Medicare Inpatient Hospital Payments: CMS Has Used External Data 
for New Technologies in Certain Instances and Medicare Remains	 
Primary Data Source (26-SEP-07, GAO-07-46).			 
                                                                 
Under Medicare, hospitals generally receive fixed payments for	 
inpatient stays based on diagnosis-related groups (DRG), a system
that classifies stays by patient diagnoses and procedures. The	 
Centers for Medicare & Medicaid Services (CMS) annually uses its 
own data to reclassify DRGs. CMS also makes add-on payments for  
stays involving new technologies that meet three eligibility	 
criteria. Stakeholders may submit data that are external to CMS  
as part of a DRG reclassification request or an add-on payment	 
application. The Medicare Prescription Drug, Improvement, and	 
Modernization Act of 2003 required GAO to examine whether CMS	 
could improve its use of external data, including using data	 
collected by other government agencies for DRG payments. As	 
discussed with the committees of jurisdiction, GAO examined (1)  
to what extent CMS has used external data in determining payments
for inpatient stays involving new technologies, and (2) to what  
extent can external data from other government agencies be used  
by CMS in determining DRG payments for inpatient stays involving 
new technologies. GAO interviewed officials from CMS and industry
stakeholders. GAO interviewed officials from Bureau of Labor	 
Statistics (BLS), Department of Veterans Affairs (VA), Department
of Defense (DOD), and Agency for Healthcare Research and Quality 
(AHRQ) because these agencies may have data useful to CMS. GAO	 
also reviewed regulations and other CMS materials.		 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-07-46						        
    ACCNO:   A76730						        
  TITLE:     Medicare Inpatient Hospital Payments: CMS Has Used       
External Data for New Technologies in Certain Instances and	 
Medicare Remains Primary Data Source				 
     DATE:   09/26/2007 
  SUBJECT:   Data collection					 
	     Data integrity					 
	     Diagnosis related groups				 
	     Eligibility determinations 			 
	     Health care cost control				 
	     Health care costs					 
	     Hospital care services				 
	     Medical care evaluation				 
	     Medical expense claims				 
	     Medical records					 
	     Medical technology 				 
	     Medicare						 
	     Medicare Provider Analysis and Review		 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-07-46

   

     * [1]Results in Brief
     * [2]Background

          * [3]CMS Annually Revises DRG Classifications and Payment Weights
          * [4]DRG Classifications and Payment Weights for Any Given Fiscal
          * [5]Add-on Payments for New Technologies Can Supplement DRG Paym

     * [6]CMS Has Used External Data to Inform DRG Reclassification an

          * [7]CMS Has Used External Data to Inform the DRG Reclassificatio
          * [8]CMS Has Used External Data to Evaluate Applications for New

     * [9]Data from Other Government Agencies Have Limitations for Set
     * [10]Concluding Observations
     * [11]Agency and Other External Comments
     * [12]Appendix I: Comments from the Centers for Medicare & Medicai
     * [13]Appendix II: Comments from the Department of Defense
     * [14]Appendix III: GAO Contact and Staff Acknowledgments

          * [15]GAO Contact
          * [16]Acknowledgments

               * [17]Order by Mail or Phone

Report to Congressional Committees

United States Government Accountability Office

GAO

September 2007

MEDICARE INPATIENT HOSPITAL PAYMENTS

CMS Has Used External Data for New Technologies in Certain Instances and
Medicare Remains Primary Data Source

GAO-07-46

Contents

Letter 1

Results in Brief 5
Background 6
CMS Has Used External Data to Inform DRG Reclassification and to Evaluate
New Technology Add-on Payment Applications 12
Data from Other Government Agencies Have Limitations for Setting DRG
Payments for Inpatient Stays Involving New Technologies 17
Concluding Observations 20
Agency and Other External Comments 20
Appendix I Comments from the Centers for Medicare & Medicaid Services 22
Appendix II Comments from the Department of Defense 23
Appendix III GAO Contact and Staff Acknowledgments 24

Table

Table 1: Three Hypothetical Scenarios in Which CMS, during Fiscal Year
2007, Could Determine Eligibility under the Cost Criterion for Fiscal Year
2008 Add-on Payments 15

Figure

Figure 1: Process by Which CMS Used the MEDPAR File to Propose and
Finalize DRGs for Fiscal Year 2007 DRG Payments 10

Abbreviations

AAMC Association of American Medical Colleges
AHA American Hospital Association
AHRQ Agency for Healthcare Research and Quality
BBA Balanced Budget Act of 1997
BIO Biotechnology Industry Organization
BIPA Medicare, Medicaid, and SCHIP Benefits Improvement and Protection
  Act of 2000
BLS Bureau of Labor Statistics
CMS Centers for Medicare & Medicaid Services
CPI Consumer Price Index
DOD Department of Defense
DOL Department of Labor
DRG diagnosis-related groups
FDA Food and Drug Administration
FSS federal supply schedule
HCUP Healthcare Cost and Utilization Project
HHS Department of Health and Human Services
IPPS inpatient prospective payment system
MEDPAR Medicare Provider Analysis and Review
MFC Most-Favored Customer
MMA Medicare Prescription Drug, Improvement, and Modernization
  Act of 2003
MTF Military Treatment Facility
PPI Producer Price Index
TMA TRICARE Management Activity
VA Department of Veterans Affairs
X STOP X STOP Interspinous Process Decompression System

This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. However, because this
work may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this material
separately.

United States Government Accountability Office
Washington, DC 20548

September 26, 2007

The Honorable Max Baucus
Chairman
The Honorable Charles E. Grassley
Ranking Member
Committee on Finance
United States Senate

The Honorable Charles B. Rangel
Chairman
The Honorable Jim McCrery
Ranking
Member Committee on Ways and Means House of Representatives

At $119.4 billion, spending for hospital inpatient services accounted for
more than a third of total Medicare spending in fiscal year 2005. Most of
these dollars were spent on care provided to Medicare beneficiaries by the
approximately 4,000 acute care hospitals that bill Medicare under its
inpatient prospective payment system (IPPS). Under the IPPS, a hospital
generally receives a fixed, prospectively determined payment amount for
each inpatient stay.^1 Paying prospectively encourages hospitals to
operate efficiently, as they retain the difference if the payment for the
inpatient stay exceeds the hospital's cost of providing the stay.

Medicare law requires that IPPS payments account for variation in the
costs of providing different types of inpatient stays.^2 Consequently, the
Centers for Medicare & Medicaid Services (CMS), the agency within the
Department of Health and Human Services (HHS) that administers Medicare,
classifies inpatient stays using a system of diagnosis-related groups
(DRG). The number of DRGs changes from year to year. For fiscal year 2007,
each inpatient stay billed to Medicare is assigned to one of 538 DRGs
based on patient diagnosis and procedures performed. Inpatient stays
assigned to the same DRG are expected to have clinical and cost
similarities. Hospitals are paid for an inpatient stay based on the
assigned DRG's weight, which reflects the relative costliness of all
inpatient stays assigned to that DRG compared to inpatient stays assigned
to all DRGs.^3 DRG classifications and weights are the basis for DRG
payments to hospitals. In addition to the DRG-based payment, hospitals may
receive a supplemental payment, known as an outlier payment, if the costs
of the inpatient stay substantially exceed the DRG-based payment for that
stay.

