect real changes in LTCHs' case-mix in prior years.

Regarding the impact of the revised short-stay outlier policy on Medicare payments to LTCHs, we continue to believe that the revisions we established to the short-stay outlier payment adjustments in the RY 2007 LTCH PPS final rule were highly appropriate and that they provide fair and reasonable payment for short-stay patients in LTCHs, which are required to meet the same certification criteria as short-term acute care hospitals set forth in section 1861(e) of the Act and generally have an average length of stay of greater than 25 days. Therefore, our present policy under the short-stay outlier policy at § 412.529, effective for discharges beginning on or after July 1, 2006, is to base Medicare payment on the least of 100 percent of the estimated costs of the discharge, 120 percent of the LTC-DRG per diem payment amount multiplied by the length of stay, the full LTC-DRG payment, or a LTCH PPS payment based on a blend of the IPPS-comparable per diem payment amount (capped at the full IPPS comparable payment amount) and a payment based on 120 percent of the LTC-DRG per diem amount.

We believe that this finalized policy clearly demonstrates our rationale, which is that as the length of a short-stay outlier case increases, the case begins to resemble a more “typical” LTCH stay as defined under section 1886(d)(1)(B)(IV)(I) of the Act and envisioned by the statutes authorizing the establishment of the LTCH PPS. Furthermore, the estimated 3.7 percent decrease in payments cited by the commenters will only have an impact on payments to those LTCHs that continue to admit a large number of very short-stay patients. We believe that the previous short-stay outlier policy, under which Medicare paid the least of 120 percent of the estimated cost of the case, 120 percent of the per diem LTC-DRG multiplied by the length of stay, or the full LTC-DRG, inadvertently provided an incentive for a LTCH to inappropriately admit patients who could otherwise have been treated in acute care hospitals and paid for under the IPPS. Therefore, we believe the provisions of the short-stay outlier policy that were finalized in the RY 2007 LTCH PPS final rule (71 FR 27845 through 27872) will result in fair and equitable payment for short-stay patients at LTCHs.

In response to the commenter who provided a chart that indicated industry-wide margins of approximately zero percent because of the proposed changes in the LTC-DRG relative weights that are anticipated to result from the 1.4 percent payment reduction, we continue to believe that our case-mix analysis (case-mix index) and Medicare margins analysis are sound. In the RY 2007 final rule for the LTCH PPS, we calculated “revenue-weighted” Medicare margins, which are the sum of hospital inpatient Medicare revenue (payments) minus the sum of hospital inpatient Medicare expenses (costs) divided by the sum of hospital inpatient Medicare revenue (payments). This margin analysis, which is also utilized by MedPAC in its analyses, is used to evaluate the overall financial status of LTCHs in general. In our analysis of the latest available LTCH data, we found that LTCH Medicare margins for FY 2003 (the first year of the LTCH PPS) were 7.8 percent, and preliminary data for FY 2004 based on the most recent HCRIS data revealed an even higher Medicare margin of 12.7 percent. Moreover, our analysis of LTCHs' payments and costs per discharge based on the latest available cost report data supports our adjustment to account for changes in coding practices that do not reflect changes in real case mix because it shows that, while payments (revenue) increased approximately 15 percent from FY 2002 to FY 2003, costs (expenses) per discharge increased by only 8 percent for the same period.

Thus, payments to LTCHs from FY 2002 to FY 2003 increased almost twice as much as the increase of costs for the same period. We also noted that even though we established a zero update to the Federal payment rate for RY 2007, we continue to believe that, based on the sizeable Medicare margins among LTCHs, the standard rate for the RY 2007 LTCH PPS will not affect beneficiary access to LTCH services because LTCHs will continue to be paid adequately to reflect the cost of resources needed to treat Medicare beneficiaries. We also note that MedPAC's March 2006 Report to Congress on Medicare Payment Policy included similar data on margins and, based on its indepth evaluation of payment adequacy for LTCHs for 2006, MedPAC recommended that there be no update to the LTCH PPS Federal rate for RY 2007.

In addition, we do not believe that it would be appropriate to “postpone implementation” of the proposed reweighting of the LTC-DRGs pending an analysis of the impact on LTCH payment adequacy of this change, as well as other payment changes established for LTCHs for RY 2007. The annual recalibration of the LTC-DRG relative weights, which is based on patient data, is one of the cornerstones of all prospective payment systems. To reiterate, we believe that the policies finalized for RY 2007, including the zero percent update to the standard Federal rate and the payment adjustment for short-stay outlier cases, do not provide any impediment to the ability of LTCHs to continue to maintain the quality or the availability of appropriately delivered LTCH services to Medicare beneficiaries.

Comment: Several comments questioned the methodology that we used that distinguishes between payment “reductions” resulting from the zero update to the standard Federal rate finalized in the RY 2007 LTCH PPS final rule and payment reductions resulting from the proposed reweighting of the LTC-DRGs for FY 2007. One commenter asserted that CMS has utilized the same rationale as a basis to propose to reduce the FY 2007 LTC-DRG relative weights that were used to apply a zero percent update in the RY 2007 LTCH PPS final rule. The commenters believed CMS has double-counted the same phenomenon.

Another commenter stated that, because the LTC-DRG relative weights are not updated in a budget-neutral manner, through the annual recalibration of the weights, the LTC-DRG system will “self-correct over time” without the need for any lowering of the Federal payment rate. The commenter believed that this non-budget neutral weight recalibration will continue to correct for the case-mix creep until coding improvement reaches a plateau, at which point annual case-mix variation will reflect actual variations in case-mix intensity. Citing our justification of “apparent” as opposed to real case-mix increase based on FY 2004 LTCH data for the zero percent update to the Federal rate for RY 2007, the commenter believed that CMS has overpenalized LTCHs by a net 4.2 percent. The commenter recommended that CMS work with the industry to establish an update system that eliminates the possibility of “over reduction” due to case-mix creep by one of the following options: implementing a budget neutral recalibration system and address case-mix creep through the update; or alternatively, maintaining the current non-budget neutral weight recalibration system but foregoing any future Federal rate update reduction for case-mix creep.

Response: The commenters have expressed concern that, if we finalize the proposed change in the FY 2007 LTC-DRG relative weights, the change would result in an estimated 1.4 percent decrease in payments. Because we have already finalized the zero update to the RY 2007 standard Federal rate, the commenters believe we will have reduced payments to LTCHs twice for the same phenomenon. We would like to remind the commenters that the “zero percent” update to the Federal rate for RY 2007 did not reduce LTCH PPS payments from their previous level. Instead, the Federal rate remained at $38,086.04 from RY 2006 to RY 2007. Furthermore, we disagree and do not believe that LTCHs are being penalized twice, once through adjustment of the standard Federal rate and again due to the proposed and finalized recalibration of the LTC-DRG relative weights for FY 2007.

In the LTCH PPS RY 2007 final rule, we addressed a similar allegation by commenters that we were “unfairly penalizing” LTCHs twice in proposing the zero percent update to the standard Federal rate as a remedy for inappropriate Medicare payments to LTCHs resulting from “case-mix creep” (that is, the “apparent” case-mix index increase) between FYs 2003 and 2004. At that time, several commenters stated that CMS had already corrected any coding issues from FY 2004 by the annual recalibration of the LTC-DRGs for FY 2006 based on case-mix changes from FYs 2003 and 2004, which resulted in an estimated decrease of 4.2 percent in payments to LTCHs.

In the RY 2007 LTCH PPS final rule (71 FR 27882), we presented the explanation of the distinction between the annual reweighting of the LTC-DRGs, which we expect to result in appropriate payments for the forthcoming fiscal year's LTCH discharges, and determinations regarding the appropriate application of adjustments to the market basket increase applied to the standard Federal rate which was established to account for payments made in a prior year that were based on improved coding rather than increased patient severity (71 FR 27821). At that time, we reviewed the discussion in the FY 2006 IPPS final rule (70 FR 47701-47702) in which we estimated that a payment reduction of −4.2 percent would result from the FY 2006 recalibration of the LTC-DRG relative weights, which were based on LTCH claims data from the FY 2004 MedPAR file. We stated “* * * [t]hus FY 2004 LTCH claims data, which reflected improved coding, were used to determine the LTC-DRG relative weights used to pay LTCH PPS discharges occurring during FY 2006. While it is true that the reweighting of the LTC-DRGs using FY 2004 LTCH claims served to update the relative weights based on actual claims data in each LTC-DRG, which also reflects coding improvements that occurred in FY 2004, the recalibration of LTC-DRG weights only corrects for any coding improvement for the purpose of making accurate LTCH PPS payments in FY 2006.” (71 FR 27822)

However, annual recalibration does not serve to account for payments that were made based on improved coding (rather than patient severity) in prior years. The case-mix adjustment to the market basket in determining the RY 2007 Federal rate is meant to reduce current payments to account for the increase payments that occurred in FY 2004 that resulted from the CMI increase that is attributable to “case-mix” creep in that year 71 FR 27822).

We also explained the rationale and computations underlying our update for RY 2007 in that same final rule: “In the RY 2007 LTCH PPS proposed rule, we proposed to offset the market basket by an amount equal to the increase in case mix that was due solely to improved documentation and coding rather than changes in real case mix. At the time of the proposed rule, that increase was within rounding error of the market basket, and therefore resulted in a proposed Federal rate for RY 2007 that was equal to the RY 2006 Federal rate, and not a reduction to the RY 2006 Federal rate.” (71 FR 27821). Therefore, this policy determination regarding the market basket increase of zero percent for RY 2007 was based on changes in the LTCHs' case-mix indices in conjunction with a broader analysis of trends in the LTCH industry (noted most recently by MedPAC in the Commission's March 2006 Report to the Congress (page 211)) and in particular, driven by a detailed analysis of LTCH margins since the implementation of the LTCH PPS. As we stated in that same final rule, we believe that, in determining the Federal rate update for RY 2007, it is appropriate to apply an adjustment to the most recent estimate of the LTCH PPS market basket to eliminate the effects of coding and classification changes that do not reflect changes in real case-mix. This adjustment is necessary to account for prior year payments that were made based on improved coding practices (rather than increased patient severity) (71 FR 27821). Furthermore, we note that FY 2004 LTCH claims data were used to determine the adjustment to the market basket to account for changes in coding practices in establishing the zero percent update to the Federal rate for RY 2007, while FY 2005 LTCH claims data were used to determine the proposed and final FY 2007 LTC-DRG relative weights. Because LTCH claims data from different years were used to determine the two adjustments noted by the commenters, we further disagree that we “double counted the same phenomenon.”

Regarding our margins analysis, based on data from the LTCHs' cost reports received as of December 31, 2005, updated LTCH margins analysis for the LTCH PPS RY 2007 final rule continued to show high Medicare margins among LTCHs since the implementation of the LTCH PPS in FY 2003. As noted in the RY 2007 LTCH PPS final rule, “[w]e calculated ‘revenue-weighted’ Medicare margins, which are the sum of hospital inpatient Medicare revenue (payments) minus the sum of hospital inpatient Medicare expenses (costs) divided by the sum of hospital inpatient Medicare revenue (payments). This margin calculation, also utilized by MedPAC in its analyses, is used to evaluate the overall financial status of LTCHs in general. In an analysis of the latest available LTCH cost reports, we found that LTCH Medicare margins for FY 2003 (the first year of the LTCH PPS) were 7.8 percent and preliminary cost report data for FY 2004 based on the most recent update to the cost report data in HCRIS reveal an even higher Medicare margin of 12.7 percent. For periods prior to the implementation of the LTCH PPS (that is, FY 1999 through FY 2002), we found that aggregate Medicare margins ranged between a minimum of −2.3 percent in FY 2000, and a maximum of 1.5 percent in FY 2002.” (71 FR 27823).

We wish to emphasize that, as we specified in the RY 2007 LTCH PPS proposed rule, the large observed increase in LTCH case-mix was not accompanied by a corresponding increase in Medicare costs. This was consistent with our belief expressed earlier that a significant part of this observed increase in case-mix was “apparent” and not “real.” In conjunction with an increase in real case-mix (that is, patient severity), we would have expected to see a significant increase in costs per discharge, even taking into account LTCH operating efficiencies, to pay for the resources needed to treat sicker patients. Consistent with MedPAC's most recent research discussed in its March 2006 Report to Congress (section 4C), our margins analysis indicated that, in spite of the estimated real increase in case-mix (severity of patients), payments to LTCHs under the LTCH PPS are generally more than adequate to cover the Medicare costs of the inpatient hospital services provided to LTCH patients.

Therefore, for the reasons discussed above, we disagree with the commenters who believe that we “double counted the same phenomenon.” To summarize, the purpose of the adjustment to the market basket which was to account for changes in coding practices that resulted in a zero percent update to the Federal rate for RY 2007 and the changes in payments that will result from the proposed and final reweighting of the LTC-DRGs are different. Specifically, the objective of our adjustment to the standard Federal rate update for RY 2007 was to adjust payments to account for prior year payments made by the Medicare program that were due to changes in coding practices, that did not reflect actual costs of beneficiary care. However, the annual recalibration of the relative weights for LTC-DRGs reflects the variation in coding practices and charges from the previous year and it helps ensure that the LTC-DRG relative weights in the upcoming fiscal year will result in appropriate payments to LTCHs for the resources they expend to treat patients. This was the case for FY 2006, when LTC-DRG relative weight recalibrations were estimated to result in a payment decrease of 4.2 percent and it was also the case for the estimated 1.4 percent decrease based on the proposed LTC-DRG relative weights for FY 2007. It is also the case for the estimated 1.3 percent decrease in this final rule due to the recalibration of the LTC-DRG relative weights.

Therefore, in response to the commenter who presented an “either/or” scenario suggesting that we should adjust payments based on case-mix variation through the present (that is, not budget neutral) recalibration of the LTC-DRG relative weights but forego any future Federal rate update for case-mix creep, or we should address “case-mix creep” through the annual update in the Federal rate but implement a budget neutral recalibration system, we do not believe that this approach is appropriate, given that, as discussed in greater detail above, the purposes of the case-mix adjustments in each context are distinct. It is possible that if coding practices stabilize and reach “a plateau,” as one of the commenters suggested, and case-mix variation only reflects real variations in case-mix intensity, the “self-correcting” mechanism of the annual recalibrations of LTC-DRG relative weights may be a reliable indication of actual costs at LTCHs by DRG. However, we emphasize that there is a distinct difference between the payment adjustments that could result from data-driven determinations that we consider, as described earlier, when we promulgate our policy regarding the annual application of the market basket update to the standard Federal rate and the data-driven effects of the recalibration of the LTC-DRG relative weights. Moreover, we do not believe that the zero update to the standard Federal rate implemented for RY 2007, which was intended to adjust for payments that were reflective of payments that were made based on improved coding rather than patient severity in 2004, and the reweighting of the relative weights for the LTC-DRGs, which would only address making appropriate payments for FY 2007, have resulted in an “over reduction” of payments to LTCHs, or overpenalized the LTCH industry.

As we have stated most recently in the RY 2007 LTCH PPS final rule, we discussed a potential framework to update payments to LTCHs that would account for appropriate factors that affect efficient delivery of services and care to Medicare beneficiaries (71 FR 27818), and we have solicited comments on the presentation of a model for such a framework presented in Appendix A of that final rule. Presently, however, in the absence of a more comprehensive update framework, we believe that it is necessary and appropriate for us to evaluate the need of applying an adjustment to the full market-basket increase, based upon the best available data and policy considerations. Similarly, we believe it is appropriate to update the LTC-DRG relative weights based on the latest available data because the more recent data ensure that the LTC DRG relative weights for FY 2007 best reflect the resources actually used in the treatment of LTCH patients.

Comment: Several commenters discussed the impact of policies that we proposed under the IPPS for short-term, acute care hospitals (that is, the adoption of severity-adjusted DRGs; and the implementation of HSRVcc (cost-based weights) methodology for calibration of DRG weights) in their evaluation of the proposed 1.4 decrease in the LTC-DRG payments based on the proposed LTC-DRG changes for FY 2007. Both commenters urged us not to implement the proposed LTC-DRG relative weights because they believe that the discussion of the severity-adjusted DRGs in the proposed rule emphasized the fact that the LTC-DRG classifications, as they currently exist, do not accurately capture the full measure of severity for LTCH patients.

One commenter commissioned a study by the Lewin Group that utilized claims data from the FY 2005 MedPAR file and cost report data from FY 2003 to simulate the HSRVcc methodology set forth in the proposed rule. The commenter stated that the result was that, rather than a estimated 1.4 percent payment reduction, the HSRVcc method of determining LTC-DRG relative weights resulted in an estimated 1.5 percent increase in LTCH PPS payments. The commenter added that this indicates that there can be reasonable differences as to what is the most accurate method of establishing relative weights under PPSs and that the Secretary should adjust the LTC-DRG weights this year on a budget-neutral basis, thus eliminating the estimated 1.4 percent decrease based on the proposed LTC-DRG relative weights. The commenter recommended that, although the authorizing legislation contemplates that CMS use the most recently available LTC-DRG weights for an annual update, the Secretary could use his broad authority to modify the LTC-DRG payments, as appropriate, and in order to accurately reflect current LTCH patient care. The commenter believed that the FY 2006 LTC-DRG relative weights should be maintained for FY 2007 because they more accurately account for the expected resources to be used by LTCH patients in FY 2007.

Another commenter noted that, based on the discussion in the FY 2007 proposed rule, CMS believes that severity-adjusted DRGs would improve the accuracy of the DRG system under the IPPS, and consequently, the commenter believed that, for FY 2008, severity-adjusted LTC-DRGs could be considered because they may better account for differences in severity of illnesses and associated costs across hospitals. This commenter further stated that higher weighted LTC-DRGs (and the LTCHs that treat them) are more vulnerable to the payment reductions proposed for FY 2007 based on proposed LTC-DRG relative weights because payment rates for higher acuity LTCH patients will be diluted by the FY 2005 upcoding of many lower severity cases to the higher weighted DRGs. In addition, the commenter pointed to the revised short-stay outlier policy established in the RY 2007 LTCH PPS final rule which, they believe, is intended to reduce the number of lower-acuity patients being treated in LTCHs, and stated that those LTCH patients that are not short-stay outlier cases will be more typical of LTCH patients and, therefore, have higher acuity. The same commenter also mentioned that the FY 2005 data that are being proposed to be used to reweight the LTC-DRGs for FY 2007 represent a system “still in flux” because the system is still transitioning to full payment under the LTCH PPS and only a portion of each case is being paid based on LTC-DRGs. For these reasons, the commenter urged CMS to postpone further LTC-DRG rate reductions and instead recommended that CMS address coding improvements comprehensively in FY 2008 under the LTCH PPS in the context of the improved severity measures proposed under the IPPS for FY 2008 (or earlier).

Response: We understand that the commenters are concerned with the 1.4 percent decrease in estimated aggregated LTCH PPS payments for FY 2007 due to the proposed reweighting of the LTC-DRGs. We also understand that the commenters believe that the adoption of a severity-adjusted patient classification system under the LTCH PPS applied to the LTC-DRGs and the use of cost-based weights (HSRVcc) methodology could result in a different estimated aggregate payment change for FY 2007. However, as we discussed in greater detail below, we do not agree that the FY 2006 relative weights would more accurately represent resource use by LTCH patients for FY 2007 and that it would be necessary or appropriate to postpone the finalization of the annual reweighting of the LTC-DRGs. The current (FY 2006) LTC-DRG relative weights were determined based on FY 2004 LTCH claims data from the MedPAR files. For FY 2007, we proposed to use our existing relative weight methodology (established when the LTCH PPS was implemented for FY 2003) and FY 2005 LTCH claims data from the MEDPAR files to recalibrate the LTC-DRG relative weights, as these were currently the most recent complete LTCH claims data. As was proposed, for this final rule, we are using the March 2006 update of the FY 2005 MedPAR files because this is currently the most recent and complete LTCH claims data. We believe that the FY 2005 data are the best LTCH data available that reflect LTCHs' current treatment practice and coding patterns. Therefore, because the FY 2005 LTCH claims data better reflects current LTCH behavior than the FY 2004 LTCH claims data that was used to determine the FY 2006 LTC-DRG relative weights, we believe that using this updated (FY 2005) LTCH claims data with our existing relative weight methodology will result in LTC-DRG relative weights for FY 2007 that will best reflect the resources actually utilized by LTCHs in treating their Medicare patients.

With respect to the accuracy of the current LTCH-DRG system, we note the following. For FY 2003, we decided to adopt the current LTC-DRG system stating, “the LTC-CMS-DRG system is a system that is familiar to hospitals because it is based on the current DRG system under the acute care hospital inpatient prospective payment system. We believe that the familiarity of the LTC-CMS-DRG model may best facilitate the transition from the reasonable cost-based system to the prospective payment system as well as providing continuity in payment methodology across related sites of care (for example, an acute care hospitalization for a patient with a chronic condition)” (67 FR 55966). However, we have noted that we believed that there may be significant advantages in the use of severity-adjusted LTC-DRGs. In fact, when we were developing the LTCH PPS for FY 2003, we seriously considered using a specially modified version of the APR-DRGs (67 FR 55966-55967). At that time, we stated:

“The LTC-APR-DRGs, a condensed version of 3M's all-patient refined DRGs (APR-DRGs) for acute care hospitals, was developed by 3M Health Information Systems, for exclusive use in LTCHs. The LTC-APR-DRG system was designed to reflect the clinical characteristics of LTCH patients. This case-mix classification model contains 26 base LTC-APR-DRGs, subdivided by 4 severity of illness levels to yield 104 classification levels. In this system, the patient's secondary diagnoses, their interaction, and their clinical impact on the primary diagnosis determine the severity level assigned to each of the 26 LTC-APR-DRGs” (67 FR 55966).

When we decided to use the same patient classification system as the IPPS, following a comprehensive analysis of both the LTC-APR-DRGs and the existing DRG system (modified by the use of quintiles for low volume DRGs) for the particular purposes of patient classification at LTCHs, we indicated that we believed that either classification system would result in appropriate payments for LTCHs under the PPS. However, we noted several issues to consider concerning the LTC-APR-DRG system, including—

“* * * its complexity, its clinical subjectivity, and its utility as it relates to other Medicare prospective payment systems. The LTC-APR-DRG model provides a clinical description of the population of LTCHs, patients exhibiting a range of severity of illness with multiple comorbidities as indicated by secondary diagnoses. The clinical interaction of the primary diagnosis with these comorbidities determines the severity level of the primary diagnoses, resulting in the final assignment to a LTC-APR-DRG by the GROUPER software designed for this system” (67 FR 55966).

We further noted that “* * * determining whether particular comorbidities increase the cost of a case for a LTCH patient is complicated by the nature of the clinical characteristics of these patients. More specifically, many LTCH patients have numerous conditions that may not all be relevant to the cost of care for a particular discharge. Although the patient actually has a specific condition, including this condition among secondary diagnoses coded under the LTC-APR-DRG system may assign an inaccurate severity level to the primary diagnosis and result in inappropriate LTC-APR-DRG payment. We also believe that reliance on existing comorbidity information submitted on LTCH bills could result in significant variation in the assignment of the specific LTC-APR-DRGs” (67 FR 55967).

We concluded our explanation in the FY 2003 final rule for the LTCH PPS by stating that “[e]ven though we are using LTC-DRGs in the LTCH prospective payment system in this final rule, we may have the opportunity to propose a severity-adjusted patient classification for LTCHs in the future, particularly if the acute care hospital inpatient prospective payment system moves in this direction” (67 FR 55967). As we noted in the FY 2007 IPPS proposed rule, if and when a severity-adjusted patient classification system is adopted under the IPPS, we would need to consider whether to propose revisions to the patient classification system used under the LTCH PPS. Any proposed changes to the patient classification system would be done through notice and comment rulemaking (71 FR 24051). Subsequently, in 2005, MedPAC recommended we refine the entire inpatient acute care CMS DRG system to take into account severity of illness and apply HSRV weights to DRGs. However, we believe that it is advantageous to the LTCH community to wait for CMS to first finalize its policies regarding any refinements to the DRG system for the IPPS so that we can fully analyze what the effects of such changes would be on LTCH PPS payments. To the extent any changes for severity-adjusted DRGs for the IPPS system have been finalized, an analysis could then be performed to determine whether it is appropriate to propose the same severity-adjusted patient classification for LTCHs. As we stated in the FY 2007 IPPS proposed rule:

“At that time, we would need to consider whether to propose revisions to the patient classification system under the LTCH PPS. Any proposed changes to the patient classification system would be done through notice and comment rulemaking” (71 FR 24051).

The commenters cited the virtues of the severity-adjusted DRGs and one commenter commissioned the above described study to assess the validity of our proposed update to the LTC-DRG relative weights for FY 2007. In response to these comments, we reiterate that, while we understand that applying the severity-adjusted DRGs under the LTCH PPS could have an impact on setting relative weights used in determining LTCH PPS payments, we would consider their use in the LTCH PPS after we evaluate any DRG refinements for the IPPS, as noted above.

We note that while severity-adjusted DRGs had been proposed under the IPPS system for FY 2008 (or earlier), we did not propose to revise the current patient classification system used under the LTC PPS. Because, as we explained above, we believe any refinement due to severity-adjusted DRGs for the IPPS system would need to be evaluated to determine whether it is appropriate to use the same severity-adjusted DRGs for LTCHs, we will, at that time, take into consideration such issues as the impact of treating higher acuity patients.

We have noted that some commenters believe it is not appropriate that LTCHs be impacted by decreasing payments because of the upcoding of lower acuity patients to higher weighted LTC-DRGs, as discussed in the previous responses. However, as we discussed in the FY 2007 IPPS proposed rule (71 FR 24413), many of the LTC-DRG relative weights proposed for FY 2007 are lower than the current (FY 2006) LTC-DRG relative weight because based on the latest available LTCH claims data, we continue to observe an increase in the number of relatively lower charge cases being assigned to LTC-DRGs with higher relative weights in prior years. As explained previously, we believe that using updated (FY 2005) LTCH claims data will result in LTC-DRG relative weights for FY 2007 that best reflect the resources actually utilized by LTCHs in treating their Medicare patients and thereby act to ensure appropriate LTCH PPS payments in FY 2007. The commenter is correct in noting that it was our intention, when we revised the short-stay outlier policy described above, to reduce the number or type of short-stay patients being treated in LTCHs that do not utilize the resources of “typical” LTCHs. Many of these very short stay cases require more appropriate treatment at another hospital setting, such as an acute care hospital. Therefore, we are not convinced that reducing the number of short stay patients treated at LTCHs will necessarily result in higher LTC-DRG weights in all LTC-DRGs or even in higher weighted LTC-DRGs.

Moreover, since the implementation of the LTCH PPS in FY 2003, we have accounted for very short-stay and short-stay outliers cases in our LTC-DRG relative weight methodology. Specifically, we have removed cases with a length of stay of 7 days or less because we believed that they could “significantly bias payments against inlier cases” (67 FR 55989). In addition, the methodology includes a step to adjust charges for the effects of short-stay outliers by “counting a short-stay outlier as a fraction of a discharge based on the ratio of the length of stay of the case to the average length of stay for the LTC-DRG.” Without this adjustment, we maintained at that time that we believed that “the relatively lower charges of the short-stay outlier cases bring down the average charge for all cases within a LTC-DRG * * * [and] result in an ‘underpayment’ to nonshort-stay outlier cases * * *” (67 FR 55990). Therefore, we do not believe that the changes that we have made in the short-stay outlier policy in the RY 2007 LTCH PPS final rule will affect the DRG weights because our methodology has always accounted for this potential effect so that a reduction in short-stay outlier cases will not necessarily result in a significant change to the DRG weights.

During the previous 4 years, while we phased in to full payments under the LTCH PPS, we have reweighted the LTC-DRGs, with the result that for the first year, there was an estimated negligible increase in average payments based upon the reweighting of the LTC-DRGs (FY 2004 + 0.4 percent) and a negligible decrease in estimated payments based on the LTC-DRG update in FY 2005 (FY 2005, −0.5 percent). For the subsequent 2 years, there were decreases (FY 2006, −4.2 percent; proposed FY 2007, −1.4 percent). Although the LTCH PPS has been evolving, we believe that using the updated (FY 2005) LTCH claims data with our existing relative weight methodology will result in LTC-DRG relative weights for FY 2007 that will best reflect the resources actually utilized by LTCHs in treating its Medicare patients since the FY 2005 data is the best LTCH data available that reflects LTCHs' current treatment practice and coding patterns. Therefore, we do not find it either necessary or appropriate to postpone the FY 2007 update of the LTC-DRG relative weights until we consider the adoption of a classification system with “improved severity measures.”

Comment: Numerous commenters suggested that CMS forgo the proposed approximately 1.4 percent decrease in estimated aggregate LTCH PPS payments and, instead, establish a policy of budget neutrality for the annual updates of the LTC-DRG relative weights. The commenters believed a policy of budget neutrality would mitigate the estimated LTCH PPS payment reductions that CMS estimates would result from the proposed changes to the LTC-DRGs and relative weights for FY 2007. MedPAC also endorsed adopting a policy of budget neutrality for the annual recalibration of the LTC-DRG weights and noted that the adoption of the budget neutrality process that CMS uses in recalibrating the annual weights for the IPPS for the LTCH would avoid the estimated decrease in payments of 1.4 percent for FY 2007.

One commenter asserted that the absence of a budget neutrality adjustment for the annual recalibration of the LTC-DRGs provides a negative incentive for efficiency, because assigning cases that appropriately use fewer hospital resources to a particular LTC-DRG will result in a lower weight for that LTC-DRG. Therefore, the commenter urged CMS not to implement the proposed reweighting for FY 2007 prior to a full analysis of the impact of the proposed reweighting along with other payment policy changes provided in the RY 2007 LTCH PPS on the overall adequacy of payments to LTCHs. In addition, the commenters expressed eagerness to review the recommendations currently under development by RTI International for patient and facility criteria for LTCHs. Several commenters further suggested that no additional reimbursement reductions under the LTCH PPS should be imposed until the RTI report is complete and the industry works with CMS to implement its findings.

Response: We understand that the commenters are concerned with the estimated decrease in payments under LTCH PPS based upon the changes in the LTC DRGs and relative weights proposed for FY 2007. However, as discussed above, we are not postponing the proposed FY 2007 reclassification and recalibration of the LTC-DRGs. In addition, the payment policies that were finalized in the RY 2007 LTCH PPS final rule, such as the zero update to the standard Federal rate and the revised short-stay outlier policy, will be effective for LTCH discharges beginning on July 1, 2007, as established in that rule.

We further acknowledge that the commenters and also MedPAC are urging us to establish a budget neutrality requirement for the annual reclassification and recalibration of the LTC-DRGs so that, in future years, the LTCH PPS would avoid an estimated decrease in aggregate payments such as the estimated 1.3 percent based on the LTC-DRG weights that we are finalizing for FY 2007.

In the responses to comments addressed above, we have noted several reasons for the annual fluctuations in LTC-DRG relative weights that resulted in an estimated increase in aggregate payments for FY 2004, a negligible estimated decrease in aggregate payments for FY 2005, and decreases in aggregate payments for FYs 2006 and 2007. We reiterate that the LTCH PPS has existed since FY 2003, and we believe that several factors are occurring that affect the changes to the relative weights, including actual improvements in coding so that cases are appropriately assigned to LTC-DRGs. Each year, we recalibrate the LTC-DRG relative weights based on the most recent available LTCH claims data, which reflect current LTCH patient mix and coding practices. The annual recalibration of the LTC-DRG relative weights to which LTCH cases are assigned will appropriately reflect more or less resource use than the previous year's LTC-DRG relative weights.

We understand the concerns expressed by the commenters regarding this fiscal year's estimated decrease in payments based upon the proposed (and finalized) FY 2007 reweighting of the LTC-DRGs. However, we remind the commenters that establishing a budget-neutrality policy for the LTC-DRG weights would have precluded the increase in payments that occurred during FY 2004 as well as any increase that an analysis of future data may warrant.

Under the IPPS, there is a statutory requirement in section 1886(d)(4)(C)(iii) of the Act that requires that, beginning with FY 1991, reclassification and recalibration changes be made in a manner that assures that the aggregate payments are neither greater than nor less than the aggregate payments that would have been made without the changes. However, there is no statutory or regulatory requirement that the annual update to the LTC-DRG classifications and relative weights be done in a budget neutral manner. In addition, after FY 2003, the year that the LTCH PPS was implemented, there was no statutory requirement for budget neutrality for any component of the LTCH PPS.

However, as we have already noted, the LTCH PPS, having been first implemented for cost reporting periods beginning on or after the start of FY 2003, will soon end its transition period and payment will be based solely on the Federal rate with cost reporting periods beginning in FY 2007. In the RY 2007 LTCH PPS final rule, we provided that we would reevaluate all payment adjustments that were originally considered for the LTCH PPS prior to its implementation and also determine the appropriateness of a one-time prospective adjustment to the standard Federal rate (§ 412.523(d)(3)) so that the effect of any significant differences between actual payments and estimated payments for the first year of the LTCH PPS would not be perpetuated in the PPS for future years. Given the considerable discretion granted to the Secretary under the BBRA of 1999 and the BIPA of 2000 to develop the LTCH PPS, it is possible, however, that at the same time, the Secretary would consider using his broad authority to establish a policy of budget neutrality for the annual update of the LTC-DRG classifications and relative weights. As noted above, currently the best available LTCH data (FY 2005) are from the second full year of the PPS, and LTCHs may still be modifying their behavior to the change in payment methodology. If, upon reevaluation of our payment policies based on future LTCH data as the data become available, we find that it would be appropriate to propose making the updates to the LTC-DRGs and relative weights in a budget neutral manner, the public will have the opportunity to submit comments on any proposed change during the rulemaking process.

The commenters mentioned their eagerness to review the recommendations currently being developed by RTI regarding the feasibility of patient and facility level admissions criteria for LTCHs. We anticipate that RTI will submit its final report and recommendations during RY 2007. We place considerable importance on RTI's work, and we will encourage a dialogue with the public based on the report. We note that, while we believe the report will have a substantial impact on future Medicare policy for LTCHs, we still believe that the retention of many of the specific payment adjustment features of the LTCH PPS presently in place and the development of additional or revised adjustments may still be both necessary and appropriate for purposes of protecting the integrity of the Medicare trust fund.

Comment: Several commenters believed that the changes to the LTC-DRG relative weights will have a more significant impact on high case-mix providers than on low-case mix providers. One commenter referred to a LTCH which, as a high acuity provider, will experience an approximate 5 percent drop in total case mix index. This commenter requested that CMS make a weighted average calculation available when it publishes the impacts of changes in the relative weights. The commenter further suggested that CMS produce an impact statement focusing on changes across all DRGs that will enable providers to understand the impacts on their individual LTCHs.

Response: We believe we published a comprehensive description of the impact of the reweighting of the LTC-DRGs for FY 2007 in the proposed rule (71 FR 24413). Specifically, in section VII, Effects of Other Proposed Policy Changes, in subsection A, under the heading, Effects of LTC-DRG Reclassifications and Relative Weights for LTCHs, we included a detailed analysis of the impact that would result from our proposals.

In that section, we stated: “When we compared the GROUPER Version 23.0 (FY 2006) LTC-DRG relative weights to the proposed GROUPER Version 24.0 (FY 2007) proposed LTC-DRG relative weights, we found that approximately 62 percent of the LTC-DRGs would have a higher relative weight under Version 23.0, while the remaining approximately 38 percent of the LTC-DRGs would have a higher relative weight under Version 24.0. We also found that, based on FY 2005 LTCH cases, the GROUPER Version 23.0 LTC-DRG relative weights were, on average, approximately 3.1 percent higher than the proposed GROUPER Version 24.0 LTC-DRG relative weights. In addition, based on an analysis of the most recent available LTCH claims data from the FY 2005 MedPAR file, we continue to observe that the average proposed LTC-DRG relative weight decreases due to an increase of relatively lower charge cases being assigned to LTC-DRGs with higher relative weights in the prior year.

Contributing to this increase in these relatively lower charge cases being assigned to proposed LTC-DRGs with higher relative weights in the prior year are improvements in coding practices, which are typical when moving from a reasonable cost-based payment system to a PPS. The impact of including additional cases with relatively lower charges in LTC-DRGs that had a relatively higher relative weight in the GROUPER Version 23.0 (FY 2006) is a decrease in the average relative weight for those LTC-DRGs in the proposed GROUPER Version 24.0. As noted above in section II.F. of the preamble to this proposed rule, LTCHs are a specialized provider type that typically do not treat a broad spectrum of patients in their facilities with many different diagnoses. While there are 526 valid proposed GROUPER Version 24.0 LTC-DRGs, 191 LTC-DRGs have no LTCH cases. In addition, another 173 LTC-DRGs are categorized as ‘low volume’ (that is, have less than 25 cases annually). Consequently, only about 162 LTC-DRGs are used by most LTCHs on a ‘regular basis’ (that is, nationally LTCHs discharge, in total, an average of 25 or more of these cases annually).

Of these 162 LTC-DRGs that are used on a regular basis, we found that approximately 60 percent of the LTC-DRGs would have higher relative weights under GROUPER Version 23.0 in comparison to proposed GROUPER Version 24.0, and the remaining 40 percent of the 162 LTC-DRGs that are used on a ‘regular basis’ would have higher relative weights under proposed GROUPER Version 24 in comparison to GROUPER Version 23.0. In addition, about 25 percent of the 162 LTC-DRGs that are used on a ‘regular basis’ would experience a decrease in the average charge per case as compared to the average charge per case in that DRG based on FY 2004 data, which generally results in a lower relative weight. Moreover, of the 162 LTC-DRGs that are used on a ‘regular basis,’ approximately 63 percent of those LTC-DRGs would experience a change in the average charge per case from FY 2004 LTCH data as compared to FY 2005 LTCH data that is less than the increase in overall average LTCH charges across all LTC-DRGs from FY 2004 to FY 2005 of about 8.3 percent. Accordingly, those LTC-DRGs would also have a proposed reduction in their relative weight as compared to the relative weight in FY 2006. For those LTC-DRGs in which the average charge within the LTC-DRG increase is less than 8.3 percent, the proposed relative weights for those LTC-DRGs would decrease because the average charge for each of those LTC-DRGs is being divided by a larger number (that is, the average charge across all LTC-DRGs). For the reasons discussed above, we believe that the proposed changes in the LTC-DRG relative weights, which include a significant number of LTC-DRGs with lower proposed relative weights, would result in approximately a 1.4 percent decrease in estimated aggregate LTCH PPS payments” (71 FR 24413).

The above paragraphs, published in the FY 2007 IPPS proposed rule, clearly indicated the impact of the reweighting of the LTC-DRGs. All of the impact percentages listed are “weighted averages,” as was the proposed estimated 1.4 percent decrease in aggregate LTCH PPS payments. That is, all LTCH cases in the December 2005 update of the FY 2005 MedPAR file were used to determine the LTC-DRG impact figures presented in the FY 2007 IPPS proposed rule. Therefore, the latest data on the types of patients treated across all LTCHs were used to determine the impact and not just the proposed changes to the LTC-DRG weights. The proposed and final FY 2007 reweighting of the LTC-DRGs may indeed have a more significant impact on a high acuity provider because many of the proposed and final LTC-DRG weights in relatively high weighted LTC-DRGs would decrease compared to their current values. However, we also note that Medicare payments for several of the highest acuity LTC-DRGs have yielded substantial margins. For example, an analysis of MedPAR data from FY 2004 indicated that, for LTC-DRG 475 (Ventilator Support) with a relative weight of 2.1358 for FY 2004, average aggregate (dollar weighted) margins for all providers was 21.09 percent, and for LTC-DRG 87 (Pulmonary Edema/Respiratory Support) with a relative weight of 1.6513 for FY 2004, average aggregate margins were 26.93 percent. Even for cases requiring somewhat less resource intensity, such as LTC-DRG 416 (Septicemia) with a relative weight of 0.9191 for FY 2004, which is also one of the diagnoses most frequently found in LTCHs, the aggregate margin is 11.54 percent and for LTC-DRG 249 (After Care Musculoskeletal) with a relative weight of 0.7829 for FY 2004, the margin is 9.69 percent. Therefore, we believe that the reweighting of the LTC-DRGs for FY 2007, even for those high-acuity providers who experience a more significant impact, should not impede the efficient and effective delivery of care to Medicare beneficiaries, because, as described above, several of the highest-acuity LTC-DRGs have yielded substantial margins. Furthermore, even though the recalibration of the LTC-DRG relative weights will result in a decrease in the relative weight for some high-acuity LTC-DRGs, because the recalibration is based on the most recent available LTCH claims data (FY 2005), it ensures the most accurate payments for FY 2007 based on current LTCH treatment and coding practices.

In response to the commenter who requested impacts that reflected a weighted average calculation, as noted above, the impact of the proposed changes to the LTC-DRGs for FY 2007 presented in the FY 2007 IPPS proposed rule (71 FR 24413) are based on a weighted average calculation. That is, all FY 2005 LTCH cases in the December 2005 update of the MedPAR data were used to determine the impact figures presented in the proposed rule. This means that only the proposed changes to the relative weights for LTC-DRGs that had LTCH cases in those DRGs based on the FY 2005 LTCH data contributed to the impact. This continues to be true for the impact of the final LTC_DRG weights which are based on the most recent update of the FY 2005 MedPAR data. It also means that, for example, LTCH cases in LTC-DRG 475 represent approximately 12 percent of all LTCH cases in FY 2005 and therefore, 12 percent of the impact presented in the proposed rule was due to the proposed change in the LTC-DRG weight for LTC-DRG 475. We believe that the commenter may have mistakenly believed that we measured the impact of the proposed LTC-DRG changes based on the changes proposed for each LTC-DRG without accounting for the volume of LTCH cases treated in each LTC-DRG. In addition, we note that if a provider is eager to determine the specific impact of the annual proposed LTC-DRG reweighting on an individual LTCH or a particular weight, the provider needs only to compare an application of the LTC-DRG weights published in the previous year's final rule (Table 11) of its cases to the proposed LTC-DRG relative weights that are published in the current year's proposed rule (Table 11; 71 FR 24395-24403) as applied to the same set of cases.

Comment: Several commenters maintained that the proposed 1.4 percent decrease in aggregate payments to LTCHs due to the proposed LTC-DRG reclassification and recalibration for FY 2007, in addition to payment cuts established for RY 2007 represent a “misinterpretation” of MedPAC's recommendation in its March 2006 Report to the Congress for a zero update for LTCHs. MedPAC cited Medicare margins for 2004 of 9.0 for the LTCH industry and projected 7.8 percent margins for 2006, but the commenters believed that these projections did not factor in the impact of the “25 percent policy” for co-located LTCHs or the estimated payment reductions associated with the revised short-stay outlier policy.

Response: As we have noted elsewhere in earlier responses to comments, the estimated 1.4 percent decrease aggregate in LTCH PPS payments due to the proposed LTC-DRG reclassification and recalibration for FY 2007 is a data-driven result of the annual recalibration of the relative weights for LTC-DRGs based on the latest available LTCH claims data from the MedPAR files (FY 2005). Therefore, for FY 2007, based on the updated LTC-DRGs classifications and relative weights, estimated payments to LTCHs will be 1.3 percent less than they would have been based on the prior fiscal year's (that is, FY 2006) classifications and relative weights for the same LTCH cases. Similarly, LTCH claims data from the FY 2006 MedPAR files will be used to determine the proposed LTC-DRG relative weights for FY 2008, and the resulting aggregate LTCH PPS payments, absent a regulatory or statutory change implementing recalibration of relative weights in a budget neutral manner, may either decrease or increase, based upon the FY 2006 data and DRG classification changes. In setting the annual relative weights for the LTC-DRG system for FY 2007, we have followed the requirements established with the implementation of the LTCH PPS in FY 2003 (67 FR 55984-55995). Although the proposed recalibrated LTC-DRG relative weights were estimated to result in 1.4 percent decrease in LTCH PPS payments for FY 2007 (and based on final policies established in this final rule, the updated LTC-DRGs for FY 2007 are estimated to result in a 1.3 percent decrease in aggregate LTCH PPS payments for FY 2007, as noted above), we do not believe that this adjustment is relevant to MedPAC's recommendation for the zero percent update to the LTCH PPS Federal rate for RY 2007.

The annual LTC-DRG update is separate from the Federal rate update; specifically, their purposes are different and independent. The standard Federal base rate is an estimate of the national average cost per case which is adjusted by the LTC-DRG relative weights to reflect the resource consumption of the particular case; that is, a case with a relative weight of 2.0 is twice as costly/uses twice the resources as a case with a relative weight of 1.0. The LTC-DRG relative weights are recalibrated annually based on the most recent available LTCH data to reflect resources used by LTCHs in treating each type of case. The update to the Federal rate is to adjust the Federal rate to account for various adjustments to that rate, including inflation.

MedPAC's data analysis, in its March 2006 Report to the Congress, indicated that the average Medicare margin for LTCHs was 9.0 percent for FY 2005 and was projected at 7.8 percent for 2006. (As we stated in our RY 2007 LTCH PPS final rule, MedPAC also noted that “LTCH HwHs were found to have higher margins than freestanding LTCHs in RY 2005” (71 FR 27823).) Based on its analysis, MedPAC stated that “* * * evidence from the indicators we have examined suggests that LTCHs can accommodate the cost of caring for Medicare beneficiaries in 2007 without an increase in the base rate” (p. 218). Consistent with MedPAC's recommendation, after incorporating an adjustment to account for changes in coding practices that did not reflect “real” case-mix, we finalized a zero percent update for FY 2007 (71 FR 27819). As stated earlier, this adjustment is not a function of, or related to, the update to the relative weights for LTC-DRGs.

The commenters' also reference the 25-percent threshold payment adjustment for co-located LTCHs (§ 412.534) established beginning in FY 2005 and the newly revised short stay outlier payment policy (§ 412.529) beginning in RY 2007. We believe that the commenters are seeking to connect these adjustments, which are also estimated to result in a decrease in aggregate LTCH PPS payments in the absence of a change in admission practices by LTCHs to the estimated impact of the updated LTC-DRG relative weights. However, the policies cited by the commenters are not related to the impact of the updating of the LTCH-DRG relative weights, but each independently, furthers the goal of establishing fair and reasonable Medicare payments under the LTCH PPS.

The HwH “25 percent rule,” that is, the special payment provisions for LTCH HwHs and satellites, was established at § 412.534 in the FY 2005 IPPS final rule. Under that policy, we provide a payment adjustment for those patients discharged from co-located LTCHs (that is, HwHs and satellites) admitted from host hospitals that exceeded a specified threshold percentage (in most cases, 25 percent). Medicare patients who reach high-cost outlier status in the host hospital are excluded from the count of the percentage of patients admitted directly from the host. As we discussed in the FY 2005 IPPS final rule, when we implemented the “25 percent rule,” we were unable to estimate the impact of this policy because we anticipated behavioral changes by both the host and the co-located LTCHs resulting from the provision that exempted high-cost outliers from the percentage threshold calculation (69 FR 49771).

MedPAC further addressed this issue in the March 2006 Report, where it noted that it “* * * cannot foresee how HwHs/ behavior will change in response to this rule. CMS has discussed scenarios (CMS 2005). For example, patients admitted to an HwHs from the host hospital after becoming an outlier are not counted in the limit, thus HwHs may admit more outlier cases under this rule. Alternatively, host hospitals may discharge fewer patients to their HwHs because of constraints from the 25 percent rule, in which case HwHs' volume might fall. In cities where there is another LTCH, an acute care hospital might discharge patients to a different long-term care hospital than the one on its grounds. The Office of Inspector General or the QIOs may want to monitor acute care hospitals' and HwHs' behavior in response to the 25 percent rule. Because we have no evidence of how HwHs will react, we have not modeled margins incorporating this policy change.” (p. 218)

Because the policy at § 412.534 exempts patients admitted from the host hospital if they had already achieved high-cost outlier status under the IPPS, from the LTCHs' percentage threshold calculation (as noted above), we believe that even with some adjustments resulting in a decrease in payments to some co-located LTCHs, Medicare payments to co-located LTCHs on average will continue to exceed the Medicare costs of the inpatient hospital services provided to its patients, even with a zero percent update to the Federal rate for RY 2007 (71 FR 27823). Furthermore, we believe that the 25-percent threshold policy and the short stay outlier payment revision that that we have established, first for co-located LTCHs at § 412.534 for FY 2004 and the revisions to the short-stay outlier policies at § 412.529 that we finalized for RY 2007, each have a firm and consistent basis in our general policy considerations under the LTCH PPS.

As we noted in the RY 2007 final rule for the LTCH PPS, we do not believe that the change to the short-stay outlier policy will result in an adverse impact on LTCHs. As a result of the change to the short-stay outlier payment formula, we believe that LTCHs will have an incentive to significantly reduce the number of very short-stay cases that they admit. We believe that, by paying appropriately for short-stay outlier cases and by removing the financial incentive for LTCHs to admit those very short stay cases that could otherwise receive appropriate treatment at an acute care hospital (and paid under the IPPS), LTCHs will change their admission patterns for these patients. We further believe that payment decreases to LTCHs resulting from this policy would only occur if LTCHs were to continue to admit the same number of short-stay outlier patients with very short lengths of stay. We believe this policy is needed to assure that payments for short-stay outlier cases are appropriate.

Therefore, we disagree with the commenter that we have “misinterpreted” MedPAC's recommendation of a zero percent update for 2007 in our proposed update to the LTC-DRGs for FY 2007. We maintain that the rationale for each of the policy features mentioned by the commenter, when evaluated independently, is clear and reasonable. In addition, they are independent of the DRG recalibration that occurs every year based on an established formula. We strongly disagree with the allegations that their implementation represents a “misinterpretation” of MedPAC's margin analysis and recommendation (discussed above) in the March 2006 Report to the Congress. As discussed above, this update is not based on MedPAC's analysis and we believe that updating the LTC-DRG relative weights for FY 2007 based on FY 2005 LTCH claims data will result in more appropriate LTCH PPS payments since the relative resource intensity of each LTC-DRG (that is, the relative weight) will be determined from the most recent available LTCH data (FY 2005) reflecting LTCHs' current practice and treatment patterns.