^1Throughout this report, we use "stay" to represent a patient's
hospitalization.

^242 U.S.C. S 1395ww(d)(4)(C) (2000).

A major challenge for the IPPS is to maintain a system of DRGs that
accounts for the use of new technologies, which can substantially change
the costs hospitals incur in providing different types of inpatient
stays.^4 For example, a new technology may be clinically advantageous yet
so expensive that a hospital's overall cost increases substantially when
it provides the technology as part of an inpatient stay. In contrast, the
use of an alternative new technology may decrease the overall cost of an
inpatient stay--even if the technology is expensive--because it can reduce
complications and the length of time patients spend in the hospital.
Hospitals consider a range of factors--in addition to payment--before they
adopt a new technology, including the extent of its clinical benefit or
the needs of their patient populations.

To address changes in the cost of inpatient care, including the use of new
technologies, CMS annually revises the DRGs using data that are "internal"
to the Medicare program--that is, inpatient claims, which are bills
hospitals submit to CMS for inpatient services rendered to Medicare
beneficiaries.^5 CMS compiles data from these inpatient claims into an
electronic file, known as the Medicare Provider Analysis and Review
(MEDPAR) file. The MEDPAR file includes all charges from inpatient claims
for inpatient services provided to all Medicare beneficiaries across all
hospitals paid under the IPPS. In a process known as DRG reclassification,
CMS uses the data from the MEDPAR file to revise the assignment of
diagnoses and procedures to particular DRGs to ensure that each DRG
continues to represent inpatient stays with cost and clinical
similarities. Once inpatient stays assigned to various DRGs have been
reclassified, CMS calculates the payment weights for all the DRGs, so that
each weight reflects the average expected costliness of inpatient stays
that will be assigned to that DRG relative to inpatient stays that will be
assigned to all DRGs. Because CMS does not have a direct measure of the
actual cost of inpatient stays, it has relied on a proxy measure--the
amount hospitals charge Medicare on claims. Until October 1, 2006, CMS had
used the average charges of all inpatient stays assigned to each DRG to
represent the relative costliness of inpatient stays in that DRG compared
to the average charge for all inpatient stays. Effective October 1, 2006,
CMS uses the average estimated cost per inpatient stay to measure the
relative costliness of each DRG.

^3Under the IPPS, payment also includes adjustments for geographic
variation in hospital wages, indirect expenses related to medical
education, a hospital's caseload of low-income patients, and other
factors.

^4For purposes of this report, "new technology" is defined as a new
medical device, drug, or procedure.

^5CMS also uses hospital cost reports, which are submitted annually to
CMS. Cost reports contain each hospital's aggregate information on charges
for services and the actual costs of providing those services to all
patients, as well as information on total charges and estimates of costs
for services provided to Medicare beneficiaries.

The DRG classifications and weights are based on data from inpatient
claims for inpatient services provided 2 fiscal years prior.^6 As a
result, certain DRG weights do not reflect the cost of the most recent
technologies, or those adopted by hospitals in the previous fiscal year.
Manufacturers of new technologies have raised concerns that CMS's reliance
on data from the MEDPAR file to annually revise the DRG classifications
and weights may result in inadequate payments to hospitals for inpatient
stays involving the new technologies. Furthermore, they have raised
concerns that inadequate payments could jeopardize beneficiary access to
these technologies.

To address concerns about the timeliness of data from the MEDPAR file with
respect to new technologies, the conference report for the Balanced Budget
Act of 1997 (BBA)^7 directed CMS to consider using data that are
"external" to the Medicare program--for example, claims data from other
payers--when it reclassifies and weights the DRGs,^8 to the extent that
doing so is "feasible" and the data are "reliable" and "validated."^9 In
response to the BBA conference report, CMS instituted a policy whereby
manufacturers and other stakeholders--for example, hospitals--could submit
external data to CMS to help demonstrate that inpatient stays involving a
particular new technology are costlier on average than other inpatient
stays assigned to the same DRG, and should be assigned to a DRG with a
higher payment weight. The Medicare, Medicaid, and SCHIP Benefits
Improvement and Protection Act of 2000 (BIPA) required CMS to make
additional payments for inpatient stays that, because they involve new
technologies that may increase overall costs, would be inadequately paid
under DRG-based payments alone.^10 In response to the BIPA provision, CMS
implemented "add-on payments" to hospitals for certain expensive new
technologies that meet three eligibility criteria for being new, costly,
and a substantial clinical improvement over existing technologies. CMS
stated that it could use data from external sources to identify
technologies that are appropriate for these add-on payments, because those
technologies would not be represented in inpatient claims. CMS projected
that it would spend approximately $32 million in fiscal year
2007--approximately 0.03 percent of total IPPS spending--on add-on
payments.^11

^6For example, payments for fiscal year 2007 are based on data from claims
for services provided in fiscal year 2005.

^7Pub. L. No. 105-33, 111 Stat. 251.

^8External data can also include, but are not limited to, manufacturer
invoices and data from hospital data vendors.

^9H. R. Rep. No. 105-217, at 734 (1997) (Conf. Rep.). External data are
not defined in statute. We define external data as data that are not
collected from hospitals by CMS.

Manufacturers of new technologies have stated that CMS has not
sufficiently used external cost data they may provide on behalf of their
products when it reclassifies and weights DRGs or evaluates new
technologies for the add-on payments. The Medicare Prescription Drug,
Improvement, and Modernization Act of 2003 (MMA), required us to examine
whether CMS could improve its use of external data under the IPPS,
including whether data collected by other government agencies would be
best suited for CMS to use in setting payments for inpatient stays.^12 As
discussed with the committees of jurisdiction, we examined (1) to what
extent CMS has used external data in determining payments for inpatient
stays involving new technologies, and (2) to what extent external data
from other government agencies can be used by CMS in determining DRG
payments for inpatient stays involving new technologies.

To examine to what extent CMS has used external data in determining
payments for inpatient stays involving new technologies, we reviewed IPPS
regulations and other CMS materials and interviewed officials from CMS and
HHS.^13 We interviewed representatives from two hospital associations--the
American Hospital Association and the Association of American Medical
Colleges--and two associations of technology manufacturers--the Advanced
Medical Technology Association and the Biotechnology Industry
Organization. Additionally, we interviewed officials from four individual
drug and device manufacturers and one hospital that we identified as
having submitted external data to CMS.

^10Pub. L. No. 106-554, app. F, sec. 533(b), S 1886(d)(5)(K)(ii)(III), 114
Stat. 2763, 2763A-548 (codified at 42 U.S.C. S 1395ww(d)(5)(K)(ii)(III)
(2000)).