Comment: Several commenters, including MedPAC, recommended that we adopt severity-adjusted DRGs as the patient classification system for the LTCH PPS. In particular, MedPAC analyzed FY 2004 CMS LTCH data using both standardized charges and standardized hospital-specific costs (removing the effect of local wages) using Version 23 (FY 2006) of the GROUPER and stated that, on a preliminary basis, CS DRGs are relatively homogeneous in resource use for the kinds of cases treated in LTCHs. They believed that this indicates that the CS DRGs proposed for IPPS hospitals may also “be promising for LTCHs.”

Response: We are aware of the heightened interest in severity-adjusted DRGs by the provider community, and in section II.C.6.of this final rule, we discuss the revisions that we are making to the DRG classifications structure for the IPPS and our expectations for adopting severity adjustments for DRGs under the IPPS in FY 2008. We appreciate the data analysis that MedPAC produced to demonstrate the potential utility of CS DRGs for classifying patients being treated in LTCHs. It is possible that the modified version of the APR DRGs or another severity-adjusted patient classification system may account for differences in severity of illness and associated costs among hospitals. In section II.C. of this preamble, we discuss the issues that we are dealing with respect to the adoption of a severity adjusted DRG system. Once we have addressed those issues under the IPPS, we would need to consider whether it is appropriate to propose similar revisions to the patient classification system under the LTCH PPS. As stated in the FY 2007 IPPS proposed rule, we would emphasize that any proposed changes to the patient classification system for LTCHs would be done through notice and comment rulemaking.

G. Add-On Payments for New Services and Technologies

1. Background

Sections 1886(d)(5)(K) and (L) of the Act establish a process of identifying and ensuring adequate payment for new medical services and technologies (sometimes collectively referred to in this section as “new technologies”) under the IPPS. Section 1886(d)(5)(K)(vi) of the Act specifies that a medical service or technology will be considered new if it meets criteria established by the Secretary after notice and opportunity for public comment. Section 1886(d)(5)(K)(ii)(I) of the Act specifies that the process must apply to a new medical service or technology if, “based on the estimated costs incurred with respect to discharges involving such service or technology, the DRG prospective payment rate otherwise applicable to such discharges under this subsection is inadequate.”

The regulations implementing this provision establish three criteria for new medical services and technologies to receive an additional payment. First, § 412.87(b)(2) defines when a specific medical service or technology will be considered new for purposes of new medical service or technology add-on payments. The statutory provision contemplated the special payment treatment for new medical services or technologies until such time as data are available to reflect the cost of the technology in the DRG weights through recalibration. There is a lag of 2 to 3 years from the point a new medical service or technology is first introduced on the market and when data reflecting the use of the medical service or technology are used to calculate the DRG weights. For example, data from discharges occurring during FY 2005 are used to calculate the FY 2007 DRG weights in this final rule. Section 412.87(b)(2) provides that a “medical service or technology may be considered new within 2 or 3 years after the point at which data begin to become available reflecting the ICD-9-CM code assigned to the new medical service or technology (depending on when a new code is assigned and data on the new medical service or technology become available for DRG recalibration). After CMS has recalibrated the DRGs, based on available data, to reflect the costs of an otherwise new medical service or technology, the medical service or technology will no longer be considered ‘new’ under the criterion for this section.”

The 2-year to 3-year period during which a medical service or technology can be considered new would ordinarily begin with FDA approval, unless there was some documented delay in bringing the product onto the market after that approval (for instance, component production or drug production had been postponed until FDA approval due to shelf life concerns or manufacturing issues). After the DRGs have been recalibrated to reflect the costs of an otherwise new medical service or technology, the special add-on payment for new medical services or technology ceases (§ 412.87(b)(2)). For example, an approved new technology that received FDA approval in October 2005 and entered the market at that time may be eligible to receive add-on payments as a new technology until FY 2008 (discharges occurring before October 1, 2007), when data reflecting the costs of the technology would be used to recalibrate the DRG weights. Because the FY 2008 DRG weights will be calculated using FY 2006 MedPAR data, the costs of such a new technology would likely be reflected in the FY 2008 DRG weights.

Section 412.87(b)(3) further provides that, to receive special payment treatment, new medical services or technologies must be inadequately paid otherwise under the DRG system. To assess whether technologies would be inadequately paid under the DRGs, we establish thresholds to evaluate applicants for new technology add-on payments. In the FY 2004 IPPS final rule (68 FR 45385, August 1, 2003), we established the threshold at the geometric mean standardized charge for all cases in the DRG plus 75 percent of 1 standard deviation above the geometric mean standardized charge (based on the logarithmic values of the charges and transformed back to charges) for all cases in the DRG to which the new medical service or technology is assigned (or the case-weighted average of all relevant DRGs, if the new medical service or technology occurs in many different DRGs). Table 10 in the Addendum to the FY 2004 IPPS final rule (68 FR 45648) listed the qualifying threshold by DRG, based on the discharge data that we used to calculate the FY 2004 DRG weights.

However, section 503(b)(1) of Pub. L. 108-173 amended section 1886(d)(5)(K)(ii)(I) of the Act to provide for “applying a threshold * * * that is the lesser of 75 percent of the standardized amount (increased to reflect the difference between cost and charges) or 75 percent of 1 standard deviation for the diagnosis-related group involved.” The provisions of section 503(b)(1) apply to classification for fiscal years beginning with FY 2005. We updated Table 10 from the Federal Register document that corrected the FY 2004 final rule (68 FR 57753, October 6, 2003), which contained the thresholds that we used to evaluate applications for new service or technology add-on payments for FY 2005, using the section 503(b)(1) measures stated above, and posted these new thresholds on our Web site at: http://www.cms.hhs.gov/AcuteInpatientPPS/08_newtech.asp. In the FY 2005 IPPS final rule (in Table 10 of the Addendum), we included the final thresholds that were being used to evaluate applicants for new technology add-on payments for FY 2006. (Refer to section IV.D. of the preamble to the FY 2005 IPPS final rule (69 FR 49084, August 11, 2004) for a discussion of a revision of the regulations to incorporate the change made by section 503(b)(1) of Pub. L. 108-173.) Table 10 of the Addendum to the FY 2006 final rule (70 FR 47680) contained the final thresholds that are being used to evaluate applications for new technology add-on payments for FY 2007.

Section 412.87(b)(1) of our existing regulations provides that a new technology is an appropriate candidate for an additional payment when it represents “an advance that substantially improves, relative to technologies previously available, the diagnosis or treatment of Medicare beneficiaries.” For example, a new technology represents a substantial clinical improvement when it reduces mortality, decreases the number of hospitalizations or physician visits, or reduces recovery time compared to the technologies previously available. (Refer to the September 7, 2001 final rule (66 FR 46902) for a complete discussion of this criterion.)

The new medical service or technology add-on payment policy provides additional payments for cases with high costs involving eligible new medical services or technologies while preserving some of the incentives under the average-based payment system. The payment mechanism is based on the cost to hospitals for the new medical service or technology. Under § 412.88, Medicare pays a marginal cost factor of 50 percent for the costs of a new medical service or technology in excess of the full DRG payment. If the actual costs of a new medical service or technology case exceed the DRG payment by more than the 50-percent marginal cost factor of the new medical service or technology, Medicare payment is limited to the DRG payment plus 50 percent of the estimated costs of the new technology.

The report language accompanying section 533 of Pub. L. 106-554 indicated Congressional intent that the Secretary implement the new mechanism on a budget neutral basis (H.R. Conf. Rep. No. 106-1033, 106th Cong., 2nd Sess. at 897 (2000)). Section 1886(d)(4)(C)(iii) of the Act requires that the adjustments to annual DRG classifications and relative weights must be made in a manner that ensures that aggregate payments to hospitals are not affected. Therefore, in the past, we accounted for projected payments under the new medical service and technology provision during the upcoming fiscal year at the same time we estimated the payment effect of changes to the DRG classifications and recalibration. The impact of additional payments under this provision was then included in the budget neutrality factor, which was applied to the standardized amounts and the hospital-specific amounts.

Section 1886(d)(5)(K)(ii)(III) of the Act, as amended by section 503(d)(2) of Pub. L. 108-173, provides that there shall be no reduction or adjustment in aggregate payments under the IPPS due to add-on payments for new medical services and technologies. Therefore, add-on payments for new medical services or technologies for FY 2005 and later years have not been budget neutral.

Applicants for add-on payments for new medical services or technologies for FY 2008 must submit a formal request, including a full description of the clinical applications of the medical service or technology and the results of any clinical evaluations demonstrating that the new medical service or technology represents a substantial clinical improvement, along with a significant sample of data to demonstrate the medical service or technology meets the high-cost threshold, no later than October 15, 2006. Applicants must submit a complete database no later than December 30, 2006. Complete application information, along with final deadlines for submitting a full application, will be available at our Web site: http://www.cms.hhs.gov/AcuteInpatientPPS/08_newtech.asp. To allow interested parties to identify the new medical services or technologies under review before the publication of the proposed rule for FY 2008, the Web site will also list the tracking forms completed by each applicant.

2. Public Input Before Publication of a Notice of Proposed Rulemaking on Add-On Payments

Section 1886(d)(5)(K)(viii) of the Act, as amended by section 503(b)(2) of Pub. L. 108-173, provides for a mechanism for public input before publication of a notice of proposed rulemaking regarding whether a medical service or technology represents a substantial clinical improvement or advancement. The process for evaluating new medical service and technology applications requires the Secretary to—

• Provide, before publication of a proposed rule, for public input regarding whether a new service or technology represents an advance in medical technology that substantially improves the diagnosis or treatment of Medicare beneficiaries.
• Make public and periodically update a list of the services and technologies for which applications for add-on payments are pending.
• Accept comments, recommendations, and data from the public regarding whether a service or technology represents a substantial clinical improvement.
• Provide, before publication of a proposed rule, for a meeting at which organizations representing hospitals, physicians, manufacturers, and any other interested party may present comments, recommendations, and data regarding whether a new service or technology represents a substantial clinical improvement to the clinical staff of CMS.

In order to provide an opportunity for public input regarding add-on payments for new medical services and technologies for FY 2007 before publication of the FY 2007 IPPS proposed rule, we published a notice in the Federal Register on December 23, 2005 (70 FR 76315) and held a town hall meeting at the CMS Headquarters Office in Baltimore, MD, on February 16, 2006. In the announcement notice for the meeting, we stated that the opinions and alternatives provided during the meeting would assist us in our evaluations of applications by allowing public discussions of the substantial clinical improvement criterion for each of the FY 2007 new medical service and technology add-on payment applications before the publication of the FY 2007 IPPS proposed rule.

Approximately 35 participants registered and attended the town hall meeting in person, while additional participants listened over an open telephone line. The participants focused on presenting data on the substantial clinical improvement aspect of their products, as well as the need for additional payments to ensure access to Medicare beneficiaries. In addition, we received written comments regarding the substantial clinical improvement criterion for the applicants. We considered these comments in our evaluation of each new application for FY 2007 in the proposed rule and in this final rule. We have summarized these comments or, if applicable, indicated that no comments were received, at the end of the discussion of the individual applications.

We received two general comments about application of the newness and substantial clinical improvement criteria.

Comment: One commenter encouraged CMS to amend the definition of substantial clinical improvement for the IPPS new technology provision to conform to the OPPS definition of substantial clinical improvement used in 2001. Specifically, AdvaMed requested that after “decreased pain, bleeding, or other quantifiable symptom,” CMS should insert the following language: “such as convenience, durability, ease of operation or make other improvements in quality of life.”

Response: We believe we addressed this concern in the FY 2006 IPPS final rule (70 FR 47360). We use similar standards to evaluate substantial clinical improvement in the IPPS and OPPS and, in both systems, we employ identical language to explain and elaborate on the kinds of considerations that are taken into account in determining whether a new technology represents a substantial clinical improvement. We do not believe a change to the regulations text is necessary.

Comment: One commenter suggested that CMS should not use “substantial similarity” to evaluate newness without also determining whether the product is a substantial clinical improvement. The commenter argued that CMS is applying a concept that is not defined in regulations. If CMS applies the concept as part of determining whether a product is new without evaluating substantial clinical improvement, the commenter recommended that CMS should define substantial similarity through notice and comment rulemaking.

Response: We addressed this comment in the FY 2006 IPPS final rule (70 FR 47350 through 47351). We refer readers to that final rule for a detailed response to this comment.

Section 1886(d)(5)(K)(ix) of the Act, as added by section 503(c) of Pub. L. 108-173, requires that, before establishing any add-on payment for a new medical service or technology, the Secretary shall seek to identify one or more DRGs associated with the new technology, based on similar clinical or anatomical characteristics and the costs of the technology and assign the new technology into a DRG where the average costs of care most closely approximate the costs of care using the new technology. No add-on payment will be made if the new technology is assigned to a DRG that most closely approximates its costs.

At the time an application for new technology add-on payments is submitted, the DRGs associated with the new technology are identified. We only determine that a new DRG assignment is necessary or a new technology add-on payment is appropriate when the payment under these currently assigned DRGs is not adequate and the technology otherwise meets the newness, cost, and substantial clinical improvement criteria.

In this final rule, we evaluate whether new technology add-on payments will continue in FY 2007 for the three technologies that currently receive such payments. In addition, we present our evaluations of three applications for add-on payments in FY 2007.

Comment: One commenter stated that section 503 of Pub. L. 108-173 provided new funding for new technology add-on payments by no longer requiring that these payments be budget neutral. The commenter stated that this provision was enacted to ensure that the IPPS would better account for new drugs, devices, and services. However, the commenter believed that CMS continues to resist approval of new technologies and considers only a few technologies a year for add-on payments.

Another commenter called upon CMS to be more willing to indicate its preliminary views regarding whether a new technology application meets the criteria for add-on payments in the proposed rule. The commenter expressed particular concern that CMS had not given a strong indication of whether any of the initial new technology applications would meet the substantial clinical improvement criterion and noted that doing so would enhance stakeholder dialogue with CMS on the evaluation of the new technology criteria during the comment period.

Another commenter believed that CMS' definition of new technology is contrary to the statute. The commenter explained that CMS uses the FDA approval date to determine newness while the statute clearly requires that new technology add-on payments begin on the date an ICD-9-CM code is issued. The commenter urged CMS to use the date an ICD-9-CM code is issued to determine whether a technology is new instead of the FDA approval date.

Response: With respect to the comment that CMS resists approval of new technologies and considers only a few technologies a year for add-on payments, we note that we encourage companies with new-technologies that believe that they may meet the new technology criteria to apply for add-on payments. In our view, we have not resisted approving new technologies or been overly stringent in our application of the criteria. Our review of new technology focuses on the merits of the application and the requirements under the statute. The experience of our review process indicates that a significant number of new technologies have met the criteria. In fact, we have approved over 50 percent (6 of 11) of applications where we had to apply judgment about whether the technology met the criteria for an add-on payment. From FY 2003 to FY 2006, we received a total of 25 applications, but only 21 were unique (four applicants applied twice in subsequent years for the same technology). Of the applications that we received, 8 were already beyond the timeline to be considered new, 1 had not received FDA approval, and 1 did not meet the cost criterion. In our view, we denied these applications using objective criteria and without having to apply any subjective judgment. Of the remaining 11 applications, 6 were approved for new technology add-on payment, while the other 5 were not approved because we determined that these applications were not substantially different from older technologies or did not meet the substantial clinical improvement criterion. Therefore, to date, we have approved over 50 percent of applications where we needed to apply judgment about whether a new technology met the criteria for an add-on payment. These statistics obviously reflect the recent experience of new technology applications, and, depending on the ability of applications to meet the criteria in the future, will likely change. We note that the merits of each application determine whether it should be approved. The aggregate statistics reflect the ability of applicants to satisfy the criteria, and should not be construed as a measure of the appropriateness of the review process. We also note that over the years, the cost criterion has been lowered, giving applicants a lower threshold to meet the cost criterion. We encourage and welcome additional applications in future years so that we can continue to make payments for those technologies that meet the criteria and warrant new technology add-on payments.

With respect to the comment that CMS should be more willing to indicate our preliminary views regarding whether a new technology meets the criteria for an add-on payment, we provided our initial concerns regarding the two pending applications in the proposed rule. For the C-Port® System, we described our concerns about both the newness (“various forms of surgical staples and clips have been used for more than a decade in a wide range of surgical procedures”) and substantial clinical improvement (“the applicant submitted evidence suggesting that device does not always produce reliable anastomoses”) criteria and also indicated that the device appears to meet the cost threshold (71 FR 24071). Similarly, for the X STOP Interspinous Process Decompression System, we indicated our belief that “the device satisfies the newness and cost threshold criteria” and described our concerns about substantial clinical improvement (71 FR 24072). As a result of information provided in the proposed rule, the applicants were afforded the opportunity to address the specific concerns we raised. For example, the applicant for the C-Port® system was able to address our concerns about similarity to predicate devices to allow us to determine that the device meets the newness criterion. Similarly, the applicant for X STOP was able to address the concerns we raised in the proposed rule about whether the device meets the substantial clinical improvement criterion during the comment period.

Finally, with respect to the comment that CMS should use the issuance of an ICD-9-CM code as the date on which “newness” would begin, we have addressed this issue several times before, including in the FY 2005 IPPS final rule (69 FR 49002) and the FY 2006 IPPS final rule (70 FR 47343).

Comment: One commenter proposed that CMS allow manufacturers to apply for a new technology add-on payment on an ongoing basis and recommended that the agency issue quarterly updates announcing the approval of new technology add-on payments, similar to the outpatient setting.

Response: Section 1886(d)(5)(K)(i) of the Act requires that new technology add-on payments be established after notice and opportunity for public comments (in the publication required by subsection (e)(5) for a fiscal year or otherwise). In addition, pursuant to section 1886(d)(5)(K)(viii) of the Act, we are also required to hold an annual town hall meeting prior to the IPPS proposed rule to obtain public input about whether a new technology meets the substantial clinical improvement criterion. Given the requirements in the statute, it is not feasible to process applications on a quarterly basis.

Comment: Some commenters expressed disappointment that CMS has not increased the payment rate for new technology add-on payments from a maximum of 50 percent to a maximum of 80 percent of the marginal cost factor of the new medical service or technology, consistent with the outlier payment methodology. The commenters stated that increasing the marginal cost factor from 50 percent to 80 percent would offer some stability and consistency for hospitals thus enabling hospitals to more easily provide their patients access to new technologies.

Other commenters noted that CMS has approved so few technologies for new technology add-on payments that it would make more sense to compensate hospitals with a full add-on payment by paying on a cost basis using the average sales price plus six percent for FDA approved drugs and biologicals and list price plus a percentage for devices. The commenters believed that such a payment methodology would ensure that, “providers recoup their costs, Medicare pays a fair rate, and that payment is harmonized across treatment settings.” Finally, one commenter requested that CMS provide clear guidance and greater transparency as to how determinations of newness will be made for a technology that already has an ICD-9-CM code but is later approved by the FDA for a new indication.

Response: We did not propose any changes to the marginal cost factor in the proposed rule. Furthermore, we continue to believe that a 50-percent marginal cost factor is appropriate for the reasons described in detail in the new technology final rule (66 FR 46919, September 7, 2001).

Also, we have already discussed the situation in which a technology is described under an existing ICD-9-CM code, but subsequently receives approval for a new indication from the FDA. That discussion can be found in the September 7, 2001 new technology final rule (66 FR 46915) and in the FY 2005 IPPS final rule (69 FR 49011) concerning InFUSE® Bone Graft for tibia fractures.

Comment: Several commenters stated that CMS did not address how the proposed changes to the DRGs would affect new technology add-on payments. Another commenter stated that it is essential that CMS maintain new technology add-on payments for FY 2007 and beyond. Another commenter recommended that CMS broaden the new technology criteria to ensure that new technologies are accounted for within a cost-based DRG system.

Response: Although we are adopting a system of cost relative weights in this final rule (section III.C. of this preamble), we will continue to apply the cost criterion using standardized charges consistent with the statute. The statute requires that we apply “a threshold specified by the Secretary that is the lesser of 75 percent of the standardized payment amount (increased to reflect the difference between costs and charges) or 75 percent of one standard deviation for the diagnosis-related group involved.” Changes to the DRG system to better recognize severity in the DRG will also have no effect on our application of the new technology criteria. Any changes to the DRG system will merely result in us calculating different thresholds for the revised DRGs. In addition, once a technology is approved for new technology add-on payments, we will continue to use the ICD-9-CM code to identify the technology for determining when new technology add-on payments are appropriate.

Finally, section 1886(d)(5)(K) and (L) of the Act establishes a process of identifying and ensuring adequate payment for new medical services and technologies. Because no changes have been made to this section of the statute, we will continue to make new technology add-on payments for FY 2007 and beyond for those technologies that meet the criteria.

Comment: One commenter recommended that, because CMS proposed to implement a cost-based weight DRG system, CMS should reconsider whether applicants for FY 2007 new technology add-on payments meet the cost criterion based on a revised data set.

Response: As stated above, Table 10 of the Addendum to the FY 2006 IPPS final rule (70 FR 47680) contained the final thresholds that are being used to evaluate applications for new technology add-on payments for FY 2007. We use the thresholds contained in Table 10 that were published in the previous year's final rule (that is, FY 2006) to determine whether a technology is inadequately paid for the next fiscal year (that is, FY 2007). We publish Table 10 in the proposed rule in order to give the public notice and the opportunity to submit comments before we finalize the thresholds in the final rule. Also, it is necessary for applicants to have the thresholds from Table 10 during the application process so that both the applicants and CMS can establish if the applicant's technology meets the cost criterion. Further, as we note above, we believe that the statute requires us to establish the cost thresholds using charges.

Comment: One commenter noted that section 503 of Pub. L. 108-173 included a provision to expand the inpatient new technology add-on payment to include a broader range of technologies. The commenter added that this legislation was made to ensure that adequate payments were made to hospitals until hospital charges include the costs for these technologies. The commenter explained that CMS’ narrow interpretation has created a situation where few, if any, products can qualify for new technology add-on payments and a process that is opaque and thus, costly, especially for small companies, to apply for add-on payments. The commenter requested that CMS provide greater opportunity for technologies to qualify for add-on payments to ensure patient access to new technologies as Congress intended.

Response: Section 503 of Pub. L. 108-173 amended the law to: (1) require that we establish diagnosis and procedure codes annually on April 1 as well as October 1; (2) change the application of the cost threshold; (3) require a process for obtaining public input on new technology applications prior to the proposed rule; and (4) eliminate the budget neutrality requirement for new technology add-on payments. We believe that we have implemented section 503 as Congress intended.

As we discussed in the FY 2006 IPPS final rule (70 FR 47344), we do not believe that our criteria present an inordinately cumbersome burden for small companies that want to apply for new technology add-on payments. We have received applications for FY 2007 from relatively small companies compared to some of the companies that have applied in the past. Further, we have already been approached by other small companies seeking new technology add-on payments for FY 2008. We encourage potential applicants to contact us before their technology is available on the market if they have questions about the new technology application process.

Comment: One commenter requested that it be given the opportunity to work closely with CMS to help refine the regulatory framework under which CMS evaluates new innovative treatments for Medicare beneficiaries. The commenter suggested ideas such as creating a pathway for small companies under FDA review to elect to meet with CMS to discuss coverage, payment, and coding issues. In addition, the commenter recommended that CMS establish a committee and annual public workshop to assist emerging technologies and small companies with the new technology add-on payment process.

Response: We have been committed to providing ample opportunity for applicants and other interested parties to make their views known to us throughout the application process, at the annual public meeting, and during the comment period on the proposed rule. We encourage interested parties to contact CMS staff for more information about the new technology add-on application process. Interested parties may contact Tiffany Swygert at (410) 786-4642 or Michael Treitel at (410) 786-4552.

Comment: One commenter requested that CMS broaden the definition of substantial clinical improvement. The commenter explained that, in the outpatient setting, CMS views as a separate factor “improvements in the medical technology itself that are so significant that we may wish to recognize them for separate payment even though they do not directly result in substantial clinical improvements.” For example, technological advancements may result in improvement of a product's “convenience, durability [or] ease of operation such as the strength of materials, increased battery life, [and] miniaturization.” The commenter suggested that CMS could recognize these additional improvements along with others when evaluating substantial clinical improvement in the inpatient setting.

Response: The commenter's specific reference to language that was included in the November 2, 2001 OPPS final rule was taken out of context. The language quoted above by the commenter from that OPPS final rule stated that CMS “may,” under the OPPS, recognize technologies for separate payment even though they do not directly result in substantial clinical improvements. To date, under the OPPS, we have only applied the explicit substantial clinical improvement criteria to pass-through device category applications. In the OPPS context, CMS has not found any applications for technologies “that are so significant that we may wish to recognize them for separate payment (as opposed to packaged payments) even though they do not result in substantial clinical improvements” (67 FR 66783). In fact, the historical OPPS experience has indicated that, in general, highly significant advances in medical technology from characteristics such as longer battery life commonly result in substantial clinical improvements that may be appropriately evaluated according to the substantial clinical improvement criteria alone. We have not made a determination to apply these standards within the IPPS. However, as noted in the FY 2005 IPPS final rule (69 FR 49021), we will continue to consider whether to employ specific factors such as those identified for the OPPS in the IPPS.

Comment: One commenter urged CMS not to use the FDA section 510(k) approval process as a bar to a determination of meeting the newness criterion because the “predicate” devices identified through the section 510(k) approval process are not necessarily substantially similar to the new technology; rather the approval indicates that the new device is at least as safe and effective as its predicate(s).

Response: We appreciate the commenter's concern and agree that the mere existence of a predicate device(s) identified in the FDA section 510(k) approval process should not automatically preclude a product from meeting the newness criterion. Although we may consider the predicate devices that are listed in the FDA section 510(k) approval, we will evaluate whether a new technology is substantially similar to existing products on a case-by-case basis. We refer readers to the discussion in the FY 2006 final rule (70 FR 47350-47352) for more detailed information on substantial similarity.

3. FY 2007 Status of Technologies Approved for FY 2006 Add-On Payments

a. Kinetra® Implantable Neurostimulator (Kinetra®) for Deep Brain Stimulation

Medtronic, Inc. submitted an application for approval of the Kinetra® implantable neurostimulator device for new technology add-on payments for FY 2005. In the IPPS final rule for FY 2005 (69 FR 49019, August 11, 2004), we approved Kinetra® for new technology add-on payments.

As noted above, the period for which technologies are eligible to receive new technology add-on payments is 2 to 3 years after the product becomes available on the market and data reflecting the cost of the technology are reflected in the DRG weights. This technology received FDA approval on December 16, 2003. Therefore, the technology will be beyond the 2- to 3-year period during which it can be considered new during FY 2007. Therefore, we proposed in the FY 2007 IPPS proposed rule (71 FR 24070), to discontinue add-on payments for the Kinetra® rechargeable, implantable neurostimulator device for FY 2007.

The manufacturer submitted a request that we consider a higher-paying DRG assignment for dual array neurostimulator pulse generator cases. We have taken this request into consideration and have reviewed the FY 2005 Medicare charge data for cases that use implantable neurostimulator for deep brain stimulation. Our findings and a full discussion of this issue can be found in section II.D.2.a. of the preamble of this final rule.

Comment: A number of commenters were concerned that the expiration of the new technology add-on payment for Kinetra® will lead to inadequate payments for full system Kinetra® implants. One commenter requested that CMS reconsider its decision to end payments for the Kinetra® implantable neurostimulator. Other commenters thanked CMS for its efforts in granting add-on payments for the Kinetra® during the last 2 years.

Response: As noted above, the Kinetra® technology will be beyond the 2-year to 3-year period during which it can be considered new during FY 2007. Therefore, we are finalizing our proposal from the FY 2007 IPPS proposed rule (71 FR 24070) to discontinue add-on payments for the Kinetra® rechargeable implantable neurostimulator for FY 2007.

b. Endovascular Graft Repair of the Thoracic Aorta

W. L. Gore & Associates, Inc. submitted an application for consideration of its Endovascular Graft Repair of the Thoracic Aorta (GORE TAG) for new technology add-on payments for FY 2006. The manufacturer argued that endovascular stent-grafting of the descending thoracic aorta provides a less invasive alternative to the traditional open surgical approach required for the management of descending thoracic aortic aneurysms. The GORE TAG device is a tubular stent-graft mounted on a catheter-based delivery system, and it replaces the synthetic graft normally sutured in place during open surgery. The device was initially identified using ICD-9-CM procedure code 39.79 (Other endovascular repair (of aneurysm) of other vessels). The applicant also requested a unique ICD-9-CM procedure code. As noted in Table 6B of the FY 2006 IPPS final rule (70 FR 47637), new procedure code 39.73 (Endovascular implantation of graft in thoracic aorta) was assigned to this technology.

In the FY 2006 IPPS final rule (70 FR 47356), we approved the GORE TAG device for new technology add-on payment for FY 2006. We noted that any substantially similar device that is FDA-approved before or during FY 2006 that uses the same ICD-9-CM procedure code as GORE TAG and is assigned to the same DRGs as those approved for new technology add-on payments may also receive the new technology add-on payment associated with this technology in FY 2006.

FDA approved GORE TAG on March 23, 2005. The technology remains within the 2- to 3-year period during which it can be considered new. Therefore, as we proposed (71 FR 24070), we are continuing add-on payments for the endovascular graft repair of the thoracic aorta for FY 2007.

Comment: Some commenters supported our proposal to continue new technology add-on payments for GORE TAG for FY 2007.

Response: We thank the commenters for their support and, as noted above, we are continuing new technology add-on payments for GORE TAG for FY 2007.

c. Restore® Rechargeable Implantable Neurostimulator

Medtronic Neurological submitted an application for new technology add-on payments for its Restore® Rechargeable Implantable Neurostimulator for FY 2006. The Restore® Rechargeable Implantable Neurostimulator is designed to deliver electrical stimulation to the spinal cord to block the sensation of pain. The technology standard for neurostimulators uses internal sealed batteries as the power source to generate the electrical current. These internal batteries have finite lives, and require replacement when their power has been completely discharged. According to the manufacturer, the Restore® Rechargeable Implantable Neurostimulator “represents the next generation of neurostimulator technology, allowing the physician to set the voltage parameters in such a way that fully meets the patient's requirements to achieve adequate pain relief without fear of premature depletion of the battery.” The applicant stated that the expected life of the Restore® rechargeable battery is 9 years, compared to an average life of 3 years for conventional neurostimulator batteries. We approved new technology add-on payments for all rechargeable, implantable neurostimulators for FY 2006. Cases involving these devices, made by any manufacturer, are identified by the presence of newly created ICD-9-CM code 86.98 (Insertion or replacement of dual array rechargeable neurostimulator pulse generator).

As noted above, the period for which technologies are eligible to receive new technology add-on payments is 2 to 3 years after the product becomes available on the market and data reflecting the cost of the technology are reflected in the DRG weights. The FDA approved the Restore® Rechargeable Implantable Neurostimulator in 2005. However, as noted above and in the FY 2006 IPPS final rule (70 FR 47358), at least one similar product was approved by the FDA as early as April 2004. Nevertheless, consistent with current policy (70 FR 47362) and decisions for prior products (that is, bone morphogenetic products and CRT-D devices), as we proposed (71 FR 24070 through 24071), we are continuing new technology add-on payments for rechargeable, implantable neurostimulators in FY 2007 because the product will be beyond the 3-year period only in the latter 6 months of the fiscal year.

Comment: Some commenters supported our decision to continue add-on payments for the Restore® Rechargeable Implantable Neurostimulator.

Response: We appreciate the commenters' support and as noted above, we are continuing new technology add-on payments for Restore® Rechargeable Implantable Neurostimulator for FY 2007.

4. FY 2007 Applications for New Technology Add-On Payments

a. C-Port® Distal Anastomosis System

Cardica, Inc. submitted an application for new technology add-on payments for FY 2007 for its Cardica C-Port® Distal Anastomosis System. The manufacturer stated that the C-Port® System is indicated for all patients requiring a vein as a conduit during a coronary bypass operation for bypassing a coronary artery stenosis or occlusion. The manufacturer contended that the C-Port® System is specifically designed to create a reliable and consistent end-to-side anastomosis between a conduit, such as a venous graft, and a small arterial vessel during the bypass surgery. The device consists of eight stainless steel clips and a delivery system. Once the vein graft has been loaded into the device and the device positioned against the target vessel, the anastomosis is created by pushing a single button. Cardica, Inc. stated the main purpose of the device is to replace a conventional hand-sewn, distal anastomosis with an automated, compliant, mechanical anastomosis.

We received the following public comments at the new technology town hall meeting regarding whether this technology meets the substantial clinical improvement criteria:

Comment: The manufacturer argued that this technology meets the substantial clinical improvement criterion because:

• It achieves higher patency rates at 6 months compared to conventional hand-sewn anastomoses.
• Use of the device will result in less surgeon-to-surgeon variability in the quality of the anastomosis compared to hand sewing.
• The device leads to reduced operative time.
• The product allows for the creation of an anastomosis during minimally invasive surgery.

In addition, we received written comments expressing support for approval of new technology add-on payments for the C-Port® System. These commenters noted that—

• The device allows the anastomosis to be completed quickly, reducing patient complications during surgery from ischemia.
• The device will allow for smaller incisions during heart surgery and physicians will not have to position their hands in the chest cavity in order to hand-sew the anastomosis.
• The rapidly deployed anastomosis clamp provides patients with a surgical alternative where one would otherwise not be available due to the comorbidities associated with the more invasive CABG procedures.

Response: We appreciate the time and effort the applicant took to present at the town hall meeting. We indicated in the proposed rule that we would consider the information presented in the written comments and at the town hall meeting, and invoted interested parties to submit objective data that would support the assertions presented above by the commenters.

The C-Port® System was granted section 510(k) approval from the FDA on November 10, 2005. While the device appeared to meet the criteria for being considered new based on its FDA approval date, we were concerned that various forms of surgical staples and clips have been used for more than a decade in a wide range of surgical procedures. In fact, the FDA found that the C-Port® System “is substantially equivalent to the predicate devices with regard to indications, device characteristics, method of use, labeling and materials.” Thus, given its similarity to other devices currently on the market, we were concerned that the C-Port® System may not qualify as new. In the FY 2007 IPPS proposed rule, we solicited specific comments on whether this device is new and how it could be distinguished from predicate devices that perform the same or a similar function.

We received the following public comments in response to the proposed rule.

Comment: The manufacturer commented that the C-Port® System meets the newness criterion for the following reasons:

• The FDA section 510(k) approval process identifies predicate devices as having “a similar, not necessarily identical use and function.”
• There is no other “fully-integrated anastomotic system cleared by the FDA for the creation of an anastomosis between a blood vessel graft and a target coronary artery.” There are no “clip or staple-based automated distal coronary anastomotic devices such as [C-Port®] approved by the FDA.” The manufacturer argued that while the devices they identified in the FDA section 510(k) approval process are similar to C-Port® system, none of them are identical.
• C-Port® was FDA approved in November 2005, thus enabling the device to still qualify as new based on its FDA approval date.
• There is no clinical precedence for the use of a stapling device in creating distal coronary anastomoses, and there are no ICD-9 CM codes for stapling devices—the lack of the procedure code means that CMS does not have charge data for C-Port® and that the device's costs are not reflected in the current DRG weights.
• CMS approved Kinetra® in 2004 and stated that the Kinetra® device was not “significantly different in terms of how it achieves its desired clinical results from its predecessor Soletra®.” The manufacturer believed that the approval of Kinetra® sets precedence for C-Port® approval.

Response: We appreciate the manufacturer's clarification of the questions we posed in the proposed rule about whether the C-Port® would meet the newness criterion. The additional information submitted has allowed us to determine that the C-Port® meets the newness criterion.

In response to the commenter's statement about Kinetra®, we indicated that Soletra® and Kinetra® achieve the desired clinical result through the same stimulation mechanism. However, we did not find Soletra® and Kinetra® to be substantially similar products. We noted that Soletra® controls symptoms only on one side of a patient's body, while Kinetra® provides bilateral control of neurological symptoms through a single device. We determined in the FY 2005 IPPS final rule (69 FR 49019) that Kinetra® represented a substantial clinical improvement over the previous Soletra® device.

In the proposed rule, we also noted that there is currently no ICD-9-CM code used to identify how the anastomosis is performed. The surgical technique used to graft the bypass to the arterial vessel is part of the surgical procedure itself and is not separately identified in our current coding structure. Although there is not an explicit code to identify C-Port®, the hospital's charge for the device will be included on its bill. The hospital is permitted to charge for all items and services it furnishes irrespective of whether a particular item is identified by an explicit ICD-9-CM code. The charges included on hospital bills for the device will be part of the relative weight calculation 2 years later (that is, FY 2005 hospital charge data are used to set the FY 2007 relative weights).

Comment: The manufacturer of C-Port® urged CMS to differentiate between “distinct procedures involving the creation of anastomosis” by creating the following codes: (a) Anastomosis, manual; and (b) anastomosis, automated, using single or multiple clip array deployment technology. The manufacturer commented that a new code should be created for C-Port® because the C-Port® Distal Anastomosis procedure is not a typical part of the bypass procedure code and the use of the C-Port® system requires training and proctoring for physicians and OR staff to use the equipment because the C-Port® system comprises new steps and preparation in the bypass procedure. Finally, the manufacturer stated that CMS set a precedent for the creation of a new code by creating a code for a drug-eluting stent even though ICD-9-CM procedure codes already existed for stent procedures and by creating a new code to distinguish single versus dual channel-pulse generator devices (Kinetra® by Medtronic).

Response: While the use of the C-Port® device may represent a difference in technique of creating a distal anastomosis, we do not agree that it is a distinct procedure. Historically, we have subdivided procedures involving the insertion of specific devices that are designed to achieve a specific therapeutic purpose, but we have not assigned a code for specific tools used to perform surgery. Kinetra®, a stent and a pacemaker, is an example of a device that is implanted in a patient to treat an illness that is appropriately assigned a code. To date, we have not used a code to identify a specific type of surgical tool such as a scalpel, saw, or clamp. Similarly, we view C-Port® as a surgical tool (albeit far more sophisticated or innovative than those just mentioned) that should also not be recognized by its own ICD-9-CM code.

The applicant made several arguments in support of the device meeting the cost criterion. Cardica, Inc. estimated that the cost of each device will be approximately $1,200. The applicant assumed a hospital markup of 100 percent, with an average use of 2.5 C-Port® devices per case. Therefore, it estimated that the total average charge per patient will be $6,000. The C-Port® System would be used when a coronary artery bypass graft is performed. Thus, we assessed whether it meets the cost criterion in relation to the threshold for DRGs 106 (Coronary Bypass with Percutaneous Transluminal Coronary Angioplasty), 547 (Coronary Bypass with Cardiac Catheter with Major CV Diagnosis), 548 (Coronary Bypass with Cardiac Catheter without Major CV Diagnosis), 549 (Coronary Bypass without Cardiac Catheter with Major CV Diagnosis), and 550 (Coronary Bypass without Cardiac Catheter without Major CV Diagnosis). We note that the data analysis for this technology is slightly unusual, as the DRGs to which the technology would have been assigned in FY 2005 (the MedPAR data we are currently using) are DRGs 107 and 109. These DRGs were terminated in FY 2006, and 4 new coronary bypass DRGs were created for these cases (DRGs 547, 548, 549, and 550). The manufacturer provided estimates showing a case-weighted threshold for DRGs 106, 547, 548, 549 and 550 of $75,373. The applicant projected a 20-percent market penetration for the device in FY 2007 or its use in approximately 23,000 cases across the 5 DRGs. The applicant submitted data showing average standardized charges for cases using the C-Port® System of $80,887. Therefore, the applicant argued that the device meets the cost threshold for a new technology add-on payment. Our internal data analysis of the technology, using the FY 2005 MedPAR data and Table 10 thresholds for FY 2005, shows a case-weighted threshold of $68,416. We identified cases using coronary bypass procedure codes 36.10, 36.11, 36.12, 36.13 and 36.14, and concluded that the case-weighted average standardized charge for these bypass cases was $79,394. Thus, our internal data also suggested that the device meets the cost threshold.

As we discussed in the proposed rule, the applicant made several arguments in support of the device meeting the substantial clinical improvement criterion. The manufacturer argued that the C-Port® creates a reliable and fully compliant end-to-side anastomosis between a vein graft and a coronary artery, in less time than is required to create a hand-sewn distal anastomosis. The applicant also stated that the C-Port® System integrates deployment of the anastomotic clips and creation of the arteriotomy, thus enabling deployment to occur without occlusion of blood flow through the target vessel. However, we note that the applicant submitted evidence suggesting that the device does not always produce reliable anastomoses; specifically, a study of 130 patients receiving 132 devices reported 13 incomplete anastomoses in 12 patients, and the study also noted that additional manual stitches were required in the majority of the patients studied. Therefore, we were concerned that these studies suggested that the C-Port® System may not represent a substantial clinical improvement over the traditional hand-sewn technique. At the town hall meeting, the applicant noted that these results were associated with inexperience preparing the target vessel, vein thickness assessment, proper device alignment and anastomosis site selection rather than problems with the device itself. The applicant believed that these problems will become infrequent as surgeons have more experience with the device. In the FY 2007 IPPS proposed rule, we solicited further information from commenters that would suggest how the product meets the substantial clinical improvement criterion.

We received the following comment in response to the proposed rule.

Comment: The manufacturer submitted the following comments to be considered in our evaluation of whether C-Port® met the substantial clinical improvement criterion:

• Intraoperative anastomotic failures with the hand-sewn technique occur in approximately 10 percent of patients. Falk, et al., evaluated vein graft patency using a meta-analysis of 28 published studies with over 28,000 grafts and found that occlusion within 30 days occurs in about 12 percent of vein grafts while occlusion within 6 months occurs in 20 percent.
• The C-Port® device may mitigate some of the negative factors found in hand-sewn anastomoses that impact vein graft patency. Post-operative vein graft patency rates using the hand-sewn technique were 88 percent at 30 days and 80 percent at 6 months (data obtained from historical controls); whereas patency rates using the C-Port® device were 99 percent at discharge and 96 percent at 6 months.
• In the greater than 1-year followup group, none of the patients in the pivotal C-Port® study required a reintervention.
• The “10 percent failure rate” cited in a C-Port® publication referred to a failure in surgeons using the device (due to lack of experience using it), not a failure of the device itself.

Response: We are concerned that information presented by the applicant does not demonstrate that this technology is a sufficient improvement over hand-sewing the distal anastomosis. Although patency rates using the C-Port® device were reportedly higher than those found using the hand-sewn technique (99 percent at discharge and 96 percent at 6 months compared to 88 percent at 30 days and 80 percent at 6 months), we also found that the data on the hand-sewn patency rates was derived from a meta-analysis of over 28,000 bypass grafts to different coronary vessels, many of which may have been comparatively poor candidates for bypass grafting, suggesting a possible selection bias in the arteries in the C-Port® study. We believe that a clinical study demonstrating substantial clinical improvement in outcomes is necessary for this technology because the comparison is of the CABG procedure using the C-Port® device to the hand-sewn technique. In some cases, our approval of a technology was based on a clinical assessment that at least one of the criteria for evaluating substantial clinical improvement listed in the new technology final rule (66 FR 46914) was met. For example, our approval of the Restore rechargeable neurostimulator was based on evidence that showed it decreased the “rate of subsequent * * * therapeutic interventions” by avoiding a surgery to replace a battery. Similarly, we approved GORE TAG because it “offers a treatment option for patient population unresponsive to, or ineligible for, currently available treatments.” In these cases, we were less reliant on a clinical study to demonstrate improvement over an existing technology than our clinical judgment that the product achieved its intended purposes which itself is a substantial clinical improvement. With C-Port® or with a hand-sewn anastomosis, the treatment is the same (a CABG for coronary artery vessel disease). Thus, clinical studies demonstrating an improvement in CABG outcomes using the C-Port® device relative to the hand-sewn technique are critical to approving the device for new technology add-on payments.

Given the relatively high rates of success of both the hand sewn and the automated technique, we were not able to determine that the C-Port® device is a substantial clinical improvement over the traditional hand-sewn technique. Accordingly, after consideration of the comments received, we are not approving the C-Port® Distal Anastomosis System for FY 2007 new technology add-on payment.

There are several potential criteria listed in the new technology final rule that C-Port® could potentially meet. For instance, it is possible that C-Port® will reduce recovery time or lead to more rapid beneficial resolution of the disease process treatment. Given the potential benefits of C-Port®, it is likely that we would approve the technology for add-on payments with a study that more definitively demonstrates substantial clinical improvement. For instance, our main concern with the study presented was that the control group and the study population used to demonstrate substantial clinical improvement may not have been directly comparable. If there was a study that showed similar improvements in patency rates between the control group and a study population where the patients were directly comparable in their coronary artery vessel disease, we believe it would be more likely to demonstrate that the substantial clinical improvement criterion was met.

b. NovoSeven® for Intracerebral Hemorrhage

The Pinnacle Health Group in conjunction with Novo Nordisk Inc. (the manufacturer) submitted an application for new technology add-on payments for FY 2007 for NovoSeven® for Intracerebral Hemorrhage. However, the applicant withdrew its application for new technology add-on payment on June 07, 2006.

We received the following public comments regarding this application for new technology add-on payments in response to the FY 2007 IPPS proposed rule.

Comment: One commenter supported approving new technology add-on payments for NovoSeven®. The commenter believed that the availability of an add-on payment would help facilitate patient access to this important and costly therapy.

Response: We appreciate the commenter's response to the proposed rule. We note that, during the comment period, the applicant withdrew its application from consideration for new technology add-on payments for FY 2007.

We appreciate the applicant for its submittal of an application for new technology add-on payments and encourage a resubmission of an application upon FDA approval of its technology.

c. X STOP Interspinous Process Decompression System

St. Francis Medical Technologies submitted an application for new technology add-on payments for the X STOP Interspinous Process Decompression System for FY 2007. Lumbar spinal stenosis describes a condition that occurs when the spaces between bones in the spine become narrowed due to arthritis and other age-related conditions. This narrowing, or stenosis, causes nerves coming from the spinal cord to be compressed, thereby causing symptoms including pain, numbness, and weakness. It particularly causes symptoms when the spine is in extension, as occurs when a patient stands fully upright or leans back. The X STOP device is inserted between the spinous processes of adjacent vertebrae in order to provide a minimally invasive alternative to conservative treatment (exercise and physical therapy) and invasive surgery (spinal fusion). It works by limiting the spine extension that compresses the nerve roots while still preserving as much motion as possible. The device is inserted in a relatively simple, primarily outpatient procedure using local anesthesia. However, in some circumstances, the physician may prefer to admit the patient for an inpatient stay. The manufacturer described the device as providing “a new minimally invasive, stand-alone alternative treatment for lumbar spinal stenosis.”

The X STOP Interspinous Process Decompression system received pre-market approval from the FDA on November 21, 2005. The device is currently described by ICD-9-CM code 84.58 (Implantation of Interspinous process decompression device) (excluding: fusion of spine (codes 81.00 through 81.08, and 81.30 through 81.39)). This ICD-9-CM code went into effect on October 1, 2005.

The manufacturer provided data in support of the device meeting the cost threshold criterion. The applicant stated that there would be an average of 1.6 units used per case. Each unit costs $5,500; therefore, the technology is expected to cost $8,800 per case. The device is currently assigned to DRGs 499 (Back and Neck Procedures Except Spinal Fusion with CC) and 500 (Back and Neck Procedures Except Spinal Fusion without CC). The manufacturer projected that there would be approximately 424 patients eligible to receive the device in DRG 499 in FY 2007, while there may be approximately 1,700 patients who receive the device in DRG 500. The manufacturer also provided data for cases involved in the clinical trials. The average standardized charge for the cases in FY 2004 was $24,065. The weighted threshold for DRGs 499 and 500 is $20,096. However, the manufacturer argued that because significantly less than 20 percent of patients receiving the X STOP experienced complications or had comorbidities, the threshold should be calculated by estimating that 20 percent of patients would be assigned to DRG 499 and 80 percent would to DRG 500. The manufacturer stated in its application that, using this methodology, the applicable threshold should be $19,796. Using either calculation, it appears that the technology meets the cost threshold for new technology add-on payments.

The applicant also submitted information in support of its claim of meeting the substantial clinical improvement criterion. The manufacturer stated that the X STOP device is placed between the spinous processes to limit extension of the symptomatic level(s), yet allowing flexion, axial rotation, and lateral bending (that is, the device limits pressure on the spinal nerves and the resulting pain symptoms when the patient is in an upright position or leans backward while also preserving the patient's ability to turn side-to-side, bend forward, and to turn to either side). The applicant contended that this technology provides an alternative with improved clinical outcomes to conservative and surgical treatments. The manufacturer further stated that the device may offer a new alternative to lumbar spinal decompression procedures such as laminectomy and laminotomy. Additional information included in the application suggested that the device preserves spinal motion and is superior to a spinal decompression procedure that requires concomitant fusion (with or without instrumentation). The applicant argued that the advantages over spinal decompression include reduced risk, shorter hospital stay, and earlier improvement in pain and function. The manufacturer further contended that disease progression at adjacent levels is minimal following X STOP implantation compared to the known risk associated with surgical decompression and concomitant fusion. The applicant stated that the X STOP is comparable to traditional surgical decompression of lumbar spinal stenosis with respect to improved quality of life postoperatively. According to the applicant, the device provides advantages over nonoperative care, including better symptom relief, improved function, and increased patient satisfaction.

We received the following public comments through the new technology town hall meeting process regarding this application for add-on payments.

Comment: The applicant asserted that the X STOP Interspinous Process Decompression system has the following advantages:

• It retains spinal anatomy and all spinal structures.
• The device allows for increased function and less pain after implantation as evidenced by radiographic measures that showed increases in the spinal canal area by 18 percent, diameter by 9 percent, and subarticular diameter (the route that the nerves exit the spine) by 50 percent. In lateral view: area increased by 25 percent and width by 41 percent.
• The X STOP is a reversible procedure that causes no damage to facets or disks.
• The device allows for a treatment option for patients that cannot undergo surgeries with general anesthesia.
• The rate of complications associated with implantation of the device is below 1 percent.