^1171 Fed. Reg. 47,870, 48,344 (Aug. 18, 2006).

^12Pub. L. No. 108-173, S 942(c), 117 Stat. 2066, 2422.

To examine to what extent external data from other government agencies can
be used by CMS in determining DRG payments for inpatient stays involving
new technologies, we reviewed IPPS regulations and other CMS materials and
interviewed officials from CMS and HHS. In addition, we interviewed
officials from the Bureau of Labor Statistics (BLS) in the Department of
Labor (DOL) because BLS collects price information for new technologies.
We also interviewed officials from the Department of Veterans Affairs
(VA), the Department of Defense (DOD), and the Agency for Healthcare
Research and Quality (AHRQ) in HHS because we identified these agencies as
having inpatient stay information, including cost or charge data, that
they collect for their own purposes that could be useful to Medicare. We
conducted our work from December 2004 through January 2006 and from August
2006 through August 2007 in accordance with generally accepted government
auditing standards.

Results in Brief

CMS has used external data for two purposes: to inform DRG
reclassification and to evaluate new technology add-on payment
applications. To inform DRG reclassification, CMS accepts the submission
of external data that are intended to demonstrate that inpatient stays
involving a new technology are costlier on average than the other
inpatient stays in the same DRG. CMS uses data from the MEDPAR file to
validate the external data submitted. Specifically, when external data are
submitted for a proposed DRG reclassification for a procedure or new
technology, CMS's policy is to find the same or similar evidence in the
MEDPAR file. Generally, CMS will not make a reclassification decision for
a DRG involving a new technology if the technology is so new that it does
not appear in the MEDPAR file. To evaluate new technology add-on payment
applications, CMS has generally used external data in conjunction with
data from the MEDPAR file to evaluate whether a new technology meets one
of the three eligibility criteria, specifically the criterion related to
cost.

^13We did not examine or evaluate the add-on payment application process,
CMS's policy for accepting external data, or the adequacy of the add-on
and outlier payments made to hospitals.

Data from other government agencies have limitations for CMS's setting DRG
payments for inpatient stays involving new technologies. This is because,
when setting DRG payments, CMS generally needs data that are
representative of the Medicare population, timely, and complete in that
the data include the total charge or other measure of costliness for all
services provided during an inpatient stay, including new technologies.
The data we identified from BLS, VA, DOD, and AHRQ were either not
representative of the Medicare population, were no timelier than data from
the MEDPAR file, or were not complete.

Data from the MEDPAR file remain the primary data source for setting DRG
payments because they include all charges from inpatient claims for
inpatient services provided to all Medicare beneficiaries across all
hospitals paid under the IPPS. CMS needs these data to determine payment
for each DRG relative to other DRGs. In instances where data from the
MEDPAR file have lacked charge information for certain stays involving new
technologies, CMS has used external data to inform the DRG
reclassification process and to evaluate new technology add-on payment
applications. To set DRG payments, CMS needs data that meet criteria of
being representative, timely, and complete. Although BLS, VA, DOD, and
AHRQ collect data for their own purposes that could potentially be useful
to CMS, these data are limited in their utility to set DRG payments
because they do not always meet CMS's criteria.

In commenting on a draft of this report, CMS stated that it agreed with
our findings and reiterated its commitment to using external data when
appropriate. DOD offered no comments on the draft of this report. VA
agreed with the facts as they pertain to the department. We also sent a
draft to DOL. DOL did not provide comments. Industry association reviewers
agreed with our findings but said that we should have discussed CMS's use
of data from sources other than the federal government. As we discussed in
the draft report, CMS has used external data from sources other than the
federal government including manufacturer data to inform DRG
reclassification and evaluate new technology add-on applications.

Background

Under the IPPS, hospitals are not paid separately for each item or service
they provide. Rather, payment is based on the DRG to which the entire
inpatient stay is assigned. Each of the 538 DRGs has a classification,
that is, an assigned combination of any of the approximately 17,000
diagnoses and procedures codes. When codes for these diagnoses and/or
procedures appear together on a claim,^14 the inpatient stay is assigned
to the appropriate DRG and paid accordingly.^15 In addition, CMS
determines if the inpatient stay is eligible for an outlier payment beyond
the DRG payment.^16 Hospitals can receive outlier payments for individual
inpatient stays determined to be extremely costly if a hospital can
demonstrate that the estimated cost of an individual inpatient stay
exceeds a cost threshold established by CMS.

CMS Annually Revises DRG Classifications and Payment Weights

Medicare law requires CMS to revise the DRG classifications and payment
weights at least annually to reflect changes in treatment patterns, new
medical services and technologies, and other factors that may change the
relative costliness of an inpatient stay.^17 To accomplish this, CMS
assembles a MEDPAR file from inpatient claims for a fiscal year, so that
the file contains one record for each inpatient stay provided during that
year. A MEDPAR record includes the admission and discharge dates, patient
and hospital identifiers, and codes that identify the diagnosis and the
procedures delivered during the inpatient stay. The record also contains
the hospital's total charge for the inpatient stay. The total charge
represents the charges for all services--including any new technology,
drugs, or supplies--provided during the inpatient stay. The total payment
to the hospital is also included in the MEDPAR record. MEDPAR records do
not indicate the hospital's actual cost for the inpatient stay or the cost
of individual procedures, which are not recorded on claims by hospitals.

^14Hospitals bill procedures and diagnoses with codes provided by The
International Classification of Diseases, 9th Revision, Clinical
Modification Hyattsville, Md.: Centers for Disease Control and Prevention,
National Center for Health Statistics (Jan. 11, 2007).

^15For example, a claim with a diagnosis code for heart failure and a
procedure code for the implantation of a pacemaker would be assigned to
DRG 115, "Permanent Cardiac Pacemaker Implant with Acute Myocardial
Infarction, Heart Failure, or Shock." CMS would then multiply the weight
for DRG 115 by a base payment amount, or the amount that Medicare would
pay for an average unit of service if no other payment adjustments
applied, to determine the hospital's base payment amount for the stay. The
higher the DRG weight the more costly the stays assigned to that DRG are
estimated to be and the higher the payment.

^16For example, in fiscal year 2007, a hospital receives an outlier
payment if its estimated cost for a stay is at least $24,485 more than its
DRG payment for that stay. The actual amount of the outlier payment will
equal 80 percent of the difference between the two amounts.

^1742 U.S.C. S 1395ww(d)(4)(C) (2000).