Response: In the proposed rule, we indicated that we would evaluate these assertions as we further considered this application for new technology add-on payments for the final rule. We also noted that the study that the applicant summarized at the town hall meeting for the X STOP used a randomized study that targeted lumbar spinal stenosis patients with mild to moderate symptoms. The control group did not require operative care. In the proposed rule, we solicited information from the comments that demonstrates how the study populations showed substantial clinical improvement compared to the control group.

We believe that the device satisfies the newness and cost threshold criteria for new technology add-on payments. However, in the FY 2007 IPPS proposed rule, we expressed our concern that the information included with the application may raise issues about substantial clinical improvement. During the FDA approval process, the Center for Devices and Radiological Health (CDRH) Advisory Panel voted against premarket approval (PMA) in August 2004 because of concerns about proper patient selection as well as the lack of objective endpoints, especially radiographic endpoints. The Panel also mentioned the overall low clinical efficacy rate in the study population. The device subsequently received PMA approval, but only on the condition that it be used in the context of a long term (5 year) follow-up study. In the proposed rule, we solicited information from commenters that addressed the concerns raised by the CDRH Advisory Panel or other information bearing on the issue of whether this product meets the substantial clinical improvement criterion.

We note that the town hall meeting produced contradictory information regarding whether this procedure is generally performed in inpatient or outpatient settings. The presenter indicated that over 90 percent of his patients were treated as outpatients. The manufacturer noted that 90 percent of non-U.S. patients and approximately two-thirds of U.S. patients since FDA approval have been treated in inpatient settings. While the setting where the procedure is typically performed has no bearing on whether the product represents a substantial clinical improvement, we noted that we believe the physician should select the most appropriate site to perform the procedure based on the clinical needs of the patient.

We received the following comments in response to the FY 2007 IPPS proposed rule.

Comment: The manufacturer commented that the contradictory information we noted in the proposed rule about whether the procedure in general performed in the outpatient or inpatient setting was likely the result of the presenter at the town hall meeting misspeaking when he said that the device was used in the outpatient setting about 90 percent of the time. Although the device may be used with local anesthesia, the manufacturer predicted that many clinicians attending to Medicare patients will choose general anesthesia and will use the procedure in an inpatient setting. The manufacturer stated that the X STOP device is currently used in the inpatient setting about 90 percent of the time.

Response: We appreciate the commenter's clarification of this point. As we indicated in the proposed rule, the site of service has no bearing on whether we will determine the technology to be a substantial clinical improvement. However, given the similarity in the criteria we apply in the two settings for determining substantial clinical improvement, we note that a decision to approve a device for inpatient new technology add-on payment may have implications for outpatient new technology pass-through payment.

Comment: In response to our request for additional information supporting that the X STOP device meets the substantial clinical improvement criterion, the manufacturer reiterated many of the comments that it submitted through the new technology town hall meeting process. Mainly, the commenter stated that X STOP offers an alternative to surgery that is associated with fewer and less severe complications, is a reversible procedure, and offers a faster recovery time than more invasive surgery. The commenter also stated that X STOP meets the criterion when compared to other disease management modalities for lumbar spinal stenosis patients, as evidenced by symptom relief, physical functioning, treatment satisfaction, and health-related quality of life, and that use of X STOP results in—

• Comparable treatment efficacy when compared to laminectomy
• Lower rates of intraoperative complications compared to surgical decompression with or without concomitant fusion
• Lower reoperation rates for unresolved stenosis systems compared to other surgical treatments.

In addition, the manufacturer stated that it addressed the issues that the Advisory Panel to the FDA cited as reasons for voting against approving X STOP. Those issues were in regards to proper patient selection, a lack of objective endpoints, especially radiographic endpoints and an overall low clinical efficacy rate in the study population. The manufacturer claimed that it addressed the concerns of the Advisory Panel by submitting additional data and analyses to the FDA that—

• Identified patients with LSS and moderately impaired physical function at baseline as the appropriate indication.
• Supplemented “the showing of the mechanism of effect on the spine in cadavers with in vivo clinical radiographic data.”
• Addressed the issue of low clinical efficacy rates, by showing that the success rates using X STOP were comparable to those of more invasive procedures that are covered by Medicare.

The manufacturer further noted that the Advisory Panel wrote in its Summary of Safety and Effectiveness document that “the X STOP device met the primary clinical study endpoint for success, exceeding the success rate of the control in every statistical analysis.” Finally, the manufacturer noted that the FDA requirement that X STOP's approval was conditioned on a 5-year followup study was not uncommon for spinal implant devices and that, over the past 10 years, all nine spinal implant FDA approvals have had similar conditional requirements. The manufacturer also commented that CMS approved the INFUSE Bone Graft device and noted that the FDA required a 6-year followup study as a condition of its approval of that device.

Several commenters who were individual physicians who have had experience using the X STOP device indicated that X STOP provides an alternative to more invasive surgery such as a laminectomy after conservative treatment has failed. All of the commenters supported approving the device for new technology add-on payment. In addition to commenters' support that the device is minimally invasive and has short operative and recovery time, some of the commenters mentioned other positive outcomes that the X STOP procedure—

• Increases foraminal height and produces minimal reversal of the lordosis, as measured by post operative x-rays;
• Reduced the pain reported by patients by half in some cases;
• Provided alleviation of neurogenic claudication symptoms; and
• Benefited patients with significant comorbidities, including cardiothoracic problems, specifically chronic obstructive pulmonary disease or coronary artery disease

In addition, some commenters noted that the X STOP device can very easily be implanted in the outpatient setting (assuming appropriate patient-selection), thus allowing high inpatient costs to be avoided.

Response: We appreciate the commenters' submittal of comments in support of X STOP. With respect to substantial clinical improvement, we continue to be concerned that the FDA Advisory Panel noted the overall low clinical efficacy rate in the study population and only approved the technology conditional on a 5 year followup study. Nevertheless, we note that the FDA did approve the technology, meaning that it is safe and effective (that is, it achieves its intended purpose). Further, we note that the applicant was able to address the FDA concern about lack of objective endpoints by the showing of the mechanism of effect on the spine in cadavers with in vivo clinical radiographic data. That is, the applicant was able to show that the X STOP device limits spine extension that compresses the nerve. Thus, we believe that the technology has promise for providing a less invasive alternative to procedures such as laminectomy or fusion for patients that have failed conservative treatment (exercise, physical therapy and medication). The X STOP system represents a new level of treatment on the continuum of care for patients with lumbar spinal stenosis that previously did not exist.

Accordingly, after consideration of the comments received, we are approving the X STOP Interspinous Process Decompression System for new technology add-on payment for FY 2007. However, we remain interested in seeing whether the clinical evidence from the 5-year followup study required by the FDA demonstrates that X STOP continues to be effective. Cases involving X STOP will be identified by ICD-9-CM code 84.58 (Implantation of interspinous process decompression device). These cases are generally included in DRG 499 (Back and Neck Procedures Except Spinal Fusion with CC) and DRG 500 (Back and Neck Procedures Except Spinal Fusion without CC). As noted in the proposed rule, the manufacturer submitted data to support its estimated cost per case involving the X STOP procedure of $8,800. Accordingly, we are finalizing a maximum add-on payment of $4,400 for cases that involve this technology.

5. Interim and Final Cost Threshold Tables Due to Changes to Wage Index and Budget Neutrality Factors

Table 10 of the IPPS proposed and final rules contains the cost thresholds that are used to determine whether a technology meets the criteria for new technology add-on payments. We are publishing an interim Table 10 in this final rule. We use the national adjusted operating standardized amounts in calculating the cost threshold. As noted in section III. and in the Addendum to this final rule, the final national adjusted operating standardized amounts will be published subsequent to this final rule when the wage index and budget neutrality factors are finalized for FY 2007. Therefore, we will also publish a revised version of Table 10, containing the final thresholds for FY 2008 between August 1 and October 1.

III. Changes to the Hospital Wage Index

A. Background

Section 1886(d)(3)(E) of the Act requires that, as part of the methodology for determining prospective payments to hospitals, the Secretary must adjust the standardized amounts “for area differences in hospital wage levels by a factor (established by the Secretary) reflecting the relative hospital wage level in the geographic area of the hospital compared to the national average hospital wage level.” In accordance with the broad discretion conferred under the Act, we currently define hospital labor market areas based on the definitions of statistical areas established by the Office of Management and Budget (OMB). A discussion of the FY 2007 hospital wage index based on the statistical areas, including OMB's revised definitions of Metropolitan Areas, appears under section III.B. of this preamble.

Beginning October 1, 1993, section 1886(d)(3)(E) of the Act requires that we update the wage index annually. Furthermore, this section provides that the Secretary base the update on a survey of wages and wage-related costs of short-term, acute care hospitals. The survey must exclude the wages and wage-related costs incurred in furnishing skilled nursing services. This provision also requires us to make any updates or adjustments to the wage index in a manner that ensures that aggregate payments to hospitals are not affected by the change in the wage index. The adjustment for FY 2007 is discussed in section II.B. of the Addendum to this final rule.

As discussed below in section III.G. of this preamble, we also take into account the geographic reclassification of hospitals in accordance with sections 1886(d)(8)(B) and 1886(d)(10) of the Act when calculating the wage index. Under section 1886(d)(8)(D) of the Act, the Secretary is required to adjust the standardized amounts so as to ensure that aggregate payments under the IPPS after implementation of the provisions of sections 1886(d)(8)(B) and (C) and 1886(d)(10) of the Act are equal to the aggregate prospective payments that would have been made absent these provisions. The budget neutrality adjustment for FY 2007 is discussed in section II.A.4.b. of the Addendum to this final rule.

Section 1886(d)(3)(E) of the Act also provides for the collection of data every 3 years on the occupational mix of employees for short-term, acute care hospitals participating in the Medicare program, in order to construct an occupational mix adjustment to the wage index. A discussion of the occupational mix adjustment that we are applying beginning October 1, 2006 (the FY 2007 wage index) appears under section III.C. of this preamble.

B. Core-Based Statistical Areas for the Hospital Wage Index

The wage index is calculated and assigned to hospitals on the basis of the labor market area in which the hospital is located. In accordance with the broad discretion under section 1886(d)(3)(E) of the Act, beginning with FY 2005, we define hospital labor market areas based on the Core-Based Statistical Areas (CBSAs) established by OMB and announced in December 2003 (69 FR 49027). OMB defines a CBSA, beginning in 2003, as “a geographic entity associated with at least one core of 10,000 or more population, plus adjacent territory that has a high degree of social and economic integration with the core as measured by commuting ties.” The standards designate and define two categories of CBSAs: Metropolitan Statistical Areas (MSAs) and Micropolitan Statistical Areas (65 FR 82235).

According to OMB, MSAs are based on urbanized areas of 50,000 or more population, and Micropolitan Statistical Areas (referred to in this discussion as Micropolitan Areas) are based on urban clusters with a population of at least 10,000 but less than 50,000. Counties that do not fall within CBSAs are deemed “Outside CBSAs.” In the past, OMB defined MSAs around areas with a minimum core population of 50,000, and smaller areas were “Outside MSAs.”

The general concept of the CBSAs is that of an area containing a recognized population nucleus and adjacent communities that have a high degree of integration with that nucleus. The purpose of the standards is to provide nationally consistent definitions for collecting, tabulating, and publishing Federal statistics for a set of geographic areas. CBSAs include adjacent counties that have a minimum of 25 percent commuting to the central counties of the area. (This is an increase over the minimum commuting threshold of 15 percent for outlying counties applied in the previous MSA definition.) We consider CBSAs that are MSAs to be urban, and CBSAs that are Micropolitan Statistical Areas as well as areas outside of CBSAs to be rural. In addition, where an MSA has been divided into Metropolitan Division to comprise the labor market areas for purposes of calculating the wage index (69 FR 49029).

The revised CBSAs established by OMB comprised MSAs and Micropolitan Areas based on Census 2000 data. (A copy of the announcement may be obtained at the following Internet address: http://www.whitehouse.gov/omb/bulletins/fy04/b04-03.html.) The revised definitions recognize 49 MSAs and 565 Micropolitan Areas, and extensively changed the composition of many of the MSAs that existed prior to the revisions.

The revised area designations resulted in a higher wage index for some areas and a lower wage index for others. Further, some hospitals that were previously classified as urban are now in rural areas. Given the significant payment impacts upon some hospitals because of these changes, we provided a transition period to the new labor market areas in the FY 2005 IPPS final rule (69 FR 49027 through 49034). As part of that transition, we allowed urban hospitals that became rural under the new definitions to maintain their assignment to the MSA where they were previously located for the 3-year period of FY 2005, FY 2006, and FY 2007. Specifically, these hospitals were assigned the wage index of the urban area to which they previously belonged. (For purposes of the wage index computation, the wage data of these hospitals remained assigned to the statewide rural area in which they are located.) The hospitals receiving this transition will not be considered urban hospitals; rather, they will maintain their status as rural hospitals. Thus, the hospital would not be eligible, for example, for a large urban add-on payment under the capital PPS. In other words, it is the wage index, but not the urban or rural status, of these hospitals that is being affected by this transition. The higher wage indices that these hospitals are receiving are also being taken into consideration in determining whether they qualify for the out-migration adjustment discussed in section III.I. of this preamble and the amount of any adjustment.

FY 2007 will be the third year of this transition period. We will continue to assign the wage index for the urban area in which the hospital was previously located through FY 2007. In order to ensure this provision remains budget neutral, we will continue to adjust the standardized amount by a transition budget neutrality factor to account for these hospitals. Doing so is consistent with the requirement of section 1886(d)(3)(E) of the Act that any “adjustments or updates [to the adjustment for different area wage levels] * * * shall be made in a manner that assures that aggregate payments * * * are not greater or less than those that would have been made in the year without such adjustment.”

Beginning in FY 2008, these hospitals will receive their statewide rural wage index, although they will be eligible to apply for reclassification by the MGCRB both during this transition period and in subsequent years. These hospitals will be considered rural for reclassification purposes.

Consistent with the FY 2005 and FY 2006 IPPS final rules, as we did beginning in FY 2006, for FY 2007 we are providing that hospitals receive 100 percent of their wage index based upon the CBSA configurations. Specifically, we will determine for each hospital a wage index for FY 2007 employing wage index data from FY 2003 hospital cost reports and using the CBSA labor market definitions.

C. Occupational Mix Adjustment to the FY 2007 Wage Index

As stated earlier, section 1886(d)(3)(E) of the Act provides for the collection of data every 3 years on the occupational mix of employees for each short-term, acute care hospital participating in the Medicare program, in order to construct an occupational mix adjustment to the wage index, for application beginning October 1, 2004 (the FY 2005 wage index). The purpose of the occupational mix adjustment is to control for the effect of hospitals” employment choices on the wage index. For example, hospitals may choose to employ different combinations of registered nurses, licensed practical nurses, nursing aides, and medical assistants for the purpose of providing nursing care to their patients. The varying labor costs associated with these choices reflect hospital management decisions rather than geographic differences in the costs of labor.

Comment: Some commenters expressed concern about the occupational mix adjustment relative to the proposed implementation of changes to the DRG system. A few stated that the purpose of the occupational mix adjustment is to ensure that hospitals are not paid through both the wage index and the resource-based DRG system for the additional resources needed for certain procedures. The commenters suggested that the occupational mix adjustment is not necessary if a robust severity-adjusted DRG system is implemented. Other commenters indicated that CMS should consider deferring the implementation of the proposed hospital-specific cost weighting methodology and severity DRGs until at least FY 2008 to alleviate the burden on hospitals that will be negatively affected by a redistribution of Medicare payments under the new occupational mix adjustment.

Response: We remind the commenters that an occupational mix adjustment to the wage index is required under section 1886(d)(3)(E) of the Act. Although we understand the commenters’ concerns that some hospitals may be negatively affected by the new occupational mix adjustment, we also believe that it is important for us to move forward with implementing changes in the DRG system that would recognize that some more complex cases may require a higher DRG payment because the services are provided by more highly skilled workers.

Comment: A few commenters opposed the occupational mix adjustment. One commenter believed that the initial application of the occupational mix adjustment had unintended results, benefiting fewer rural hospitals and more large urban hospitals than anticipated. The commenter stated that this problem has been compounded by the additional pressure from the decision in Bellevue Hosp. Center v. Leavitt, 443 F.3d 163 (2nd Cir. 2006), and, therefore, recommended that CMS approach Congress about repealing the mandate for the occupational mix adjustment. Another commenter indicated that the occupational mix survey is confusing and burdensome to hospitals.

Response: As held in Bellevue Hosp. Center v. Leavitt, 443 F.3d 163 (2nd Cir. 2006), adjusting the wage index for occupational mix is required by Congress. Therefore, commenters who believe that the occupational mix should be eliminated would need to approach the Congress with such concerns. As for the initial application of the occupational mix, we believe the unexpected outcomes may have been due to a combination of factors, including the newness of the survey and changing trends in hospital employment. We have modified the survey for 2006, and these modifications should reduce the risk of reporting and measurement errors. These modifications are based largely on suggestions we received from MedPAC and the hospital community. We understand the commenter's concern that completing the survey causes a burden to hospitals; however, the statute requires us to collect data on occupational mix every 3 years. In response to similar concerns expressed for the 2003 survey, we streamlined the 2006 survey and clarified the instructions in an effort to reduce the burden. We will continue to work with hospitals and associations to explore ways to improve the survey to ensure the accuracy of the occupational mix adjustment while reducing the reporting burden for hospitals.

1. Development of Data for the FY 2007 Occupational Mix Adjustment

In our initial FY 2007 IPPS proposed rule (71 FR 23996), we discussed our proposals for calculating the proposed FY 2007 occupational mix adjustment. We proposed to use the same CMS Wage Index Occupational Mix Survey and Bureau of Labor Statistics (BLS) data that we used for the FY 2005 and FY 2006 wage indices, with a few exceptions. We also proposed to adjust 10 percent of the FY 2007 wage index by a factor reflecting occupational mix. However on April 3, 2006, in Bellevue Hosp. Center v. Leavitt, 443 F.3d 163 (2nd Cir. 2006) the Court of Appeals for the Second Circuit (the Court) ordered CMS to apply the occupational mix adjustment to 100 percent of the wage index effective for FY 2007. The Court ordered CMS to “immediately * * * collect data that are sufficiently robust to permit full application of the occupational mix adjustment.” The Court also ordered that all “data collection and measurement and any other preparations necessary for full application be completed by September 30, 2006, at which time the agency is to immediately apply the adjustment in full.” For more information, we refer the readers to Bellevue Hosp. Center v. Leavitt, 443 F.3d 163, 179 (2nd Cir. 2006).

To comply with the Court's order, on April 21, 2006, we issued a Joint-Signature Memorandum (JSM-06412) to all Medicare fiscal intermediaries announcing our plans to collect new occupational mix data from hospitals. The Joint-Signature Memorandum is available on the CMS Web site at: http://www.cms.hhs.gov/AcuteInpatientPPS. Click on “Wage Index Files” and the link is titled: 2006 Occupational Mix Survey—Interim Data Collection—CMS Memo to Fiscal Intermediaries.

On May 17, 2006, we also published in the Federal Register (71 FR 28644) a second proposed rule that proposed to revise the methodology for calculating the occupational mix adjustment by applying the occupational mix adjustment to 100 percent of the wage index using the new occupational mix data collected from hospitals. The second proposed rule also proposed to modify hospitals' procedures for withdrawing requests to reclassify for the FY 2007 wage index and for supplementing the FY 2008 reclassification application with official data used to develop the FY 2007 wage index. In addition, we proposed to replace in full the descriptions of the data and methodology that would be used in calculating the occupational mix adjustment discussed in the initial FY 2007 IPPS proposed rule.

As stated earlier, section 1886(d)(3)(E) of the Act requires us to conduct a new survey at least once every 3 years. On October 14, 2005, we published a notice in the Federal Register (70 FR 60092) proposing to use a new survey, the 2006 Medicare Wage Index Occupational Mix Survey (the 2006 survey) to apply an occupational mix adjustment to the FY 2008 wage index. In the proposed 2006 survey, we included several modifications based on the comments and recommendations we received on the 2003 survey, including (1) allowing hospitals to report their own average hourly wage rather than using BLS data; (2) extending the prospective survey period; and (3) reducing the number of occupational categories but refining the subcategories for registered nurses.

We made the changes to the occupational categories in response to MedPAC comments to the FY 2005 IPPS final rule (69 FR 49036). Specifically, MedPAC recommended that CMS assess whether including subcategories of registered nurses would result in a more accurate occupational mix adjustment. MedPAC believed that including all registered nurses in a single category may obscure significant wage differences among the subcategories of registered nurses, for example, the wages of surgical registered nurses and floor registered nurses may differ. Also, to offset additional reporting burden for hospitals, MedPAC recommended that CMS should combine the general service categories that account for only a small percentage of a hospital's total hours with the “all other occupations” category because most of the occupational mix adjustment is correlated with the nursing general service category.

In addition, in response to the public comments on the October 14, 2005 notice, we modified the 2006 survey. On February 10, 2006, we published a Federal Register notice (71 FR 7047) that solicited comments and announced our intent to seek OMB approval on the revised occupational mix survey (Form CMS-10079 (2006)).

The revised 2006 survey provides for the collection of hospital-specific wages and hours data, a 6-month prospective reporting period (that is, January 1, 2006, through June 30, 2006), the transfer of each general service category that comprised less than 4 percent of total hospital employees in the 2003 survey to the “all other occupations” category (the revised survey focuses only on the mix of nursing occupations), additional clarification of the definitions for the occupational categories, an expansion of the registered nurse category to include functional subcategories, and the exclusion of average hourly rate data associated with advance practice nurses.

The 2006 survey includes only two general occupational categories: Nursing and “all other occupations.” The nursing category has four subcategories: registered nurses, licensed practical nurses, aides, orderlies, attendants, and medical assistants. The registered nurse subcategory includes two functional subcategories: management personnel and staff nurses or clinicians. As indicated above, the 2006 survey provides for a 6-month data collection period, from January 1, 2006 through June 30, 2006. However, we allowed flexibility for the reporting period begin and end dates to accommodate some hospitals' bi-weekly payroll and reporting systems. That is, the 6-month reporting period must begin on or after December 25, 2005, and must end before July 9, 2006.

To comply with the order of the court in Bellevue Hosp. Center v. Leavitt, as discussed above, we proposed to collect new survey data, instead of using the 2003 survey data proposed in the FY 2007 IPPS proposed rule, to calculate the occupational mix adjustment for the FY 2007 wage index. Because hospitals were already collecting data for the revised 2006 survey, we proposed to use the first 3 months of that data (that is, from January 1, 2006, through March 31, 2006) to calculate the FY 2007 occupational mix adjustment. In order to allow sufficient time for hospitals, fiscal intermediaries, and CMS to collect, review, and correct the new data, and for CMS to perform required analyses and apply the new data in calculating the FY 2007 occupational mix adjustment, we determined that it would be impossible for us to apply the full 6 months of data by October 1, 2006.

Comment: Several commenters stated that hospitals were sometimes unsure of the placement of certain employees on the survey. For example, hospitals were uncertain as to the category that would include surgical technicians and paramedics who are employed by the hospital and who usually work in the emergency department. The commenters urged CMS to evaluate where these employees should be placed on the survey for future collections.

The commenters also stated that they agreed with CMS’ efforts to ensure consistent reporting by specifying the cost centers for collecting nursing personnel data. They agreed that the cost centers included on the survey are where the majority of nurses are employed within hospitals. The commenters added that the use of the cost centers significantly reduces the burden for hospitals by allowing them to focus on only the listed cost centers. However, the commenters urged CMS to consider refining the list of cost centers for future collections. The commenters advised that every hospital has a different method for attributing costs to cost centers; therefore, some hospitals may have a few cost centers that contain a significant number of nursing personnel that were not included in the current survey.

The commenters recommended that CMS work with the hospital community to explore potential changes to the survey occupational categories and cost centers. Even if they are warranted, the commenters suggested that CMS should not make any changes to the ongoing survey collection, as it would necessitate the resubmission of the 1st quarter 2006 data to ensure that both 1st and 2nd quarters could be used for the FY 2008 and the FY 2009 occupational mix adjustment.

Response: We appreciate the assistance we have already received from the hospital community in developing the 2006 occupational mix survey. On May 25, 2006, in response to questions from hospitals and associations, we distributed supplemental instructions to the intermediaries, hospitals (via the intermediaries), and national hospital associations (and posted the instructions on our Web site) to clarify the placement of nursing and nonnursing personnel on the occupational mix survey. We will continue to work with MedPAC and the hospital community to determine if changes to the occupational categories and cost centers included on the survey are reasonable and necessary for future collections. We agree with not changing the instructions for the 2006 survey. As the commenters indicated, to change the survey with the 1st quarter data collection already completed would require substantial rework on the part of hospitals, fiscal intermediaries, and CMS.

Comment: A few commenters expressed concern that hospitals in States with mandatory nurse-staffing ratios for inpatient facilities and hospitals that use higher levels of registered nurses to improve the quality of care will be adversely affected by the occupational mix adjustment. One commenter stated that the current survey is designed to benefit parts of the country that make greater use of lesser skilled nurses and allied health professionals, and to reduce payment in areas that make greater use of registered nurses in nursing positions. The commenter speculated that the occupational mix adjustment will likely reduce the payments for its hospitals, thus reducing the quality of care they can provide to Medicare beneficiaries.