CMS uses data from the MEDPAR file to revise the DRGs for the coming
fiscal year. It revises the DRGs in a two-step process: reclassification
of DRGs and calculation of DRG payment weights. First, CMS incorporates
new codes into the IPPS that represent new diagnoses or procedures by
assigning them to the same DRGs as existing codes for clinically similar
diagnoses or procedures.^18 Using data from the MEDPAR file, CMS may
reclassify the DRG assignment of inpatient stays with a particular
procedure or diagnosis code if it determines the inpatient stays are more
similar in their clinical characteristics and costliness to a DRG other
than the DRG to which those stays were previously assigned.^19 CMS will
create a new DRG if it determines that the inpatient stays involving newly
identified diagnoses and procedures cannot be described by any of the
existing DRGs.^20 The classification of most DRGs does not change from
year to year.^21

The second step in revising the DRGs involves calculating weights across
all DRGs, so that the DRGs reflect the expected relative differences in
costliness of inpatient stays for the upcoming fiscal year. Prior to
fiscal year 2007, CMS annually derived each DRG's weight by dividing the
average charge per inpatient stay for that DRG by the average charge per
inpatient stay across all DRGs for a fiscal year. Effective fiscal year
2007, CMS uses charge data from the MEDPAR file and hospitals'
cost-to-charge ratios from Medicare cost reports to estimate the costs per
inpatient stay. CMS then uses these average estimated costs to measure the
relative costliness of inpatient stays that will be assigned to each DRG.

^18Codes for new procedures and diagnoses are assigned by CMS and the
National Center for Health Care Statistics, respectively, in a process
that is separate from the DRG reclassification process. New codes are
assigned for procedures and diagnoses that are deemed so different from
existing procedures and diagnoses that they warrant their own unique
identifiers.

^19CMS makes a clinical determination to reclassify a DRG based on input
from the public as well as its own clinical staff. CMS evaluates
costliness by comparing the average charge for inpatient stays with the
particular code--as calculated using data from the MEDPAR file--to the
average charges for all stays assigned to the current and proposed DRGs,
respectively.

^20For fiscal year 2008, CMS has proposed to refine the classification of
and expand the number of DRGs from 538 to 745 to better reflect severity
of illness and the cost of treating Medicare beneficiaries. 72 Fed. Reg.
24,680, 24,687 (May 3, 2007).

^21For fiscal year 2007, for example, CMS added 20 new DRGs, made changes
to the classifications of approximately 32 existing DRGs, and deleted 8
DRGs. See 71 Fed. Reg. 47,870, 47,879 (Aug. 18, 2006).

In reclassifying and weighting DRGs, CMS generally requires that the data
meet three criteria: (1) the data must be representative of the Medicare
population; (2) the data must be timely--that is, they should be the most
recent data available; and (3) the data must be complete--meaning that CMS
needs total charges or other measure of costliness for all services
provided during an inpatient stay. Charge data collected at the
inpatient-stay level allow CMS to appropriately measure relative
costliness across the DRGs.

DRG Classifications and Payment Weights for Any Given Fiscal Year Are Based on
Inpatient Claims Data That Do Not Reflect the Cost of the Most Recent
Technologies

The DRG classifications and payment weights for any given fiscal year are
based on data from the MEDPAR file for inpatient services provided 2
fiscal years prior, and therefore, do not reflect the cost of the most
recently used technologies. For example, during the summer of 2006, when
CMS was finalizing the DRGs for fiscal year 2007, the most recent data
pertained to inpatient services provided through the end of fiscal year
2005, and did not reflect the cost of technologies first adopted by
hospitals in fiscal year 2006.

The time lag in the data that are used to set DRG classifications and
weights is primarily due to two factors in combination: the time it takes
to annually finalize the DRGs, and the time it takes for CMS to process
each inpatient claim into a MEDPAR record. First, Medicare law requires
that DRG classifications and weights be revised annually and published in
the Federal Register on or before the August 1 before each fiscal year.^22
Fiscal years begin October 1 and end the following September 30. In order
to obtain public input, CMS generally publishes its proposed DRGs for the
coming fiscal year in the Federal Register each April and accepts comments
for 60 days before publishing the final DRGs by August.

The second factor that affects the incorporation of the cost of new
technologies into the MEDPAR file involves the time it takes for CMS to
process each inpatient claim into a MEDPAR record. Before a record for an
inpatient stay can be added to the MEDPAR file, the hospital must submit
the claim, a private contractor must process and pay the claim, and CMS
must create a MEDPAR record using information on the claim. It takes about
6 months from the time of the inpatient stay to the time the MEDPAR record
for that inpatient stay is created. In addition, the MEDPAR record may not
be added to the MEDPAR file until as much as 3 months later, since the
MEDPAR file is updated quarterly--in December, March, June, and September.
This means that MEDPAR records are not available to CMS until 6 to 9
months after the inpatient stay has occurred. (See fig. 1.)

^2242 U.S.C. S 1395ww(d)(6) (2000).

Figure 1: Process by Which CMS Used the MEDPAR File to Propose and
Finalize DRGs for Fiscal Year 2007 DRG Payments

Add-on Payments for New Technologies Can Supplement DRG Payments

Because DRG payments for a given fiscal year are based on claims for
inpatient services provided 2 fiscal years prior, Medicare can provide
hospitals with add-on payments, in addition to the DRG-based payments, for
inpatient stays involving certain new technologies. CMS designates
technologies for add-on payments if they meet specified criteria for being
new, costly, and a substantial clinical improvement over existing
technologies. CMS considers a technology new if no more than 2 to 3 years
have passed between the date when the technology was first introduced on
the market, as identified by CMS, and the payment year.^23 At the end of
this period, CMS assumes the costs for the technology to be fully
reflected in the most recent MEDPAR file and supplemental add-on payments
are no longer necessary. CMS considers a new technology costly if the
average amount charged by hospitals for all inpatient stays involving the
technology exceeds a charge threshold or a predetermined amount.^24 CMS
considers a new technology a substantial clinical improvement over
existing technologies if the technology has one or more unique clinical
advantages--for example, the technology diagnoses a medical condition in a
patient population where that condition was previously undetectable.

Every year, CMS accepts applications from technology manufacturers,
hospitals, and other stakeholders, in which they present evidence that
certain technologies meet the criteria for add-on payments in the coming
fiscal year. When CMS publishes its final DRG classifications and weights,
it summarizes each application, and explains why the particular technology
was approved or rejected for add-on payments. For fiscal year 2007, CMS
approved one new application and continued add-on payments for two
technologies approved for fiscal year 2006.^25 As a result, hospitals
receive an add-on payment, in addition to a DRG payment, when they submit
a claim to Medicare that includes the code for a procedure involving one
of those three technologies. The amount of the add-on payment is
determined on a claim-by-claim basis; the hospital receives up to half the
estimated cost of the technology, depending on the amount by which the
total cost of the inpatient stay is estimated by CMS to exceed the
DRG-based payment.^26

^23BIPA required that CMS collect cost data on a new technology for a 2 to
3 year period and that a new technology be eligible for add-on payments
during that period. Pub. L. No. 106-554, app. F, sec. 533(b), S
1886(d)(5)(K)(ii)(III),114 Stat. 2763, 2763A-548 (codified at 42 U.S.C. S
1395ww(d)(5)(K)(ii)(III) (2000)). CMS established that the beginning of
the 2 to 3 years would be the date it determines that the technology
became available on the market. 69 Fed. Reg. 28,196, 28,237 (May 18,
2004).

^24Specifically, the charge for the inpatient stay involving the new
technology must exceed the lower of two thresholds: (1) 75 percent of the
base payment amount adjusted to reflect charges, or (2) 75 percent of a
standard deviation beyond the average charge of all stays that fall within
the DRGs to which the new technology is assigned. 42 U.S.C. S
1395ww(d)(5)(K)(ii)(I) (2000 & Supp. III 2003).

^25Specifically, CMS approved the new technology application for the
X-STOP Interspinous Process Decompression System and continued add-on
payments for the Endovascular Graft Repair of the Thoracic Aorta and
Restore (R) Rechargeable Implantable Neurostimulator.

CMS Has Used External Data to Inform DRG Reclassification and to Evaluate New
Technology Add-on Payment Applications

CMS has used external data for two purposes: to inform DRG
reclassification and to evaluate new technology add-on payment
applications. To inform DRG reclassification, CMS accepts the submission
of external data that are intended to demonstrate that inpatient stays
involving a new technology are costlier on average than the other
inpatient stays in the same DRG. CMS uses data from the MEDPAR file to
validate the external data submitted. Generally, CMS will not make a
reclassification decision for a DRG involving a new technology if the
technology is so new that it does not appear in the MEDPAR file. To
evaluate new technology add-on payment applications, CMS has generally
used external data in conjunction with data from the MEDPAR file to
evaluate whether a new technology meets one of three eligibility criteria,
specifically, the criterion related to cost.

CMS Has Used External Data to Inform the DRG Reclassification Process on
Inpatient Stays Involving New Technologies

CMS officials told us they have used external data to inform the DRG
reclassification process. External data are submitted by stakeholders as
part of a request to reclassify--from one DRG to another--certain
procedure codes involving particular new technologies. Although CMS will
accept the submission of external data, it has used data from the MEDPAR
file to validate the external data submitted.^27 Specifically, when
external data are submitted for a proposed DRG reclassification for a
procedure or new technology, CMS's policy is to find the same or similar
evidence in the MEDPAR file. CMS encourages stakeholders to submit their
external data for DRG reclassification purposes by the December before the
issuance of the proposed rule the following April.

^26To calculate the amount of the add-on payment, CMS converts the charges
billed for the inpatient stay to an estimated cost using the hospital's
cost-to-charge ratio. If the total estimated cost for the inpatient stay
is higher than the DRG-based payment, then the add-on payment is 50
percent of the difference, up to half of the estimated cost of the new
technology.

^27CMS was directed by the conference report accompanying the BBA to only
use external data it can validate. H.R. Rep. No. 105-217, at 734 (1997)
(Conf. Rep.).

Although there is no formal application process to request a DRG
reclassification, CMS explained its policy for accepting external data
submissions in its July 30, 1999, notice of final rulemaking.^28,29 It
stated that external data submissions must be sufficiently
detailed--include applicable hospital and beneficiary identifiers,
procedure and diagnosis codes, admission and discharge dates, and total
charges for each inpatient stay involving the codes--so that CMS can
validate whether the same, or similar, inpatient stays appear in the
MEDPAR file. CMS also requires that the external data submitted comprise a
complete set, or representative sample, of cases involving the technology.
CMS will not reclassify a procedure code from one DRG to another based on
the external data submission alone. As a result, CMS generally will not
make a DRG reclassification involving a technology that is so new it does
not yet appear in the MEDPAR file.^30

CMS Has Used External Data to Evaluate Applications for New Technology Add-on
Payments

CMS has used external data to evaluate applications for new technology
add-on payments to better recognize the cost of technologies that are
clinically beneficial yet would not be fully reflected in the MEDPAR file.
CMS designates technologies for add-on payments if they meet specified
criteria for being new, costly, and a substantial clinical improvement
over existing technologies. CMS's use of external data is limited to its
evaluation of the cost criterion. CMS has generally used external data and
data from the MEDPAR file to evaluate whether a new technology that is
being considered for an add-on payment meets the criterion for being
considered costly.^31

^28CMS maintains a list of DRG issues raised throughout the year by
manufacturers, providers, and the general public. It reviews these
requests with its staff and analyzes external data if those data were
submitted. CMS does not address each and every DRG reclassification
request it receives throughout the year in its proposed or final rules for
the upcoming year, nor does it track how many times external data were
submitted with those reclassification requests.

^2964 Fed. Reg. 41,490, 41,499-504.

^30CMS has made one reclassification decision for a medical
technology--drug-eluting stents--that had not yet received Food and Drug
Administration (FDA) approval, and therefore, did not appear in the MEDPAR
file. CMS explained that it took this action because of the potential for
drug-eluting stents to significantly impact the treatment of coronary
blockages, and the expectation that hospitals would rapidly adopt the
technology upon FDA approval. 67 Fed. Reg. 49,983, 50,004 (Aug. 1, 2002).

^31CMS outlined the new technology add-on application process in its
September 7, 2001, final rule. 66 Fed. Reg. 46,902, 46,916.

As of fiscal year 2007, according to our review of CMS regulations and our
interviews with CMS officials, CMS has received few applications for
add-on payments--a total of 25, which is an average of about 5 per year
since fiscal year 2002. All but two applications were submitted by device
and drug manufacturers. When CMS receives an application for a new
technology add-on payment, it first evaluates whether the technology meets
the criterion of being new before it evaluates the technology under the
cost and clinical improvement criteria. The majority of new technology
add-on payment applications have been rejected because the technology
failed to meet the newness criterion. For these applications, CMS did not
have to review any information related to the cost and clinical
improvement criteria, including external data related to the cost
criterion. Of the 25 applications received, CMS evaluated 14 under the
cost criterion. Of these 14 technologies, CMS approved 7 for new
technology add-on payments.

When CMS evaluates new technologies under the cost criterion, it uses
external data in conjunction with data from the MEDPAR file to determine
whether the technology meets the cost criterion.^32 Table 1 illustrates
three hypothetical scenarios in which CMS, during fiscal year 2007, could
use external data in conjunction with data from the fiscal year 2006
MEDPAR file in determining if a new technology is eligible for add-on
payments for fiscal year 2008 under the cost criterion.

^32For example, in explaining the fiscal year 2007 final rule, CMS
summarized the application for the X STOP Interspinous Process
Decompression System (X STOP). The device manufacturer provided CMS with
data from clinical trials demonstrating that the total costs of inpatient
stays involving X STOP met the cost criterion. CMS verified the
standardized charge for the stays in the MEDPAR file. 71 Fed. Reg. 47,870,
48,003 (Aug. 18, 2006).