Another commenter indicated that the wage index and occupational mix adjustments penalize hospitals that invest in quality and efficiency at the same time that Congress is trying to improve quality and efficiency under the Medicare program. The commenter stated that the effect of these adjustments on hospitals that use higher levels of registered nurses reduces or eliminates the annual Medicare inflation increase provided to address the increasing costs these hospitals incur. The commenter further indicated that this reduction would not be a savings to the program, but rather it would be a redistribution of Medicare payments to hospitals that have not been as efficient or as focused on improving the quality of care.

Response: As stated earlier, the statute requires implementation of an occupational mix adjustment to the wage index. In addition, the purpose of the occupational mix adjustment is to control the effect of a hospital's employment mix on its average hourly wage for the wage index. The adjustment standardizes the employment mix for hospitals so that the wage index more accurately compares wage rates among labor market areas for a constant mix of labor. As the commenters noted, the occupational mix adjustment would lower the wage index for an area employing a mix of more highly paid and skilled labor than the national average. Although we understand the commenters’ concerns regarding the effect of the occupational mix on their areas’ Medicare payments, we disagree that the wage index and occupational mix adjustments penalize hospitals that invest in quality and efficiency. We note that CMS is moving toward adoption of a severity-based DRG system that will better recognize severity of illness and provide improved payments to those hospitals that need more highly skilled labor to care for more severely ill patients. Even under the current system, the labor costs incurred by hospitals that provide more highly skilled services are currently reflected in the hospital's DRG payments and illustrated through a higher case mix index. Reflecting the costs associated with more highly skilled labor in both the case mix and the wage index is essentially counting them twice.

To comply with the order of the court in Bellevue Hosp. Center v. Leavitt, as a final policy, we are adopting our proposal to use the new 1st quarter 2006 survey data to calculate the occupational mix adjustment for the FY 2007 wage index.

2. Timeline for the Collection, Review, and Correction of the Occupational Mix Data

The Joint-Signature Memorandum (JSM-06412) that we issued on April 21, 2006, instructed all fiscal intermediaries to immediately alert the hospitals they service to the changes in the schedule for submitting the occupational mix data files.

The Joint-Signature Memorandum provided hospitals and fiscal intermediaries with the revised schedule for the occupational mix survey data that would be used in the FY 2007 wage index. The schedule included deadlines for—

• Hospitals to submit occupational mix data. The deadline was June 1, 2006.
• Fiscal intermediary review of the submitted data. The deadline was June 22, 2006.
• Availability of the submitted data on the CMS Web site. The deadline was June 29, 2006.
• Hospitals to submit requests to their fiscal intermediaries for corrections to their interim occupational mix data. The deadline was July 13, 2006.
• Fiscal intermediaries to submit corrected interim occupational mix survey data for the January 1, 2006, through March 31, 2006 period. The deadline was July 27, 2006.

We noted that it was critical that hospitals provide information according to the dates provided in the schedule in order to be able to appeal any disputed calculations at a later point to the Provider Review Reimbursement Board (PRRB). The final deadline for the fiscal intermediaries to make occupational mix data available to CMS was July 27, 2006. These data would reflect fiscal intermediary review and the resolution of any errors or adjustments between the hospitals and fiscal intermediary. Once these data are available on the CMS Web site, changes to a hospital's occupational mix data would be allowed only in those very limited situations involving an error by the fiscal intermediary or CMS that the hospital could not have known about before its review of the final occupational mix data file. Specifically, neither the fiscal intermediary nor CMS would approve the following types of requests:

• Requests for occupational mix data corrections that were submitted too late to be included in the data transmitted to CMS by fiscal intermediaries on or before July 27, 2006.
• Requests for correction of errors that were not, but could have been, identified during the hospital's review of the June 29, 2006 occupational mix file.

Verified corrections to the occupational mix received by the fiscal intermediaries and CMS (that is, by July 13, 2006) would be incorporated into the final wage index for FY 2007, to be effective October 1, 2006.

We created the process described above to resolve all substantive occupational mix correction disputes before we finalize the wage and occupational mix data for the FY 2007 payment rates. Accordingly, hospitals that did not meet the procedural deadlines set forth above will not be afforded a later opportunity to submit occupational mix data corrections or to dispute the fiscal intermediary's decision with respect to requested changes. Specifically, our policy is that hospitals that do not meet the procedural deadlines set forth above will not be permitted to challenge later, before the PRRB, the failure of CMS to make a requested data revision. (See W.A. Foote Memorial Hospital v. Shalala, No. 99-CV-75202-DT (E.D. Mich.2001) and Palisades General Hospital v. Thompson, No. 99-1230 (D.D.C. 2003)). We also refer the reader to the FY 2000 IPPS final rule (64 FR 41513) for a discussion of the parameters for appealing to the PRRB for wage index data corrections.

We believe the occupational mix data correction process described above provided hospitals with the opportunity to bring errors in their occupational mix data to the fiscal intermediary's attention.

Because hospitals had access to the final occupational mix data by June 29, 2006, we believe they had the opportunity to detect any data entry or tabulation errors made by the fiscal intermediary or CMS before the development and publication of the final FY 2007 wage index and the implementation of the FY 2007 wage index on October 1, 2006. We believe that if hospitals availed themselves of the opportunities afforded to provide and make corrections to the occupational mix data, the wage index implemented on October 1, 2006, will be accurate. In the event that errors are identified by hospitals and brought to our attention after July 13, 2006, we will only make mid-year changes to the wage index in accordance with § 412.64(k). For a detailed discussion, see section III.J. of this preamble.

Comment: One commenter stated that the 6-month reporting period for the 2006 survey, originally planned for the FY 2008 wage index, is an improvement over the 2003 survey process. However, the commenter urged CMS to initiate a survey with a full-year reporting period for the FY 2009 wage index.

Response: We appreciate the commenter's recognition of our efforts to improve the occupational mix survey process. We also appreciate the commenter's suggestion for expanding the survey reporting period to a full year for the FY 2009 wage index. While we appreciate the willingness expressed in the comment to collect a complete year of data in order to achieve more accurate survey results, we note that hospitals are currently obligated to collect data for the period April 1, 2006, to June 30, 2006, by August 31 in order for us to use 6 months of data to apply the occupational mix adjustment for FY 2008. If we were to use a full year of 2006 survey data to apply an occupational mix adjustment for FY 2009, hospitals would have to submit data for the last 6 months of calendar year 2006. Hospitals have already been required to submit occupational mix survey data for two different 3-month periods in 2006. At this time, we believe it would be burdensome to require a third occupational mix data collection from hospitals for 2006 in order to apply the adjustment based on a full year of data for FY 2009. We also note that collecting a full year of calendar year 2007 data, from January 1, 2007, through December 31, 2007, would not provide enough time for a thorough review and correction period before the FY 2009 proposed rule would be published in April 2008. Our normal wage index review and correction process before the proposed rule publication begins in early October and ends in late February. This would mean that hospitals and intermediaries would have only approximately 2 months, from January to late February, to review and correct a year's worth of occupational mix data. We believe that such an abbreviated review and correction period would not be in the hospitals' best interest. However, we will consider expanding the survey reporting period to a full year for a future collection.

Comment: Some commenters expressed concern that the 3-month survey period for FY 2007 will lead to inaccurate results for several reasons: Having no advance notice of the expedited data collection; some hospitals had not yet begun, or had just begun, to plan for the 2006 survey data collection and had little or no resources available to complete the survey for all or part of the 3-month time period; the new survey, though improved over the previous survey, is more complicated and requires more effort to complete; due to the short timeframe for developing and submitting the data (4 months), some normal review processes had to be eliminated by hospitals; not enough time was allowed for the types of corrections that can be made during the annual wage index survey process; due to the infrequent collection of the occupational mix data, many hospitals may underestimate its importance; there was not enough time for hospital groups to review the data for individual hospitals in the area, a process that often raises questions that leads to more accurate data.

Response: We understand the commenters’ concerns about the potential for inaccurate occupational mix survey data to be used due to the abbreviated data collection and reporting periods. However, CMS has established a process that we believe will maximize the opportunity for accurate occupational mix data to be used to adjust area wage indices. Hospitals were required to submit occupational mix survey data to their fiscal intermediaries by June 1, 2006. CMS provided fiscal intermediaries with a desk review program to assist in identifying erroneous or aberrant data. Fiscal intermediaries then had 3 weeks (or until June 22) to review the data and submit it to CMS. CMS made the occupational mix survey data available on the CMS Web site on June 29 to facilitate review by hospitals, fiscal intermediaries, and others. The June 29 posting of occupational survey data resulted in hospitals, State hospital associations, wage index consultants, and others identifying errors and other aberrant data. These parties then initiated action to correct the occupational mix survey data by the July 13 deadline. While there is no additional time available to correct the survey data for the FY 2007 wage index, we will, however, allow hospitals to submit any additional revisions and corrections to both 3-month periods of data for the FY 2008 wage index. We strongly encourage hospitals to take full advantage of the FY 2008 wage index correction process. Hospitals will be notified early in the Fall of 2006 regarding the revision/correction process for the FY 2008 wage index for both the cost report wage data and the 2006 occupational mix survey data.

3. Calculation of the Occupational Mix Adjustment

In the May 17, 2006 proposed rule, we proposed a series of steps to be used in calculating the FY 2007 occupational mix adjustment factor. In this final rule, we are adopting the proposed steps with one minor exception. In response to comments (discussed below), we have made an adjustment to step 7 so that the percentage of worker salaries attributable to the nursing category is based on salaries and not on hours. For 2007, we will calculate the occupational mix adjustment factor using the following steps:

Step 1—For each hospital, determine the percentage of the total nursing category attributable to a nursing subcategory by dividing the nursing subcategory hours by the total nursing category's hours (registered nurse management personnel and registered nurse staff nurses or clinicians are treated as separate nursing subcategories). Repeat this computation for each of the five nursing subcategories: Registered nurse management personnel, registered nurse staff nurses or clinicians, licensed practical nurses; nursing aides, orderlies, and attendants; and medical assistants.

Step 2—Determine a national average hourly rate for each nursing subcategory by dividing a subcategory's total salaries for all hospitals in the occupational mix survey database by the subcategory's total hours for all hospitals in the occupational mix survey database.

Step 3—For each hospital, determine an adjusted average hourly rate for each nursing subcategory by multiplying the percentage of the total nursing category (from Step 1) by the national average hourly rate for that nursing subcategory (from Step 2). Repeat this calculation for each of the five nursing subcategories.

Step 4—For each hospital, determine the adjusted average hourly rate for the total nursing category by summing the adjusted average hourly rate (from Step 3) for each of the nursing subcategories.

Step 5—Determine the national average hourly rate for the total nursing category by dividing total nursing category salaries for all hospitals in the occupational mix survey database by total nursing category hours for all hospitals in the occupational mix survey database.

Step 6—For each hospital, compute the occupational mix adjustment factor for the total nursing category by dividing the national average hourly rate for the total nursing category (from Step 5) by the hospital's adjusted average hourly rate for the total nursing category (from Step 4).

If the hospital's adjusted average hourly rate is less than the national average hourly rate (indicating the hospital employs a less costly mix of nursing employees), the occupational mix adjustment factor would be greater than 1.0000. If the hospital's adjusted average hourly rate is greater than the national average hourly rate, the occupational mix adjustment factor would be less than 1.0000.

Step 7—For each hospital, calculate the occupational mix adjusted salaries and wage-related costs for the total nursing category by multiplying the hospital's total salaries and wage-related costs (from Step 5 of the unadjusted wage index calculation in section III.F. of this preamble) by the percentage of the hospital's total workers attributable to the total nursing category (using the occupational mix survey data, this percentage is determined by dividing the hospital's total nursing category salaries by the hospital's total salaries for “nursing and all other”) and by the total nursing category's occupational mix adjustment factor (from Step 6 above).

The remaining portion of the hospital's total salaries and wage-related costs that is attributable to all other employees of the hospital is not adjusted by the occupational mix. A hospital's all other portion is determined by subtracting the hospital's nursing category percentage from 100 percent.

Step 8—For each hospital, calculate the total occupational mix adjusted salaries and wage-related costs for a hospital by summing the occupational mix adjusted salaries and wage-related costs for the total nursing category (from Step 7) and the portion of the hospital's salaries and wage-related costs for all other employees (from Step 7).

To compute a hospital's occupational mix adjusted average hourly wage, divide the hospital's total occupational mix adjusted salaries and wage-related costs by the hospital's total hours (from Step 4 of the unadjusted wage index calculation in section III.F. of this preamble).

Step 9—To compute the occupational mix adjusted average hourly wage for an urban or rural area, sum the total occupational mix adjusted salaries and wage-related costs for all hospitals in the area, then sum the total hours for all hospitals in the area. Next, divide the area's occupational mix adjusted salaries and wage-related costs by the area's hours.

Step 10—To compute the national occupational mix adjusted average hourly wage, sum the total occupational mix adjusted salaries and wage-related costs for all hospitals in the Nation, then sum the total hours for all hospitals in the Nation. Next, divide the national occupational mix adjusted salaries and wage-related costs by the national hours.

Step 11—To compute the occupational mix adjusted wage index, divide each area's occupational mix adjusted average hourly wage (Step 9) by the national occupational mix adjusted average hourly wage (Step 10).

Step 12—To compute the Puerto Rico specific occupational mix adjusted wage index, follow Steps 1 through 11 above.

Comment: MedPAC and a few other commenters noted that Step 7 of CMS’ proposed calculation for the occupational mix adjustment uses the occupational mix survey's paid hours to determine the portion of the salaries and wage-related costs to adjust for occupational mix (that is, the total nursing portion) and the portion to remain unadjusted (that is, the all other occupations portion). One of the commenters stated that this approach was reasonable using the 2003 survey data because hospital-specific paid salaries data were not collected. However, the commenter also noted that the actual share of wages for either the nursing category or the all other occupations category could differ using an allocation that is based on paid hours compared to paid salaries. The commenters suggested that, since the 2006 survey provides for the collection of paid salaries data, CMS should use paid salaries instead of paid hours to more accurately determine the wage costs that should be adjusted for occupational mix and those that should not.

Response: As discussed above, we evaluated the commenters’ recommendation and agree that it is reasonable to use the occupational mix survey salaries instead of hours in computing the portion of a hospital's salaries and wage-related costs to adjust for occupational mix and the portion to remain unadjusted. Accordingly, we revised Step 7 of the final calculation for the occupational mix adjustment to reflect this change.

We received no other comments on the steps used in calculating the occupational mix adjustment. As a final policy, we are adopting the proposed calculation, with the change to Step 7, for the occupational mix adjustment to the FY 2007 wage index. Also, to comply with the order of the court in Bellevue Hosp. Center v. Leavitt, we will apply this adjustment to 100 percent of the wage index.

The table below is an illustrative example of the final occupational mix adjustment. (Note: We have revised this example from that included in the proposed rule to reflect the change in step 7 discussed above. We have added an additional column for provider occupational mix salaries and the Provider Percent by Total is determined by dividing the hospital's total nurse salaries (and separately, Total All Other Salaries) by Total Employee Salaries.

Because the occupational mix adjustment is required by statute, all hospitals that are subject to payments under the IPPS, or any hospital that would be subject to the IPPS if not granted a waiver, must complete the occupational mix survey, unless the hospital has no associated cost report wage data that are included in the FY 2007 wage index.

For the FY 2005 and FY 2006 final wage indices, we used the unadjusted wage data for hospitals that did not submit occupational mix survey data. For calculation purposes, this equates to applying the national nursing mix to the wage data for these hospitals, because hospitals having the same mix as the Nation would have an occupational mix adjustment factor equaling 1.0000. However, an adjustment may not be equitable in situations where the hospital has a higher or lower than average occupational mix than the Nation as a whole. If the hospital's occupational mix is higher than the average for the nation as a whole, hospitals in other areas are disadvantaged by the hospital not providing occupational mix information. If the hospital's occupational mix is lower than the average for the Nation as a whole, other hospitals in the same geographic area would be disadvantaged by the hospital not providing the information.

In the FY 2005 and FY 2006 IPPS final rules (69 FR 49035 and 70 FR 47368), we noted that we would revisit this matter with subsequent collections of the occupational mix data. In the May 17, 2006 proposed rule, for the FY 2007 wage index, we proposed to use one of four options for treating the occupational mix data for nonresponsive hospitals: (1) Assign the hospital an occupational mix adjustment factor of 1.0000 as we did for FY 2005 and FY 2006; (2) assign the hospital the average occupational mix adjustment factor for its labor market area; (3) assign the hospital the lowest occupational mix adjustment factor for its labor market area; or (4) assign the hospital the average occupational mix factor for similar hospitals, based on factors such as, geographic location, bed size, teaching versus non-teaching status and case mix. We requested comments on these or other alternatives for equitably addressing the situation of hospitals that are not responsive to the occupational mix survey.

Comment: A majority of the commenters believed that, in order for the wage index to be computed accurately, it is critical for all IPPS hospitals to complete the occupational mix survey. Many of the commenters suggested that CMS should penalize hospitals that did not submit a survey. However, the commenters indicated that no hospitals should be penalized for not completing the survey for the 1st quarter of FY 2006 (to be used in calculating the FY 2007 wage index) because of the short notification and timeframe for the collection of that data. Some suggested future penalties such as a 1 to 2 percent reduction in the hospital's wage index value or a set percentage of the standardized amount, whichever is administratively feasible. However, the commenters also suggested that any penalty should be hospital-specific and should not affect the wage index amounts for other hospitals in the area. Commenters suggested that CMS should first calculate the area wage index using proxy data for a nonresponsive hospital's occupational mix adjustment, and then CMS should assess a penalty on its wage index value or national standardized amount.

The commenters supported all of the ideas we raised in the proposed rule except option 3. Commenters unanimously opposed assigning the hospital the lowest occupational mix adjustment factor for its labor market area, because they believed this option would have the most negative impact on other hospitals in the labor market area. MedPAC recommended option 4, to assign the hospital the average occupational mix factor for similar hospitals, based on factors such as, geographic location, bed size, teaching versus nonteaching status and case-mix. MedPAC suggested other factors that CMS should consider, such as share of ICU days and types of services offered. Some commenters recommended an option that we did not describe. These commenters recommended that CMS substitute data from the previous 2003 survey for hospitals that did not submit 2006 survey data for the FY 2007 wage index. Alternatively, several commenters recommended that CMS could substitute the national average hourly wage (that is, option 1, an occupational mix adjustment of 1.0000) for nonresponsive hospitals in calculating an area's wage index, while others favored option 2 because it would have the least affect on the labor market area. One commenter recommended assigning the lower of the hospital's occupational mix adjustment in FY 2006 or the average for the hospital's labor market in FY 2007. The commenter believed that the best proxy for a hospital's missing FY 2007 data is its FY 2006 occupational mix adjustment, even though there was a change in the formula to calculate the FY 2007 adjustment. The commenter stated that CMS should provide an exception for an exogenous event affecting all hospitals in the labor market area. In this scenario, the commenter recommended using the average FY 2007 adjustment.

Response: We agree with the commenters that hospitals that did not respond to the occupational mix survey should not benefit from the participation of others. We also agree that, due to the unusual circumstances of the Court's order and the short timeframe that hospitals were provided for completing and submitting their data, it would not be fair to apply a penalty to nonresponsive hospitals for the 2007 wage index. However, we believe that section 1886(d)(5)(I)(i) of the Act provides us with the authority to penalize hospitals that do not submit occupational mix survey data. That section authorizes us to provide for exceptions and adjustments to the payment amounts under IPPS as the Secretary deems appropriate. We will give serious consideration to applying a hospital-specific penalty such as those suggested by the commenters if a hospital does not comply with regulations requiring submission of occupational mix survey data in future years. We will address this issue in the FY 2008 IPPS proposed rule.

Regarding the treatment of data for nonresponsive hospitals, we have chosen not to adopt option 3, because it would be punitive to other hospitals in the area that submitted occupational mix data. We also have not chosen option 1 because it does not provide an incentive for hospitals to respond if they have a higher mix of employees than the national average. We will not use data from the 2003 survey, as some commenters suggested, because the 2007 wage index, we believe, should be exclusively based on the newly collected data. In addition, there was concern about the sufficient robustness of such data to support 100 percent adjustments. We also do not believe it would be entirely feasible, for 2007, to implement MedPAC's recommendation, option 4, due to the wide range of parameters that could be used for developing proxies for the missing hospitals and the fact that the exact set of such parameters was not subject to comment. So many variables might be of relevance that our selection of any particular variables might be subject to controversy, and hospitals may wish to have an opportunity to comment on the exact variables that would be used. MedPAC's recommendation to add more variables to further refine the analysis could be so limiting as to result in few or no hospitals to use for comparison.

For the FY 2007 wage index, we have adopted option 2—using the average occupational mix adjustment for the labor market area. We believe this option would have the least impact on the wage index for other hospitals in the area and does not have the disadvantages of the options discussed above. Although we believe this option is the best of the ones we considered for nonresponsive hospitals for FY 2007, we reserve the right to apply a different approach in future years, including potentially penalizing nonresponsive hospitals. If there is only one hospital in the labor market area, and that hospital failed to submit occupational mix data, or, if there are no hospitals in the labor market area, we would apply the national occupational mix factor of 1.0000 in calculating the area's FY 2007 occupational mix adjusted wage index.

Comment: Some commenters recommended that CMS allow hospitals that failed to submit their 1st quarter data by June 1, 2006, to submit that data when the 2nd quarter data is due (that is, by August 31, 2006). The commenters also suggested that CMS allow hospitals that submitted their 1st quarter data by June 1, an opportunity to correct that data when the 2nd quarter data are due. The commenters indicated that allowing hospitals to submit the data at this time would improve the survey response rate and eliminate the need for penalties for hospitals that would otherwise be nonresponsive and improve the accuracy of the data for the FY 2008 and the FY 2009 occupational mix adjustment.

Response: We agree with the commenters. Hospitals that did not submit occupational mix data for the 1st quarter of 2006 will be permitted to submit 1st and 2nd quarter data by August 31. We included the 1st quarter data for some hospitals that submitted survey data after June 1. However, submissions that were received too late to include in the FY 2007 occupational mix adjustment will be included in the desk review process for the occupational mix adjustment for the FY 2008 wage index. As we previously mentioned, we will also allow hospitals an opportunity to revise both their 1st quarter and 2nd quarter 2006 occupational mix data for the FY 2008 wage index. Further, we stated that we will notify hospitals early in the Fall of 2006 regarding the revision/correction process for the FY 2008 wage index for both the cost report wage data and the 2006 occupational mix survey data.

D. Worksheet S-3 Wage Data for the FY 2007 Wage Index

The FY 2007 wage index values (effective for hospital discharges occurring on or after October 1, 2006, and before October 1, 2007) that will be published separately from this final rule will be based on the data collected from the Medicare cost reports submitted by hospitals for cost reporting periods beginning in FY 2003 (the FY 2006 wage index was based on FY 2002 wage data).

The FY 2007 wage index will include the following categories of data associated with costs paid under the IPPS (as well as outpatient costs):

• Salaries and hours from short-term, acute care hospitals (including paid lunch hours and hours associated with military leave and jury duty).
• Home office costs and hours.
• Certain contract labor costs and hours (which includes direct patient care, certain top management, pharmacy, laboratory, and nonteaching physician Part A services).
• Wage-related costs, including pensions and other deferred compensation costs.

Consistent with the wage index methodology for FY 2006, the final wage index for FY 2007 also will exclude the direct and overhead salaries and hours for services not subject to IPPS payment, such as SNF services, home health services, costs related to GME (teaching physicians and residents) and certified registered nurse anesthetists (CRNAs), and other subprovider components that are not paid under the IPPS. The final FY 2007 wage index also will exclude the salaries, hours, and wage-related costs of hospital-based rural health clinics (RHCs), and Federally qualified health centers (FQHCs) because Medicare pays for these costs outside of the IPPS (68 FR 45395). In addition, salaries, hours, and wage-related costs of CAHs will be excluded from the wage index, for the reasons explained in the FY 2004 IPPS final rule (68 FR 45397).

Data collected for the IPPS wage index are also currently used to calculate wage indices applicable to other providers, such as SNFs, home health agencies, and hospices. In addition, they are used for prospective payments to IRFs, IPFs, and LTCHs, and for hospital outpatient services. We note that, in the IPPS rules, we do not address comments pertaining to the wage indices for non-IPPS providers. Such comments should be made in response to separate proposed rules for those providers.

Comment: Several commenters addressed CMS’ policy of excluding data from CAHs when computing the wage index. They stated that, as of FY 2007, 1,191 CAHs (representing approximately 24 percent of all IPPS hospitals in FY 2000, and approximately 55 percent of all rural hospitals in FY 2000) have been removed from the wage index. The commenters indicated that CAHs have lower average hourly wages than the typical IPPS hospital and eliminating their data from the wage index overstates the national average hourly wage by an estimated 0.707 percent. They added that increases in the national average hourly wage, in turn, are offset with the application of a negative budget neutrality adjustment, which understates IPPS operating payments according to the commenters. The commenters believed that the artificial increase in the national average hourly wage has lowered the budget neutrality adjustment by an estimated $1.52 billion over 5 years (2003-2007). The commenters stated that CMS should apply a one-time positive budget neutrality adjustment in FY 2007 to compensate for the prior underpayments. They did not believe similar future adjustments would be necessary because very few hospitals are expected “to convert to CAH status now that the necessary provider designation is no longer an option.”