Table 1: Three Hypothetical Scenarios in Which CMS, during Fiscal Year
2007, Could Determine Eligibility under the Cost Criterion for Fiscal Year
2008 Add-on Payments

                                 Is the                                       
                                 charge for  What data does                   
                                 the new     the                              
                                 technology  applicant^a                      
                                 included in submit to CMS  How does CMS use  
                                 the total   to demonstrate data from the     
                    Does the     charge in   that the new   fiscal year 2006  
                    fiscal year  the fiscal  technology     MEDPAR file to    
                    2006 MEDPAR  year 2006   meets the      verify that the   
                    file contain MEDPAR file charge         new technology    
                    data on      for each    threshold^b    meets the charge  
                    stays        stay        established by threshold         
                    involving    involving   CMS for the    established by    
                    the new      the         cost           CMS for the cost  
Scenario         technology?  technology? criterion?     criterion?        
The new          Yes          Yes         The applicant  CMS conducts its  
technology                                submits an     own analysis      
became available                          analysis       using data from   
on the U.S.                               comparing the  the fiscal year   
market in 2006,                           average charge 2006 MEDPAR file  
was adopted by a                          of the stays   to validate the   
few hospitals by                          involving the  accuracy of the   
the end of that                           new technology applicant's       
year, and was                             based on data  analysis.         
provided by                               from the                         
those hospitals                           fiscal year                      
to Medicare                               2006 MEDPAR                      
beneficiaries.                            file to the                      
                                             charge                           
                                             threshold.                       
The new          Yes          No^c        Using external CMS evaluates the 
technology was                            data, the      reasonableness of 
not available on                          applicant      the applicant's   
the U.S. market                           estimates what estimated         
until 2007.                               U.S. hospitals hospital charges  
However, in 2006                          would have     for the           
it was provided                           charged for    technology. If    
to Medicare                               the technology CMS determines    
beneficiaries by                          on average in  the estimates are 
a few U.S.                                2006. The      reasonable, it    
hospitals that                            applicant then verifies the      
received the                              submits an     accuracy of the   
technology at no                          analysis       applicant's       
charge for use                            comparing the  analysis of data  
in clinical                               average charge using the fiscal  
trials.                                   of inpatient   year 2006 MEDPAR  
                                             stays          file.             
                                             involving the                    
                                             new technology                   
                                             in 2006 based                    
                                             on data from                     
                                             the fiscal                       
                                             year 2006                        
                                             MEDPAR file to                   
                                             the charge                       
                                             threshold.                       
The technology   No           Not         The applicant  CMS verifies the  
was not                       applicable  submits        accuracy of the   
available on the                          external data  applicant's       
U.S. market                               from U.S.      external data     
until 2007.                               hospitals from analysis using    
Although it was                           2007--for      data from the     
provided to                               example,       fiscal year 2006  
patients in 2006                          copies of      MEDPAR file. For  
in clinical                               bills from     example, CMS      
trials, the                               hospitals that could locate      
trials were                               provided the   stays in the      
conducted in                              technology to  fiscal year 2006  
Europe and the                            Medicare       MEDPAR file that  
patients were                             beneficiaries  likely would have 
not Medicare                              during         involved the      
beneficiaries.                            inpatient      technology had it 
                                             stays. The     been available to 
                                             applicant      U.S. hospitals.   
                                             compares the   CMS would         
                                             average charge calculate an      
                                             calculated for average charge    
                                             these stays,   for those stays,  
                                             which includes add the estimated 
                                             charges for    charge for the    
                                             the new        new technology,   
                                             technology, to and subtract the  
                                             the charge     estimated charge  
                                             threshold.     for the           
                                                            technology it may 
                                                            be replacing to   
                                                            arrive at a total 
                                                            estimated charge  
                                                            for those stays   
                                                            had they involved 
                                                            the new           
                                                            technology.       

Source: GAO based on CMS regulations and interviews with CMS officials.

^a"Applicant" refers to a manufacturer, hospital, or other stakeholder.
Although hospitals receive the add-on payment, almost all applicants for
new technology add-on payments have been technology manufacturers.

^bSpecifically, the charge for the inpatient stay involving the new
technology must exceed the lower of two thresholds: (1) 75 percent of the
base payment amount adjusted to reflect charges, or (2) 75 percent of a
standard deviation beyond the average charge of all stays that fall within
the DRG to which the new technology is assigned. 42 U.S.C. S
1395ww(d)(5)(K)(ii)(I) (2000 & Supp. III 2003).

^cData from the MEDPAR file may not include charges for a new technology if
patients were in a clinical trial and the technology was provided to
hospitals at no charge.

Data from Other Government Agencies Have Limitations for Setting DRG Payments
for Inpatient Stays Involving New Technologies

Data collected and used by other government agencies have limitations for
CMS's use in setting DRG payments for inpatient stays involving new
technologies. This is because, when setting DRG payments, CMS generally
needs data that are representative of the Medicare population, timely, and
complete in that the data include the total charge or other measure of
costliness for all services provided during an inpatient stay, including
new technologies. The data we identified from BLS, VA, DOD, and AHRQ were
either not representative of the Medicare population, were no timelier
than data from the MEDPAR file, or were not complete.

BLS collects monthly selling prices for samples of products from three
industries that may have data relevant to CMS because these data include
price information for new technologies: medical instruments,
pharmaceuticals, and biological products.^33 These data, collected from
manufacturers, are used to publish the Producer Price Index (PPI), which
tracks the inflation of prices by producers of goods and services at the
national level.^34 Because BLS cannot obtain pricing for every medical
instrument and pharmaceutical and biological product sold, it employs a
sampling methodology to track prices. Using probability statistics, BLS
selects a sample of products whose price changes over time will be
representative of the price changes characteristic of the medical
instrument and pharmaceutical and biological product industries.
Generally, BLS selects a new sample of products per industry every 7-8
years.^35 The monthly selling prices collected include prices for
transactions between manufacturers and hospitals, wholesalers, group
purchasing organizations, or other customers.

BLS data have a number of limitations that would affect CMS's use in
setting DRG payments. Because the selling prices reflect transactions
between manufacturers and a variety of purchasers such as group purchasing
organizations as well as hospitals, not all of these prices are directly
relevant for setting DRG payments. To set payments, CMS needs data that
reflect hospitals serving Medicare beneficiaries. In addition, since BLS
relies on a sample of products from each industry, and the sample is
generally updated on average every 7-8 years, it is unlikely that BLS will
have price data for a new technology that CMS does not already have, or
cannot obtain from a manufacturer. Finally, the BLS data lack information
needed by CMS on the costliness of inpatient stays involving the
technology relative to other inpatient stays; instead, they only include
price data for the technology alone.

^33BLS collects data on approximately 800 industries.

^34BLS collects data for other purposes, including the Consumer Price
Index (CPI), which tracks the price of goods purchased by consumers. BLS
also collects data on the total reimbursement to hospitals and other
medical facilities for providing a sample of medical procedures.

^35For pharmaceuticals, BLS augments its sample continuously to reflect
new products.

Two types of VA data, price data from the federal supply schedule (FSS)
and data on inpatient stays at VA hospitals, also have limitations that
would affect CMS's use in setting DRG payments. VA collects data from drug
and device manufacturers on the prices manufacturers charge their
Most-Favored Customers (MFC).^36 These data are used to negotiate prices
on the FSS, which is a schedule of prices for products used by federal
agencies. Prices on the FSS are awarded at equal to or better than the
prices manufacturers charge their MFCs. Because all federal agencies and
programs may access FSS price information on the Internet, CMS already has
access to these prices. Similar to BLS data, FSS data are not complete for
CMS's purposes because they lack information on the costliness of
inpatient stays involving the technology relative to other inpatient
stays.

VA also collects data on inpatient stays at its medical centers.^37 These
data are complete for CMS's purposes in that they include all services
provided during inpatient stays and their associated costs, including the
cost of any new technologies. However, there are still limitations for
CMS's use of these data in setting DRG payments. First, the costs of
providing care at VA medical centers may not be representative of the
costs of providing care at hospitals that provide care to Medicare
beneficiaries. VA is a provider of services and, as such, VA has the
authority to purchase new technologies at discounted rates through various
federal purchasing options, such as the FSS. Medicare, on the other hand,
is a payer--not a provider--of services and does not purchase drugs and
devices for hospitals. Therefore, Medicare does not negotiate discounts on
behalf of hospitals providing services to Medicare beneficiaries.
Furthermore, VA inpatient stay data are no timelier than MEDPAR data for
determining payments to hospitals. For example, VA's allocation of funding
to its medical centers for fiscal year 2007 is based on data spanning
fiscal years 2003 through 2005.^38 Medicare used fiscal year 2005 data to
develop fiscal year 2007 DRG payments.

^36MFC is generally the customer that receives the best discount or has
the best price arrangement on a given item.

^37Although medical care to eligible veterans is provided by the VA, some
care is provided through arrangements with affiliated academic medical
centers and other contractors.

DOD data also have limitations for CMS's use in setting DRG payments. DOD
health care delivery consists of two integrated systems: the direct care
system delivered by DOD hospitals, known as Military Treatment Facilities
(MTF), and the civilian system. The latter is coordinated by the TRICARE
Management Activity (TMA), which contracts with managed care organizations
to deliver care, including inpatient services. Data from the DOD direct
care system would not meet CMS's criterion for completeness for two
reasons. First, DOD collects overall cost data at the facility level
rather than the inpatient-stay level. CMS needs charge or cost data at the
inpatient stay level to set DRG payments. Second, while DOD uses cost and
pricing data from a variety of sources when purchasing medical products,
such as drugs and devices for its MTFs, these data alone are not
appropriate for CMS's use in setting DRG payments because CMS needs
information on the costliness of inpatient stays involving the technology
relative to other inpatient stays.

Data from the DOD civilian system also have limitations for CMS's use in
setting DRG payments. TMA pays for inpatient stays using a DRG-based
payment system that is modeled on the Medicare IPPS.^39 Although TMA's
data would be complete for CMS's purposes in that the data include total
charges for all services provided during an inpatient stay, they would not
meet CMS's criterion for representativeness. According to DOD, its
population tends to be younger and healthier and, therefore, not
comparable to the Medicare population.^40

AHRQ collects claims data from nearly all nongovernmental acute care
hospitals in 38 states and these data represent approximately 90 percent
of inpatient stays in the United States. AHRQ partners with state
organizations, which collect claims data directly from hospitals; these
data are then submitted to AHRQ. According to AHRQ, these data, which are
available to researchers through the Healthcare Cost and Utilization
Project (HCUP) claims database, are representative of the Medicare
population overall.^41 The data from the HCUP database are also complete
in that they include charge, diagnosis, and procedure information from
Medicare as well as private payers. Although data from the HCUP database
would meet CMS's criteria for being representative of the Medicare
population and are complete, these data are less timely than data from the
MEDPAR file. AHRQ data lag between 15-18 months, so, for example, if CMS
were to use data from the HCUP database to set payments for fiscal year
2007, the latest available data from AHRQ would include inpatient services
for calendar year 2004, while the latest available data from the MEDPAR
file would include inpatient services from fiscal year 2005.

^38VA's fiscal year 2007 budget for its medical centers and other medical
services was based on actuarial projections of expected enrollees and
patients for fiscal year 2007.

^3971 Fed. Reg. 60,112 (Oct. 12, 2006). DOD noted that the vast majority
of DOD's DRGs are the same as CMS's DRGs; however, DOD has additional DRGs
for neonatal cases and age-defined mental health and substance abuse
diagnoses.

^4052 Fed. Reg. 32,992, 32,998 (Sept. 1, 1987).

Concluding Observations

Data from the MEDPAR file remain the primary data source for setting DRG
payments because they include all charges from inpatient claims for
inpatient services provided to all Medicare beneficiaries across all
hospitals paid under the IPPS. CMS needs these data to determine payment
for each DRG relative to other DRGs. In instances where data from the
MEDPAR file have lacked charge information for certain stays involving new
technologies, CMS has used external data to inform the DRG
reclassification process and to evaluate new technology add-on payment
applications. To set DRG payments, CMS needs data that meet criteria of
being representative, timely, and complete. Although BLS, VA, DOD, and
AHRQ collect data for their own purposes that could potentially be useful
to CMS, these data are limited in their utility to set DRG payments
because they do not always meet CMS's criteria.

Agency and Other External Comments

In commenting on a draft of this report, CMS stated that it agreed with
our findings and reiterated its commitment to using external data when
appropriate. (See app. I.) DOD said it had no comments on the draft of
this report. (See app. II.) We received comments from VA via email. The
department agreed with the facts as they pertain to VA. We also sent a
draft of this report to DOL. DOL did not provide comments. Representatives
from American Hospital Association (AHA), Association of American Medical
Colleges (AAMC), and the Biotechnology Industry Organization (BIO)
provided oral comments on a draft of this report. They said they agreed
with our findings related to the use of external data by CMS.

^41Researchers must sign a data use agreement with AHRQ to access HCUP
data.

With regard to our finding that data from other government agencies have
limitations for CMS's use in setting DRG payments, both AAMC and BIO said
we should have discussed CMS's use of data from sources other than the
federal government. As we discussed in the draft report, CMS has used
external data from sources other than the federal government including
manufacturer data to inform DRG reclassification and evaluate new
technology add-on applications.

AAMC said it was concerned that we only examined how CMS used the external
data and did not conduct an evaluation of CMS's policy for using external
data. However, as discussed in the draft report, an examination of CMS's
policy for accepting external data was not within the scope of the report.

In addition, CMS, AAMC and AHA offered technical comments on the draft of
this report, which we incorporated as appropriate.

We are sending a copy of this report to the Administrator of CMS and
interested congressional committees. We will also provide copies to others
on request. The report is available online at no charge on GAO's Web site
at [18]http://www.gao.gov .