Other commenters asked that CMS use estimated CAH wage data to compute the FY 2007 wage index, and that an occupational mix factor of 1.0000 be assigned to these hospitals. The commenters noted that MedPAC has recommended that CAH data be included in the wage index, at least in computing the national average hourly wage. The commenters asserted that because CAHs in rural areas still compete with rural IPPS hospitals for scarce resources, their data should be included in the wage index.

Commenters also requested that CMS obtain wage data from CAHs and subject that data to the same rigorous review by the fiscal intermediaries as is done for IPPS hospitals. Another commenter suggested that an alternative to including the CAHs in wage index would be to not factor in any increases in the national average hourly wage that are attributable to the removal of CAHs’ wage data.

Response: In the August 1, 2003 final rule (68 FR 45397-8), we explained the reasons for our decision to remove CAH data from the wage index immediately upon conversion to CAH status, even if the hospital was paid under the IPPS during the cost reporting period used in calculating the current fiscal year's wage index. The primary reason for excluding CAHs from the wage index was that they are a separate provider type and are unique compared to other short term, acute care hospitals with respect to factors such as their location and bed size. We discussed the payment impact, mentioning the substantial negative impact CAHs typically have on the wage indexes in the areas where they are located, and the minimal impact they have on other areas. We also stated that we would not be holding other hospitals’ payments harmless for this change, consistent with our general wage index policy.

As the commenters indicated, in the FY 2006 IPPS final rule, we addressed a comment from MedPAC recommending that data from CAHs be included in the wage index (70 FR 47370). MedPAC had recommended that CMS begin collecting wage data from CAHs in 2005. Although we agree with MedPAC that CAHs have recently become more similar to other rural hospitals, in structure, location, and services provided, largely due to changes in the CAH statute resulting from section 405 of Pub. L. 108-173 (MMA), the wage index must be based on data from “subsection (d),” short-term, acute care hospitals, consistent with section 1886(d)(3)(E) of the Act. Therefore, we cannot use any wage data collected from CAHs in the IPPS wage index. Because Pub. L. 108-173 was enacted at the end of calendar year 2003, it would not affect the wage index at least until FY 2008, which would be computed from cost reporting periods beginning in FY 2004. Accordingly, we continue to believe that it has been prudent policy to remove the wage data for hospitals that later became CAHs from the wage index.

We do not believe that the elimination of these data has resulted in an overstated national average hourly wage, nor has the budget neutrality adjustment been inappropriately reduced. The national average hourly wage appropriately reflects only those wages paid by IPPS hospitals. To determine the budget neutrality adjustment for FY 2007, we equate IPPS payments using the FY 2006 and FY 2007 wage indices using FY 2005 MedPAR data that excludes any hospitals that became CAHs as of February 17, 2006. The calculation excludes CAHs from the determination of IPPS payments using both the FY 2006 and FY 2007 wage indices so the budget neutrality adjustment reflects only information from IPPS hospitals and is not overstated. Consequently, we will not apply a one-time positive budget neutrality adjustment in FY 2007.

E. Verification of Worksheet S-3 Wage Data

The wage data for the final FY 2007 wage index will be obtained from Worksheet S-3, Parts II and III of the FY 2003 Medicare cost reports. Instructions for completing the Worksheet S-3, Parts II and III are in the Provider Reimbursement Manual, Part I, sections 3605.2 and 3605.3. The data file used to construct the wage index will include FY 2003 data submitted to us as of June 28, 2006. As in past years, we will perform an intensive review of the wage data, mostly through the use of edits designed to identify aberrant data.

We asked our fiscal intermediaries to revise or verify data elements that resulted in specific edit failures. While some of the edits failures were resolved, we did remove the wage data of some hospitals from the final FY 2007 wage index. For the final FY 2007 wage index in this final rule, we removed the data for 229 hospitals from our database: 189 hospitals designated as CAHs by 7 or more days prior to the posting of the preliminary February public use file, and 30 hospitals were low Medicare utilization hospitals or failed edits that could not be corrected because the hospitals terminated the program or changed ownership. In addition, we removed the wage data for 10 hospitals with incomplete or inaccurate data resulting in zero or negative, or otherwise aberrant, average hourly wages. As a result, the final FY 2007 wage index is calculated based on FY 2003 wage data from 3,570 hospitals.

In constructing the final FY 2007 wage index, we will include the wage data for facilities that were IPPS hospitals in FY 2003, even for those facilities that have since terminated their participation in the program as hospitals, as long as those data do not fail any of our edits for reasonableness. We believe that including the wage data for these hospitals is, in general, appropriate to reflect the economic conditions in the various labor market areas during the relevant past period. However, we exclude the wage data for CAHs as discussed in 68 FR 45397.

Section 4410 of Pub. L. 105-33 provides that, for the purposes of section 1886(d)(3)(E) of the Act, for discharges occurring on or after October 1, 1997, the area wage index applicable to any hospital that is located in an urban area of a State may not be less than the area wage index applicable to hospitals located in rural areas in the State. This provision is commonly referred to as the “rural floor.” In the August 11, 2004 IPPS final rule (69 FR 49109), we discussed situations where a State has only urban areas and no geographically rural areas, or a State has geographically rural areas but no IPPS hospitals are located in those rural areas. As a result, these States did not have rural IPPS hospitals from which to compute and apply a “rural floor.” In that final rule, we developed a policy for imputing a “rural floor” for these States, effective for the FYs 2005, 2006, and 2007 wage indices, so that a “rural floor” could be applicable to IPPS urban hospitals in those States in the same manner that a “rural floor” is applicable to IPPS urban hospitals in States that have IPPS rural hospitals. We revised the regulations at § 412.64(h) to describe the methodology for computing the imputed “rural floors” for these States and to define an all-urban State. Specifically, § 412.64(h)(5) defines an all-urban State as “a State with no rural areas * * * or a State in which there are no hospitals classified as rural. A State with rural areas and with hospitals reclassified as rural under § 412.103 is not an all-urban State.”

We have received questions as to what area wage index CMS would apply in the instance where a new rural IPPS hospital opens in a State that has an imputed “rural floor” because it has rural areas but had no hospitals classified as rural. In addition, we have been asked whether a new IPPS hospital could submit its wages and hours data to be used in computing the wage index, even though the hospital did not file a cost report as an IPPS provider for the cost report base year that is used in calculating that wage index.

A new hospital can be an entirely new facility that did not exist before, or it can be a hospital that participated in Medicare under a previous provider number, but has acquired a new Medicare provider number (such as when a CAH converts to IPPS status, or vice versa). As a new IPPS hospital (in this case, rural), the hospital would not yet have filed any wages and hours data on a Medicare cost report. Even in the situation where a new IPPS hospital previously participated in Medicare as a non-IPPS provider, wages and hours data collected as a non-IPPS provider would not be suitable for calculating an IPPS wage index because section 1886(d)(3)(E) of the Act specifies that the wage index must be based on data from “subsection (d)” hospitals. Thus, CMS could not include wages and hours from a period during which a hospital was not an IPPS provider. Furthermore, even once the hospital files its first Medicare cost report under the new IPPS provider number, that first cost report is not used in computing the wage index for the hospital's geographic area until 4 years later (for example, we use the 2003 data to compute the wage index for FY 2007). Therefore, if a new rural IPPS hospital opens in a State that has an imputed “rural floor” and has rural areas, for FY 2007, the hospital would receive the imputed “rural floor” as its wage index. The imputed rural floor is set to expire on September 30, 2007. However, we expect that we would address the 2008 implications for a new rural hospital that is the only rural hospital in the State in the FY 2008 proposed rule.

Comment: Two commenters stated that CMS’ above policy conflicts with the policy of excluding the wage data of IPPS hospitals that convert to CAH status. The commenters also asserted that in the years before the hospital's own wage data is used, the rural hospital will be paid at the imputed rural floor, which they contend is unrelated to the hospital's own labor market costs. The commenters also asserted that if the new rural hospital's average hourly wage is greater than the imputed rural floor, the hospital would suffer underpayments until its index could be based upon its own wage data. One commenter suggested that, at least for CAHs converting to IPPS status, CMS should use wage data filed by the hospital when it was a CAH.

The commenters urged CMS to include the wage data of a new rural IPPS hospital in the wage index “as soon as a full year's cost report with the hospital operating as a PPS hospital is available.”

Response: We disagree with the commenters. Our consistent policy is that new hospitals must first develop their wage data and have it reviewed by our fiscal intermediaries prior to the wage data being included in the wage index. The submission and review process requires a 4-year period, in order to allow time for all hospitals to complete and submit their wage data for the fiscal year, for the fiscal intermediaries to review the data, for the fiscal intermediaries to present the results of their review to hospitals, for hospitals to review any potential errors in the wage index files, for us to resolve any disputes between the fiscal intermediary and the hospital, and finally, for the final wage indices to be calculated and published in advance of the fiscal year. For a discussion of the wage data review and correction process, refer to section III.J. of this preamble. This policy applies to all new hospitals, not just rural hospitals. Although a new rural IPPS hospital that previously was a CAH may be willing to provide CMS with wage data from the period during which it was a CAH, the wage index must be based on data from IPPS hospitals, consistent with section 1886(d)(3)(E) of the Act. A CAH is not an IPPS hospital; thus, we cannot include the hospital's wages and hours from the period during which it was a CAH. Indeed, even if a CAH previously existed as an IPPS hospital (that is, it previously was an IPPS hospital, converted to CAH status, and then converted back to IPPS status), its historical wage data would have been submitted from years prior to the cost reports used to calculate the FY 2007 wage index (that is, the FY 2003 cost reports). If a CAH converts back to IPPS status in FY 2007, there would be no wage data for the FY 2007 wage index because such a provider did not file Medicare cost reports as an IPPS provider in FY 2003.

We recognize, as one commenter pointed out that in the past we have noted the importance of including “all” available wage data in the wage index calculation. However, our past statements to this effect were discussing the inclusion of all IPPS hospital wage data, not data from non-IPPS hospitals. In the FY 2003 IPPS final rule (67 FR 50023), we discussed our policy of including data from IPPS hospitals that have since closed. We stated that such data should be included because, “any hospital that is in operation during the data collection period used to calculated the wage index should be included in the database, since the hospital's data reflect conditions occurring in that labor market area during the period surveyed.” Our statement, however, was directed at the inclusion of IPPS hospital data—not the inclusion of data from hospitals that were not IPPS hospitals during the data collection period. As stated earlier, section 1886(d)(3)(E) of the Act requires the wage index to be based upon a survey of “subsection (d) hospitals.”

Lastly, we think it is false logic to state that our policy excluding data from hospitals that become CAHs necessarily requires inclusion of data from hospitals that switch from CAH status to IPPS status. As stated in the FY 2003 IPPS final rule, we exclude hospitals that convert to CAH status because our analysis showed that the wage data for these hospitals, in general, are significantly different from other short-term hospitals (68 FR 45397). CAHs that convert to IPPS status, in contrast, could not, under the statute, be included in the wage index survey because they are not IPPS hospitals at the time of the survey.

Comment: A few commenters recommended that CMS propose now to extend the imputed rural floor to coincide with the rural floor established under section 4410 of Pub. L. 105-33, in order to place all 50 states on a level playing field.

Response: As stated above, our policy for imputing a “rural floor” is effective for the FYs 2005, 2006, and 2007 wage indices. We will determine the appropriateness of extending that policy beyond FY 2007 and state our proposal in the FY 2008 proposed rule. Commenters will be have sufficient time during the FY 2008 IPPS comment period to assess and comment on such a proposal.

Comment: One commenter suggested that CMS should select one national contractor as part of the Medicare Administrative Contractor (MAC) bidding process (provided for under section 1847A of the Act as added by section 911 of Pub. L. 108-173) to do wage index reviews. The commenter believed that the use of the MAC process to solicit a single “national” contractor would ensure that the wage data and occupational mix data reviews are handled consistently and accurately, so that all hospitals are subject to the same policy interpretations. The commenter noted the importance of the wage index in determining Medicare payments to hospitals and indicated that any variation among contractors in the handling of hospitals’ wage index data could be detrimental to hospitals in certain geographic regions. The commenter also stated that the inclusion of a 100 percent occupational mix adjustment intensifies the need for a contractor approach going forward.

Response: We appreciate the suggestion and will consider it as we develop our program acquisition strategies.

F. Computation of the FY 2007 Unadjusted Wage Index

The method used to compute the FY 2007 wage index without an occupational mix adjustment follows:

Step 1—As noted above, we based the FY 2007 wage index on wage data reported on the FY 2003 Medicare cost reports. We gathered data from each of the non-Federal, short-term, acute care hospitals for which data were reported on the Worksheet S-3, Parts II and III of the Medicare cost report for the hospital's cost reporting period beginning on or after October 1, 2002, and before October 1, 2003. In addition, we include data from some hospitals that had cost reporting periods beginning before October 2002 and reported a cost reporting period covering all of FY 2003. These data are included because no other data from these hospitals would be available for the cost reporting period described above, and because particular labor market areas might be affected due to the omission of these hospitals. However, we generally describe these wage data as FY 2003 data. We note that, if a hospital had more than one cost reporting period beginning during FY 2003 (for example, a hospital had two short cost reporting periods beginning on or after October 1, 2002, and before October 1, 2003), we include wage data from only one of the cost reporting periods, the longer, in the wage index calculation. If there was more than one cost reporting period and the periods were equal in length, we include the wage data from the later period in the wage index calculation.

Step 2—Salaries—The method used to compute a hospital's average hourly wage excludes certain costs that are not paid under the IPPS. In calculating a hospital's average salaries plus wage-related costs, we subtract from Line 1 (total salaries) the GME and CRNA costs reported on Lines 2, 4.01, 6, and 6.01, the Part B salaries reported on Lines 3, 5 and 5.01, home office salaries reported on Line 7, and exclude salaries reported on Lines 8 and 8.01 (that is, direct salaries attributable to SNF services, home health services, and other subprovider components not subject to the IPPS). We also subtract from Line 1 the salaries for which no hours were reported. To determine total salaries plus wage-related costs, we add to the net hospital salaries the costs of contract labor for direct patient care, certain top management, pharmacy, laboratory, and nonteaching physician Part A services (Lines 9 and 10), home office salaries and wage-related costs reported by the hospital on Lines 11 and 12, and nonexuded area wage-related costs (Lines 13, 14, and 18).

We note that contract labor and home office salaries for which no corresponding hours are reported are not included. In addition, wage-related costs for nonteaching physician Part A employees (Line 18) are excluded if no corresponding salaries are reported for those employees on Line 4.

Step 3—Hours—With the exception of wage-related costs, for which there are no associated hours, we compute total hours using the same methods as described for salaries in Step 2.

Step 4—For each hospital reporting both total overhead salaries and total overhead hours greater than zero, we then allocate overhead costs to areas of the hospital excluded from the wage index calculation. First, we determine the ratio of excluded area hours (sum of Lines 8 and 8.01 of Worksheet S-3, Part II) to revised total hours (Line 1 minus the sum of Part II, Lines 2, 3, 4.01, 5, 5.01, 6, 6.01, 7, and Part III, Line 13 of Worksheet S-3). We then compute the amounts of overhead salaries and hours to be allocated to excluded areas by multiplying the above ratio by the total overhead salaries and hours reported on Line 13 of Worksheet S-3, Part III. Next, we compute the amounts of overhead wage-related costs to be allocated to excluded areas using three steps: (1) We determine the ratio of overhead hours (Part III, Line 13) to revised hours (Line 1 minus the sum of Lines 2, 3, 4.01, 5, 5.01, 6, 6.01, 7, 8, and 8.01); (2) we compute overhead wage-related costs by multiplying the overhead hours ratio by wage-related costs reported on Part II, Lines 13, 14, and 18; and (3) we multiply the computed overhead wage-related costs by the above excluded area hours ratio. Finally, we subtract the computed overhead salaries, wage-related costs, and hours associated with excluded areas from the total salaries (plus wage-related costs) and hours derived in Steps 2 and 3.

Step 5—For each hospital, we adjust the total salaries plus wage-related costs to a common period to determine total adjusted salaries plus wage-related costs. To make the wage adjustment, we estimate the percentage change in the employment cost index (ECI) for compensation for each 30-day increment from October 14, 2002, through April 15, 2004, for private industry hospital workers from the BLS’ Compensation and Working Conditions. We use the ECI because it reflects the price increase associated with total compensation (salaries plus fringes) rather than just the increase in salaries. In addition, the ECI includes managers as well as other hospital workers. This methodology to compute the monthly update factors uses actual quarterly ECI data and assures that the update factors match the actual quarterly and annual percent changes. The factors used to adjust the hospital's data were based on the midpoint of the cost reporting period, as indicated below.

For example, the midpoint of a cost reporting period beginning January 1, 2003, and ending December 31, 2003, is June 30, 2003. An adjustment factor of 1.02948 would be applied to the wages of a hospital with such a cost reporting period. In addition, for the data for any cost reporting period that began in FY 2003 and covered a period of less than 360 days or more than 370 days, we annualize the data to reflect a 1-year cost report. Dividing the data by the number of days in the cost report and then multiplying the results by 365 accomplishes annualization.

Step 6—Each hospital is assigned to its appropriate urban or rural labor market area before any reclassifications under section 1886(d)(8)(B), section 1886(d)(8)(E), or section 1886(d)(10) of the Act. Within each urban or rural labor market area, we add the total adjusted salaries plus wage-related costs obtained in Step 5 for all hospitals in that area to determine the total adjusted salaries plus wage-related costs for the labor market area.

Step 7—We divide the total adjusted salaries plus wage-related costs obtained under both methods in Step 6 by the sum of the corresponding total hours (from Step 4) for all hospitals in each labor market area to determine an average hourly wage for the area.

Step 8—We add the total adjusted salaries plus wage-related costs obtained in Step 5 for all hospitals in the Nation and then divide the sum by the national sum of total hours from Step 4 to arrive at a national average hourly wage. Using the data as described above, the national average hourly wage is $29.6521.

Step 9—For each urban or rural labor market area, we calculate the hospital wage index value, unadjusted for occupational mix, by dividing the area average hourly wage obtained in Step 7 by the national average hourly wage computed in Step 8.

Step 10—Following the process set forth above, we develop a separate Puerto Rico-specific wage index for purposes of adjusting the Puerto Rico standardized amounts. (The national Puerto Rico standardized amount is adjusted by a wage index calculated for all Puerto Rico labor market areas based on the national average hourly wage as described above.) We add the total adjusted salaries plus wage-related costs (as calculated in Step 5) for all hospitals in Puerto Rico and divided the sum by the total hours for Puerto Rico (as calculated in Step 4) to arrive at an overall average hourly wage of $13.0915 for Puerto Rico. For each labor market area in Puerto Rico, we calculate the Puerto Rico-specific wage index value by dividing the area average hourly wage (as calculated in Step 7) by the overall Puerto Rico average hourly wage.

Step 11—Section 4410 of Pub. L. 105-33 provides that, for discharges on or after October 1, 1997, the area wage index applicable to any hospital that is located in an urban area of a State may not be less than the area wage index applicable to hospitals located in rural areas in that State. (For all-urban States, we establish an imputed floor (69 FR 49109). Furthermore, this wage index floor is to be implemented in such a manner as to ensure that aggregate IPPS payments are not greater or less than those that would have been made in the year if this section did not apply. For FY 2007, the areas affected by this provision, after the occupational mix adjustment is applied, will be by a footnote in Tables 4A-1 and 4A-2 that are to be published separate from this final rule.

G. Implementation of the FY 2007 Occupational Mix Adjustment to the Wage Index

For the final FY 2005 and FY 2006 wage indices, we used a blend of the occupational mix adjusted wage index and the unadjusted wage index. Specifically, we adjusted 10 percent of the FY 2005 and FY 2006 wage index adjustment factor by a factor reflecting occupational mix. We refer readers to the FY 2005 IPPS final rule at 69 FR 49052 and the FY 2006 IPPS final rule at 70 FR 47376 for a detailed discussion of the blended wage index.

As discussed in section III.C. of this preamble, for FY 2007, we are applying the occupational mix adjustment to 100 percent of the FY 2007 wage index. We will calculate the occupational mix adjustment using the first 3 months of the 2006 survey data, using the methodology described in section III.C. of this preamble.

Comment: One commenter suggested that, for the FY 2007 wage index, CMS should apply the Bellevue Hosp. Center v. Leavitt decision only to hospitals in the Second Circuit, and not on a nationwide basis. For States outside the Second Circuit, the commenter recommended that CMS apply the occupational mix adjustment at 10 percent, as it did in FYs 2005 and 2006. The commenter noted that there is a CMS (then HCFA) precedent for applying a court's order to only hospitals in the States in the Circuit where the decision was rendered, citing HCFA Ruling 97-2, pertaining to the inclusion of “eligible but unpaid” Medicaid days in the DSH calculation.

Response: The commenter did not address whether the 10-percent adjustment would use the new 2006 occupational mix survey data or the prior 2003 data. Therefore, it is not clear how the commenter is suggesting we apply the policy. Nevertheless, we believe the most appropriate policy is to apply the occupational mix adjustment uniformly nationwide, using the same survey data and a 100 percent adjustment for all hospitals. It is important to keep in mind that the occupational mix adjustment is an adjustment to the wage index factor that represents the ratio of a labor market area's average hourly wage to the national average hourly wage. DSH adjustments, in contrast, are not based upon individual hospital information compared to a national average. If we were to use separate sets of data depending upon geographic location, hospitals located in the Second Circuit would be compared to one national benchmark, whereas hospitals located elsewhere would be compared to a different one. We believe such a policy would undermine the calculation of the wage index that is a relative measure of differences in area wage levels that uses a uniform national baseline for purposes of comparison. In addition, we note that the New York labor market area includes counties located both inside and outside of the Second Circuit. The New York-White Plains-Wayne, NY-NJ CBSA includes three New Jersey Counties: Bergen, Hudson, and Passaic Counties. These counties are located in the Third Circuit, not the Second Circuit. Therefore, applying the Bellevue Hosp. Center v. Leavitt decision only in the Second Circuit would result in two area wage index values for the New York labor market area, adding further complexity to the wage index calculation.

Comment: One commenter believed that section 1886(d)(6) of the Act requires CMS to publish its actual wage tables and other factors by August 1. The commenter also cited the Balanced Budget Act of 1997 (BBA), Pub. L. 105-33, under which Congress moved the deadline in section 1886(d)(6) of the Act from September 1 to August 1. The commenter contended that Congress would not have needed to move the deadline if the final data were not to be published as of August 1.

Response: The relevant language of section 1886(d)(6)of the Act states: “The Secretary shall provide for publication in the Federal Register, on or before August 1 before each fiscal year * * * of a description of the methodology and data used in computing the adjusted DRG prospective payment rates under this subsection.” We believe the plain language of section 1886(d)(6) of the Act requires merely a description of the data and methodology that are used to compute the IPPS rates and does not require actual publication of the rates.

With respect to the comments about the statutory change that moved the deadline for the IPPS rule from September 1 to August 1, section 4644 of the BBA was an amendment to conform section 1886(h)(6) of the Act to the requirements of the Congressional Review Act. The Congressional Review Act does not allow a major rule to go into effect for 60-days unless there is an act of Congress allowing the rule to go into effect earlier. The publication date in section 1886(d)(6) of the Act was changed accordingly so that the IPPS final rule could take effect no sooner than 60 days after publication, or by the beginning of the Federal fiscal year on October 1 without Congress having to act. However, Congress did not alter section 1886(d)(6) of the Act with respect to the information that is to be included in the final rule. We agree with the commenter that it is our usual practice to publish the wage tables and other factors along with the final rule consistent with 42 CFR 412.8. However, due to our implementation of the Bellevue Hosp. Center v. Leavitt decision, it is not possible to follow this procedure for FY 2007. In the proposed rule, we explained our intent to post the FY 2007 occupational mix adjusted wage index tables and related impacts on the CMS Web site after we publish the FY 2007 IPPS final rule, and in advance of October 1, 2006 (71 FR 28652). We have modified 42 CFR 412.8 accordingly. The change we are making to § 412.8 is a procedural rule that we are making effective upon publication.

Comment: A few commenters expressed concern that the new occupational mix adjustment may have a negative impact on some hospitals, and they would not know how they are affected until the final FY 2007 wage index tables are published. Some commenters recommended that CMS allow hospitals more time to review their data, comment on the survey results, and make adjustments and/or revisions to their occupational mix survey data. One commenter requested that CMS publish the occupational mix regulations and data as an interim final rule with a full 60-day comment period so that providers will have an opportunity to comment further. Another commenter urged CMS to consider either delaying the implementation of the occupational mix adjustment, or consider allowing retroactive correction to any errors discovered after October 1. A few commenters recommended that CMS use its discretionary authority to “smooth out” the impact of this change on adversely affected hospitals and apply a multiyear transition of the occupational mix survey data.