If you or your staff have any questions, please contact me at (202)
512-7114 or [19][email protected] . Contact points for our Offices of
Congressional Relations and Public Affairs may be found on the last page
of this report. GAO staff who made major contributions to this report are
listed in appendix III.

A. Bruce Steinwald
Director, Health Care

Appendix I: Comments from the Centers for Medicare & Medicaid Services

Appendix II: Comments from the Department of Defense

Appendix III: GAO Contact and Staff Acknowledgments

GAO Contact

A. Bruce Steinwald, (202) 512-7114 or [email protected]

Acknowledgments

In addition to the contact above, Maria Martino, Assistant Director;
Melanie Anne Egorin; Yorick F. Uzes; and Craig Winslow made key
contributions to this report.

(290376)

GAO's Mission

The Government Accountability Office, the audit, evaluation and
investigative arm of Congress, exists to support Congress in meeting its
constitutional responsibilities and to help improve the performance and
accountability of the federal government for the American people. GAO
examines the use of public funds; evaluates federal programs and policies;
and provides analyses, recommendations, and other assistance to help
Congress make informed oversight, policy, and funding decisions. GAO's
commitment to good government is reflected in its core values of
accountability, integrity, and reliability.

Obtaining Copies of GAO Reports and Testimony

The fastest and easiest way to obtain copies of GAO documents at no cost
is through GAO's Web site ( [20]www.gao.gov ). Each weekday, GAO posts
newly released reports, testimony, and correspondence on its Web site. To
have GAO e-mail you a list of newly posted products every afternoon, go to
[21]www.gao.gov and select "E-mail Updates."

Order by Mail or Phone

The first copy of each printed report is free. Additional copies are $2
each. A check or money order should be made out to the Superintendent of
Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more
copies mailed to a single address are discounted 25 percent. Orders should
be sent to:

U.S. Government Accountability Office 441 G Street NW, Room LM Washington,
D.C. 20548

To order by Phone: Voice: (202) 512-6000
TDD: (202) 512-2537
Fax: (202) 512-6061

To Report Fraud, Waste, and Abuse in Federal Programs

Contact:

Web site: [22]www.gao.gov/fraudnet/fraudnet.htm
E-mail: [23][email protected]
Automated answering system: (800) 424-5454 or (202) 512-7470

Congressional Relations

Gloria Jarmon, Managing Director, [24][email protected] (202) 512-4400 U.S.
Government Accountability Office, 441 G Street NW, Room 7125 Washington,
D.C. 20548

Public Affairs

Susan Becker, Acting Manager, [email protected], (202) 512-4800 U.S.
Government Accountability Office, 441 G Street NW, Room 7149 Washington,
D.C. 20548

To view the full product, including the scope
and methodology, click on [25]GAO-07-46 .

For more information, contact A. Bruce Steinwald at (202) 512-7114 or
[email protected].

Highlights of [26]GAO-07-46 , a report to congressional committees

September 2007

MEDICARE INPATIENT HOSPITAL PAYMENTS

CMS Has Used External Data for New Technologies in Certain Instances and
Medicare Remains Primary Data Source

Under Medicare, hospitals generally receive fixed payments for inpatient
stays based on diagnosis-related groups (DRG), a system that classifies
stays by patient diagnoses and procedures. The Centers for Medicare &
Medicaid Services (CMS) annually uses its own data to reclassify DRGs. CMS
also makes add-on payments for stays involving new technologies that meet
three eligibility criteria. Stakeholders may submit data that are external
to CMS as part of a DRG reclassification request or an add-on payment
application. The Medicare Prescription Drug, Improvement, and
Modernization Act of 2003 required GAO to examine whether CMS could
improve its use of external data, including using data collected by other
government agencies for DRG payments. As discussed with the committees of
jurisdiction, GAO examined (1) to what extent CMS has used external data
in determining payments for inpatient stays involving new technologies,
and (2) to what extent can external data from other government agencies be
used by CMS in determining DRG payments for inpatient stays involving new
technologies. GAO interviewed officials from CMS and industry
stakeholders. GAO interviewed officials from Bureau of Labor Statistics
(BLS), Department of Veterans Affairs (VA), Department of Defense (DOD),
and Agency for Healthcare Research and Quality (AHRQ) because these
agencies may have data useful to CMS. GAO also reviewed regulations and
other CMS materials.

CMS has used external data for two purposes: to inform DRG
reclassification and to evaluate new technology add-on payment
applications. To inform  DRG reclassification, CMS accepts the submission
of external data that are intended to demonstrate that inpatient stays
involving a new technology are costlier on average than the other
inpatient stays in the same DRG. CMS uses its data from the Medicare
Provider Analysis and Review (MEDPAR) file to validate the external data
submitted. Specifically, when external data are submitted for a proposed
DRG reclassification for a procedure or new technology, CMS's policy is to
find the same or similar evidence in the MEDPAR file. Generally, CMS will
not make a reclassification decision for a DRG involving a new technology
if the technology is so new that it does not appear in the MEDPAR file. 
To evaluate new technology add-on payment applications, CMS has generally
used external data in conjunction with data from the MEDPAR file to
evaluate whether a new technology meets one of the three eligibility
criteria, specifically the criterion related to cost.

Data from other government agencies have limitations for CMS's use in
setting DRG payments for inpatient stays involving new technologies. This
is because when setting DRG payments, CMS generally needs data that are
representative of the Medicare population, timely, and complete in that
the data include the total charge or other measure of costliness for all
services provided during an inpatient stay, including new technologies.
The data we identified from other government agencies were either not
representative of the Medicare population, were not timelier than data
from the MEDPAR file, or were not complete.

Data from the MEDPAR file remain the primary data source for setting DRG
payments because they include all charges from paid inpatient claims for
inpatient services provided to all Medicare beneficiaries across all
hospitals paid under the IPPS. In instances where data from the MEDPAR
file have lacked charge information for certain stays involving new
technologies, CMS has used external data to inform the DRG
reclassification process and to evaluate new technology add-on payment
applications. To set DRG payments, CMS needs data that meet criteria of
being representative, timely, and complete. Although BLS, VA, DOD, and
AHRQ collect data for their own purposes that could potentially be useful
to CMS, these data are limited in their utility to set DRG payments
because they do not always meet CMS's criteria.

In commenting on a draft of this report, CMS stated that it agreed with
GAO's findings.

References

Visible links
  18. http://www.gao.gov/
  19. mailto:[email protected]
  20. http://www.gao.gov/
  21. http://www.gao.gov/
  22. http://www.gao.gov/fraudnet/fraudnet.htm
  23. mailto:[email protected]
  24. mailto:[email protected]
  25. http://www.gao.gov/cgi-bin/getrpt?GAO-07-46
  26. http://www.gao.gov/cgi-bin/getrpt?GAO-07-46
*** End of document. ***