Response: As we indicated above, while we understand the commenters’ concerns about the potential for inaccurate occupational mix survey data to be used due to the abbreviated data collection and reporting periods, we believe we have established a review and correction process that is intended to minimize errors. We cannot delay the implementation of, or transition in, the occupational mix adjustment for the FY 2007 wage index because the Second Circuit Court required that all “data collection and measurement and any other preparations necessary for full application should be complete by September 30, 2006, at which time we instruct the agency to immediately apply the adjustment in full.” Also, we believe that the 30-day comment period after the May 17, 2006 publication of the amended FY 2007 IPPS proposed rule provided ample opportunity for the public to comment on the new occupational mix survey data and adjustment for the FY 2007 wage index. Hospitals are usually afforded 60 days to comment on the entire IPPS rule. In addition, we cannot allow retroactive changes to the FY 2007 wage index for errors discovered after October 1, 2006, unless a hospital's correction request meets the strict criteria of § 412.64(k)(1) of our existing regulations (also see section III.J. of this preamble). However, as previously mentioned, we will allow hospitals an additional opportunity to revise both their 1st quarter and 2nd quarter 2006 occupational mix data for the FY 2008 wage index.

Comment: One commenter recommended that CMS publish the corrected 1st quarter 2006 survey data as a public use file prior to the publication of the final FY 2007 wage index tables.

Response: Intermediaries are required to transmit the corrected 1st quarter 2006 survey data to CMS by July 27, 2007. Unfortunately, due to our short timeframe after July 27 for reviewing the survey data and computing, analyzing, and publishing the final FY 2007 occupational mix adjusted wage index, we cannot publish the corrected 1st quarter survey data before we publish final FY 2007 wage index tables.

The final wage index values for FY 2007 (except those for hospitals receiving wage index adjustments under section 505 of Pub. L. 108-173) will be included in Tables 4A-1, 4A-2, 4B, 4C-1, 4C-2, and 4F that are to be posted on our Web site and published in a Federal Register notice subsequent to this final rule.

Tables 3A and 3B in the separate issuance will list the 3-year average hourly wage for each labor market area before the redesignation of hospitals, using the wages included in the calculation for the FYs 2005, 2006, 2007 wage indices. Table 3A in the separate issuance will list these data for urban areas and Table 3B in the separate issuance will list these data for rural areas. In addition, Table 2 in the separate issuance will include the adjusted average hourly wage for each hospital from the FY 2001 and FY 2002 cost reporting periods, as well as the FY 2003 period used to calculate the FY 2007 wage index. The 3-year averages will be calculated by dividing the sum of the dollars (adjusted to a common reporting period using the method described previously) across all 3 years, by the sum of the hours. If a hospital is missing data for any of the previous years, its average hourly wage for the 3-year period will be calculated based on the data available during that period.

The final wage index values in Tables 4A-1, 4A-2, 4B, 4C-1, 4C-2, and 4F and the average hourly wages in Tables 2, 3A, and 3B to be posted on our Web site and published in a subsequent Federal Register notice will include the occupational mix adjustment.

H. Revisions to the Wage Index Based on Hospital Redesignations

1. General

Under section 1886(d)(10) of the Act, the Medicare Geographic Classification Review Board (MGCRB) considers applications by hospitals for geographic reclassification for purposes of payment under the IPPS. Hospitals must apply to the MGCRB to reclassify by September 1 of the year preceding the year during which reclassification is sought. Generally, hospitals must be proximate to the labor market area to which they are seeking reclassification and must demonstrate characteristics similar to hospitals located in that area. The MGCRB issues its decisions by the end of February for reclassifications that become effective for the following fiscal year (beginning October 1). The regulations applicable to reclassifications by the MGCRB are located in §§ 412.230 through 412.280.

Section 1886(d)(10)(D)(v) of the Act provides that, beginning with FY 2001, a MGCRB decision on a hospital reclassification for purposes of the wage index is effective for 3 fiscal years, unless the hospital elects to terminate the reclassification. Section 1886(d)(10)(D)(vi) of the Act provides that the MGCRB must use the 3 most recent years' average hourly wage data in evaluating a hospital's reclassification application for FY 2003 and any succeeding fiscal year.

Section 304(b) of Pub. L. 106-554 provides that the Secretary must establish a mechanism under which a statewide entity may apply to have all of the geographic areas in the State treated as a single geographic area for purposes of computing and applying a single wage index, for reclassifications beginning in FY 2003. The implementing regulations for this provision are located at § 412.235.

Section 1886(d)(8)(B) of the Act requires the Secretary to treat a hospital located in a rural county adjacent to one or more urban areas as being located in the MSA to which the greatest number of workers in the county commute, if the rural county would otherwise be considered part of an urban area under the standards for designating MSAs and if the commuting rates used in determining outlying counties were determined on the basis of the aggregate number of resident workers who commute to (and, if applicable under the standards, from) the central county or counties of all contiguous MSAs. In light of the new CBSA definitions and the Census 2000 data that we implemented for FY 2005 (69 FR 49027), we undertook to identify those counties meeting these criteria. The eligible counties are identified under section III.H.4. of this preamble.

2. Effects of Reclassification/Redesignation

Section 1886(d)(8)(C) of the Act provides that the application of the wage index to redesignated hospitals is dependent on the hypothetical impact that the wage data from these hospitals would have on the wage index value for the area to which they have been redesignated. These requirements for determining the wage index values for redesignated hospitals is applicable both to the hospitals located in rural counties deemed urban under section 1886(d)(8)(B) of the Act and hospitals that were reclassified as a result of the MGCRB decisions under section 1886(d)(10) of the Act. In compliance with section 1886(d)(8)(C) of the Act, as well as with the rules CMS has established by regulation, the wage index values were determined by considering the following:

• If including the wage data for the redesignated hospitals would reduce the wage index value for the area to which the hospitals are redesignated by 1 percentage point or less, the area wage index value determined exclusive of the wage data for the redesignated hospitals applies to the redesignated hospitals.
• If including the wage data for the redesignated hospitals reduces the wage index value for the area to which the hospitals are redesignated by more than 1 percentage point, the area wage index determined inclusive of the wage data for the redesignated hospitals (the combined wage index value) applies to the redesignated hospitals.
• If including the wage data for the redesignated hospitals increases the wage index value for the urban area to which the hospitals are redesignated, both the area and the redesignated hospitals receive the combined wage index value. Otherwise, the hospitals located in the urban area receive a wage index excluding the wage data of hospitals redesignated into the area.
• The wage data for a reclassified urban hospital is included in both the wage index calculation of the area to which the hospital is reclassified (subject to the rules described above) and the wage index calculation of the urban area where the hospital is physically located.
• Rural areas whose wage index values would be reduced by excluding the wage data for hospitals that have been redesignated to another area continue to have their wage index values calculated as if no redesignation had occurred (otherwise, redesignated rural hospitals are excluded from the calculation of the rural wage index).
• The wage index value for a redesignated rural hospital cannot be reduced below the wage index value for the rural areas of the State in which the hospital is located.
• In cases where urban hospitals have reclassified to rural areas under 42 CFR 412.103, the urban hospital wage data are: (a) Included in the rural wage index calculation, unless doing so would reduce the rural wage index; and (b) included in the urban area where the hospital is physically located.

3. FY 2007 MGCRB Reclassifications

Under section 1886(d)(10) of the Act, the MGCRB considers applications by hospitals for geographic reclassification for purposes of payment under the IPPS. The specific procedures and rules that apply to the geographic reclassification process are outlined in § 412.230 through § 412.280.

In the FY 2007 IPPS proposed rule (71 FR 24377), we identified hospitals that have reclassifications effective in FY 2007. As specified in § 412.273, hospitals that have been reclassified by the MGCRB are permitted to withdraw an application for reclassification or terminate an existing 3-year reclassification for FY 2007. The request must be received by the MGCRB within 45 days of publication of the IPPS proposed rule.

However, as a result of our compliance with the Bellevue Hosp. Center v. Leavitt court decision, as discussed earlier, we will be recalculating wage indices using new occupational mix data and applying the occupational mix to 100 percent of the wage index. Wage tables in the IPPS proposed rule did not include the new survey data, nor did they adjust 100 percent for occupational mix. Thus, the data that hospitals might have used to make withdrawal or termination decisions are obsolete. The necessary data (including wage indices and out-migration adjustments) hospitals generally utilize in evaluating whether to withdraw or terminate a reclassification will not be available until after this IPPS final rule has been published. Therefore, in the May 17, 2006 proposed rule (71 FR 28650), in this limited circumstance, we suspended the 45-day deadline and have established the new procedure described below to withdraw from or terminate reclassifications for FY 2007. Some hospitals may have adhered to the established process and notified the MGCRB of their decision to withdraw or terminate a reclassification, in accordance with § 412.273, before publication of that proposed rule.

Because hospitals made these decisions based on information in the FY 2007 IPPS proposed rule that is now obsolete, in the May 17, 2006 proposed rule, we proposed that the MGCRB not act on these withdrawal or termination requests. Instead, we have applied the following procedures for withdrawal and termination determinations for all hospital reclassifications for FY 2007. We will make reclassification withdrawal and termination determinations based on what we perceive would be most advantageous to the hospital. We will use our best efforts to determine what would provide the hospital with the highest possible wage index. Specifically, we will choose among: section 508 reclassifications, section 1886(d)(10) reclassifications, section 505 out-migration adjustments, and certain other changes to the wage index (for example, the special exceptions policy explained in the FY 2005 IPPS rule (69 FR 49105) or Lugar status if we determine that it is in the hospital's best interest to waive the Lugar/section 1886(d)(8)(B) redesignation in order to receive the section 505 out-migration adjustment).

We also will make the final occupational mix adjusted wage indices and out-migration adjustments and our interim decisions on hospital reclassifications available to the public in the Federal Register and on the CMS Web site at: http://www.cms.hhs.gov/AcuteInpatientPPS/WIFN/list.asp after August 1, 2006, and before October 1, 2006. We will allow hospitals a 30-day period from the date the final data and our interim decisions are made available on the Web site to notify CMS in writing, with a copy to the MGCRB, of whether they wish to reverse the reclassification decision made by CMS or to choose another reclassification for which they are eligible. We will make every effort to provide the final data before September 1, 2006, so that the 30-day period to make these determinations will end before October 1, 2006, and no retroactive adjustments will be necessary. Requests to reverse a decision made by CMS must be received, in writing, no later 30 days after the data are made available on the CMS Web site at the following:

Division of Acute Care, C4-08-06, 7500 Security Boulevard, Baltimore, MD 21244, Attn: Marianne Myers;

AND a copy to

Medicare Geographic Classification Review Board, 2520 Lord Baltimore Drive, Suite L, Baltimore, MD 21244-2670.

Prior to FY 2004, hospitals had been able to apply to be reclassified for purposes of either the wage index or the standardized amount. Section 401 of Pub. L. 108-173 established that all hospitals will be paid on the basis of the large urban standardized amount, beginning with FY 2004. Consequently, all hospitals are paid on the basis of the same standardized amount, which made such reclassifications moot. Although there could still be some benefit in terms of payments for some hospitals under the DSH payment adjustment for operating IPPS, section 402 of Pub. L. 108-173 equalized DSH payment adjustments for rural and urban hospitals, with the exception that the rural DSH adjustment is capped at 12 percent (except that rural referral centers and, effective for discharges occurring on or after October 1, 2006, MDHs have no cap). (A detailed discussion of this application appears in section IV.I. of the preamble of the FY 2005 IPPS final rule (69 FR 49085). The exclusion of MDHs from the 12 percent DSH cap under Pub. L. 109-171 is discussed under section IV.F.4. of this preamble.)

Comment: Several commenters asked CMS to clarify its position on withdrawing reclassifications as well as the timeframe of submitting applications for geographic reclassification.

Response: The normal timetable of 45 days after the publication of the proposed rule for hospitals to withdraw or terminate a reclassification under section 508 of Pub. L. 108-173, section 1886(d)(10) of the Act, or section 1886(d)(8)(B) of the Act (in order to receive a section 505 out-migration adjustment) does not apply for FY 2007. For this reason, any withdrawal or termination requests submitted to the MGCRB and/or CMS following publication of the FY 2007 IPPS proposed rule are not reflected in the reclassification tables shown in this final rule.

We will make best efforts to give each hospital the highest FY 2007 wage index after reviewing applicable data using the 100 percent occupational adjusted wage index. Hospitals will have 15 days from the display date of this final rule to notify us of whether, in the absence of viewing the final 100 percent occupational mix-adjusted wage index data, they wish to choose a particular wage index for which they are eligible (such as to definitively maintain a reclassification that they received or to definitively terminate or withdraw from a reclassification). Written requests to maintain, terminate, or withdraw a reclassification, in the absence of viewing the final wage tables, must be received at the following address no later than 5 p.m. EDT 15 days from the date this final rule appears on public display at the Office of the Federal Register:

Division of Acute Care, C4-08-06, 7500 Security Boulevard, Baltimore, MD 21244, Attn: Marianne Myers.

If we do not receive notice from the hospital within such 15-day timeframe, we will make determinations for the hospital using our best efforts to determine what we believe results in the highest wage index for the hospital. If applicable, we will give the hospital its home wage index with the out-migration adjustment, if that option results in the highest wage index. In some cases, we may determine that it is most advantageous for a hospital to terminate its Lugar/section 1886(d)(8)(B) reclassification in order to receive the out-migration adjustment. Because this termination would result in the hospital losing urban status, we will separately publish a table identifying these hospitals that move from Lugar/urban status to rural status with the out-migration adjustment. For section 508 hospital individual reclassifications, we may make half-year terminations/withdrawals on behalf of hospitals, using the procedures identified in our proposed rule. That is, for a section 508 hospital that applied for an individual reclassification under section 1886(d)(10), we would give the section 508 hospital the higher of its home wage index, section 508 or 1886(d)(10) wage index for the first half of the year. For the second half of the year, we would give the section 508 hospital the higher of its home wage index or its section 1886(d)(10) reclassification. (However, in no case could such a hospital receive its home wage indes for the first half of the year and its MGCRB reclassification for the second half, or vice versa. For group reclassifications, we will apply the higher of the home wage index or the section 1886(d)(10) reclassification for the entire year. For group reclassifications that include a section 508 hospital, we will apply the decision that was on the MGCRB application for groups that followed the procedural rules (that is, the group either: (1) Withdrew from its section 1886(d)(10) reclassification for the first half of FY 2007 and will only receive a second half FY 2007 section 1886(d)(10) reclassification; or (2) the group is reclassified under section 1886(d)(10) of the Act for the entire year and the section 508 hospital withdraws from its section 508 reclassification for the first half of the FY 2007) unless the group informs us differently after publication of the final occupational mix adjusted wage indices. Groups that include a section 508 hospital will be able to make decisions as a group, separately for the first and second half of the year. Thus, the group may decide to withdraw a section 1886(d)(10) reclassification that would be applicable only for the second half of FY 2007. Again, however, in no case could a group whose 508 hospital chose to waive its 508 reclassification (and therefore accept the MGCRB reclassification for the first half of FY 2007) withdraw its MGCRB reclassification for the first half of the year, but not the second (or vice versa).

We acknowledge that hospitals may base withdrawal/termination decisions on factors other than simply what results in the highest wage index for the upcoming fiscal year. For this reason, we will allow a hospital to change a decision that is made by CMS on its behalf. Hospitals should note that we will not recalculate the wage indices or budget neutrality factors after the final notice announcing the FY 2007 occupational mix adjusted wage indices. That is, we will not further recalculate the wage indices or standardized amounts based on hospital decisions that further revise decisions made by CMS on the hospitals' behalf.

We will post the final occupational mix adjusted wage indices, out-migration adjustments, and our interim decisions on hospital reclassification on the CMS Web site, as discussed above, sometime after August 1, 2006, and before October 1, 2006. We will post the same tables on the CMS Web site that appear in the Federal Register final notice of the occupational mix adjusted wage indices to be published after August 1, 2006 and before October 1, 2006. Hospitals will be able to determine the reclassification decision applied on their behalf by reviewing Tables 9A through 9C for hospitals that are reclassified under section 1886(d)(8)(B) of the Act, section 508 of Pub. L. 108-173, or section 1886(d)(10) of the Act. The applicable wage index for these hospitals will be found on Table 2. If a hospital is not listed in Tables 9A through 9C, CMS will have made a decision not to reclassify the hospital and its home wage index will apply, including the effect of the out-migration adjustment, will be found in Table 2. The applicable out-migration adjustment for the hospital will be found in Table 4J. As indicated above, we will separately publish a table identifying hospitals that we move from Lugar/urban status to rural status with the out-migration adjustment in Table 9D. Hospitals will have 30 days after the data are placed on the CMS Web site to submit, in writing, whether they wish to revise the decision made on their behalf by CMS. Written requests to revise a decision made on behalf of a hospital by CMS must be received by CMS no later than 5 p.m. EDT, with a copy sent to the MGCRB, within 30 days from the date the information appears on the CMS Web site at http://www.cms.hhs.gov/AcuteInpatientPPS/WIFN/list.asp at the following addresses:

Division of Acute Care, C4-08-06, 7500 Security Boulevard, Baltimore, MD 21244, Attn: Marianne Myers;

AND a copy to

Medicare Geographic Classification Review Board, 2520 Lord Baltimore Drive, Suite L, Baltimore, MD 21244-2670.

If a hospital fails to notify CMS that it is revising a determination made on its behalf within 30 days from the date the information appears on the CMS Web site, the interim decision made by CMS on the hospital's behalf will be final for FY 2007. Therefore, if CMS makes a decision on a hospital's behalf to terminate or withdraw a reclassification and the hospital does not reverse or modify CMS's decision, we will deem the hospital's reclassification is withdrawn or terminated. Once CMS's decision on the hospital's behalf is in effect, it will be treated in the same manner as if the hospital(s) had made the reclassification decision on its own. Thus, for example, because a hospital cannot have overlapping reclassifications, if we decide a hospital should accept a FY 2007 through 2009 reclassification, any reclassification the hospital previously had for FY 2006 through 2008 would be permanently terminated.

Section 1886(d)(10)(C)(ii) of the Act indicates that a hospital requesting a change in geographic classification for a FY must submit its application to the MGCRB no later than the first day of the 13-month period ending on September 30 of the preceding fiscal year. Thus, the statute requires that FY 2008 reclassification applications be submitted to the MGCRB no later than September 1, 2006. Hospitals must submit applications for geographic reclassification for FY 2008 by September 1, 2006. However, because the 3-year average hourly wage of hospitals for the FY 2007 final rule will not be available by the September 1, 2006 deadline for submitting FY 2008 geographic reclassification applications, we will allow hospitals to supplement incomplete reclassification applications with the official data used to develop the FY 2007 wage index after filing their initial application. As indicated above, the 3-year average hourly wage information that will be necessary for FY 2008 reclassification applications will be available subsequent to this final rule after August 1, and before October 1, 2006. The information will be available on the CMS Web site at: http://www.cms.hhs.gov/AcuteInpatientPPS/WIFN/list.asp and then accessing the page titled “MGCRB Reclassification Data for FY 2008 Applications.” Applications and other information about MGCRB reclassifications may be obtained via the CMS Internet Web site at: http://www.cms.hhs.gov/mgcrb/, or by calling the MGCRB at (410) 786-1174. The mailing address of the MGCRB is: 2520 Lord Baltimore Drive, Suite L, Baltimore, MD 21244-2670.

Comment: Several commenters requested a revision in the geographic reclassification rules so that in the future the occupational mix adjusted average hourly wage data is used as a point of comparison for eligibility. The commenters believed this change would make reclassification decisions consistent with the new basis for the wage index. The commenters also suggested that hospitals should not be allowed to apply for reclassification if they do not provide complete occupational mix data.

Response: Section 1886(d)(10)(D)(vi) of the Act requires the MGCRB to use the 3-year average of the average hourly wage data from the most recently published hospital wage survey data, as well as the preceding 2 fiscal years’ published surveys. Because our published surveys of wage data include adjustments for occupational mix (10 percent in FYs 2005 and 2006 and 100 percent in FY 2007), the MGCRB uses mix-adjusted wage indices in making reclassification decisions. Therefore, for FY 2008 reclassification applications, the MGCRB will use the average of the average hourly wages for FYs 2005 through 2007. These data will be based on an occupational mix adjustment of 10 percent for FY 2005 and FY 2006 and 100 percent for FY 2007.

With respect to the comment about precluding hospitals that did not submit occupational mix survey data from reclassifying, we believe that due to the unusual circumstances of the Court's order and the short timeframe that hospitals were provided for completing and submitting their data, it would not be fair to apply a penalty to non-responsive hospitals for the 2008 reclassification applications. However, as indicated earlier, we will give serious consideration to applying some sort of penalty in the future if a hospital does not comply with regulations requiring submission of occupational mix survey data.

4. Procedures for Hospitals Applying for Reclassification Effective in FY 2008 and Reinstating Reclassifications in FY 2008

Applications for FY 2008 reclassifications are due to the MGCRB by September 1, 2006. We note that this deadline also applies for canceling a previous wage index reclassification withdrawal or termination under § 412.273(d). As we noted in the FY 2007 IPPS proposed rule (71 FR 24083), applications and other information about MGCRB reclassifications may be obtained, beginning in mid-July 2006, on the CMS Web site at: http://www.cms.hhs.gov/mgcrb/, or by calling the MGCRB at (410) 786-1174.

The MGCRB, in evaluating a hospital's request for reclassification for FY 2008 for the wage index, must utilize the official data used to develop the FY 2007 wage index. The wage data used to support the hospital's wage comparisons must be from the CMS hospital wage survey. Generally, the source for these data is the IPPS final rule to be published on or before August 1, 2006. However, as we stated earlier, the wage tables identifying the 3-year average hourly wage of hospitals will not be available in time to include them in this FY 2007 IPPS final rule. Therefore, we will make the data available subsequent to August 1, 2006, but before October 1, 2006.

Section 1886(d)(10)(C)(ii) of the Act indicates that a hospital requesting a change in geographic classification for a FY must submit its application to the MGCRB not later than the first day of the 13-month period ending on September 30 of the preceding FY. Thus, the statute requires that FY 2008 reclassification applications be submitted to the MGCRB by no later than September 1, 2006. For this reason, hospitals must file an FY 2008 reclassification application by the September 1, 2006 deadline even though the average hourly wage data used to develop the final FY 2007 wage indices will not yet be available. We note that, under § 412.256(c), the MGCRB must review applications and notify the hospital if it determines that the application is incomplete. We are also allowing hospitals 30 days from the date the final wage data is posted on the CMS Web site to request to cancel a withdrawal or termination in order to reinstate its reclassification for FY 2008 or FY 2009, or both fiscal years. Requests to cancel a withdrawal or termination in order to reinstate a hospital's reclassification for FY 2008 or FY 2009, or both fiscal years, should be forwarded to the following addresses:

Medicare Geographic Classification Review Board, 2520 Lord Baltimore Drive, Suite L, Baltimore, MD 21244-2670;

AND a copy to

Division of Acute Care, C4-08-06, 7500 Security Boulevard, Baltimore, MD 21244, Attn: Marianne Myers.

As outlined in § 412.256(c)(2), hospitals with incomplete applications have the opportunity to request that the MGCRB grant a hospital that has submitted an application by September 1, 2006, an extension beyond September 1, 2006, to complete its application. Thus, while hospitals must file an application for reclassification to the MGCRB by September 1, 2006, they will be able to supplement the reclassification application with official data used to develop the FY 2007 wage index after filing their initial application. We are providing that hospitals file a supplement to the reclassification application with official data used to develop the FY 2007 wage index no later than 30 days after the data are made available on the CMS Web site. These same rules will apply to canceling a withdrawal or termination of a geographic reclassification.

5. FY 2007 Redesignations Under Section 1886(d)(8)(B) of the Act

Beginning October 1, 1988, section 1886(d)(8)(B) of the Act required us to treat a hospital located in a rural county adjacent to one or more urban areas as being located in the MSA if certain criteria were met. Prior to FY 2005, the rule was that a rural county adjacent to one or more urban areas would be treated as being located in the MSA to which the greatest number of workers in the county commute, if the rural county would otherwise be considered part of an urban area under the standards published in the Federal Register on January 3, 1980 (45 FR 956) for designating MSAs (and New England County Metropolitan Areas (NECMAs)), and if the commuting rates used in determining outlying counties (or, for New England, similar recognized areas) were determined on the basis of the aggregate number of resident workers who commute to (and, if applicable under the standards, from) the central county or counties of all contiguous MSAs (or NECMAs). Hospitals that met the criteria using the January 3, 1980 version of these OMB standards were deemed urban for purposes of the standardized amounts and for purposes of assigning the wage data index.

Effective beginning FY 2005, we use OMB's 2000 CBSA standards and the Census 2000 data to identify counties qualifying for redesignation under section 1886(d)(8)(B) for the purpose of assigning the wage index to the urban area. We provided the chart below with the listing of the rural counties designated as urban under section 1886(d)(8)(B) of the Act in the FY 2007 IPPS proposed rule. For discharges occurring on or after October 1, 2006, hospitals located in the first column of this chart will be redesignated for purposes of using the wage index of the urban area listed in the second column.

The following table is subject to revision if CMS decides it is most advantageous for a county to waive its county Lugar status in order for a hospital within that county to receive a section 505 out-migration adjustment.