oversight

Medicare: Lack of Price Transparency May Hamper Hospitals' Ability to Be Prudent Purchasers of Implantable Medical Devices

Published by the Government Accountability Office on 2012-01-13.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

               United States Government Accountability Office

GAO            Report to the Chairman, Committee on
               Finance, U.S. Senate



January 2012
               MEDICARE

               Lack of Price
               Transparency May
               Hamper Hospitals’
               Ability to Be Prudent
               Purchasers of
               Implantable Medical
               Devices




GAO-12-126
                                              January 2012

                                              MEDICARE
                                              Lack of Price Transparency May Hamper Hospitals’
                                              Ability to Be Prudent Purchasers of Implantable
                                              Medical Devices
Highlights of GAO-12-126, a report to the
Chairman, Committee on Finance, U.S.
Senate




Why GAO Did This Study                        What GAO Found
Implantable medical devices (IMD)—            From 2004 through 2009, expenditures for hospital IMD procedures increased
including a variety of cardiac and            from $16.1 billion to $19.8 billion, an increase of 4.3 percent per year—a rate
orthopedic devices provided to                equal to that of Medicare spending for other hospital procedures. While cardiac
Medicare beneficiaries in inpatient or        and orthopedic procedures accounted for nearly all IMD-related expenditures,
outpatient hospital settings—represent        orthopedic procedures accounted for most of the increase in such expenditures
a significant share of hospitals’ supply      during this period. Utilization increased at a faster rate for orthopedic devices and
costs. Hospitals purchase IMDs                accounted for the majority of changes in expenditures for IMD procedures during
directly from manufacturers or through        the period.
group purchasing organizations (GPO)
and their purchasing agreements often
                                              The information GAO obtained on the amounts hospitals paid for selected IMDs
contain confidentiality clauses
                                              showed substantial variation. For a number of reasons, the detailed information
restricting them from revealing to third
parties the prices they pay for such          needed to accurately compare prices across hospitals—both the specific model
devices. Policymakers are concerned           and sale price net of discounts and rebates—was not reported by all respondents
that the lack of price transparency           for all IMDs in our study. However, data from 31 hospitals indicated substantial
inhibits competition in the device            variation in reported prices for cardiac devices. For example, the difference
market, leading to higher costs for           between the lowest and highest price hospitals reported paying for a particular
hospitals, and ultimately higher              automated implantable cardioverter defibrillator (AICD) model was $6,844. The
Medicare spending. GAO was asked to           difference between the highest and lowest price reported for another AICD model
examine (1) Medicare spending and             was $8,723. The price differences for the remaining two AICD models in our
utilization trends for procedures             study fell in between $6,844 and $8,723. The median prices across the four
involving IMDs provided to                    AICD models ranged from $16,445 to $19,007. A factor particular to the IMD
beneficiaries, and (2) what available         market that affects prices hospitals pay is the influence of physicians on
information shows about the prices            hospitals’ IMD purchasing. Although physicians are not involved in price
hospitals pay for IMDs and any factors        negotiations, they often express strong preferences for certain manufacturers
particular to the IMD market that             and models of IMDs. To the extent that physicians in the same hospital have
influence those prices. GAO analyzed          different preferences for IMDs, it may be difficult for the hospital to obtain volume
the most recently available Medicare          discounts from particular manufacturers. Also, confidentiality clauses barring
inpatient and outpatient hospital claims      hospitals from sharing price information make it difficult to inform physicians
from fiscal years 2004 through 2009.
                                              about device costs and thereby influence their preferences. Other factors that
GAO requested price information on
                                              influence IMD prices include the degree of seller competition and a hospital’s
five devices from 60 hospitals, 6
GPOs, Department of Defense (DOD)
                                              market share.
medical centers, and the Department
of Veterans Affairs (VA) health system.       These data suggest that some hospitals have substantially less bargaining power
GAO interviewed officials from GPOs,          with the small group of companies that manufacture particular IMDs and
device manufacturers, large hospital          consequently face challenges in obtaining more favorable prices. The lack of
systems, and small hospitals about the        price transparency and the substantial variation in amounts hospitals pay for
factors that affect the prices hospitals      some IMDs raise questions about whether hospitals are achieving the best prices
pay for IMDs.                                 possible. Any excess or unnecessary costs that hospitals incur through IMD
                                              pricing may be passed onto the Medicare program.

                                              The Department of Health and Human Services, VA, and DOD reviewed a draft
                                              of this report and had no general comments.



View GAO-12-126. For more information,
contact James C. Cosgrove at (202) 512-7114
or cosgrovej@gao.gov.

                                                                                       United States Government Accountability Office
Contents


Letter                                                                                     1
             Background                                                                    5
             Medicare’s Payment Systems Indirectly Account for Prices
                Hospitals Pay for IMDs; Efforts Are Under Way to Better
                Account for These Prices                                                   9
             Medicare Expenditures for IMD Procedures Increased from About
                $16 Billion to $20 Billion from 2004 through 2009—Driven
                Largely by Increased Utilization of Orthopedic Devices                   12
             Information Available on the Prices Hospitals Paid for Selected
                IMDs Shows Substantial Variation; Hospital-Physician
                Relationships Are a Particular Influence on Prices Paid                  22
             Concluding Observations                                                     29
             Agency and Industry Comments                                                30

Appendix I   GAO Contact and Staff Acknowledgments                                       32



Tables
             Table 1: Type of Detailed Information Reported on Cardiac and
                      Orthopedic IMDs, Fiscal Year 2010                                  23
             Table 2: Median Prices and Differences between the Lowest and
                      Highest Prices Hospitals Paid for Cardiac Devices, Fiscal
                      Year 2010                                                          26


Figures
             Figure 1: Components of a Total Hip Replacement                               6
             Figure 2: IMD Expenditures per Medicare FFS Beneficiary, 2004
                      through 2009                                                       13
             Figure 3: Orthopedic IMD Expenditures per Medicare FFS
                      Beneficiary, 2004 through 2009                                     15
             Figure 4: Cardiac IMD Expenditures per Medicare FFS Beneficiary,
                      2004 through 2009                                                  17
             Figure 5: Orthopedic IMD Utilization per 10,000 Medicare FFS
                      Beneficiaries, 2004 through 2009                                   19
             Figure 6: Cardiac IMD Utilization per 10,000 Medicare FFS
                      Beneficiaries, 2004 through 2009                                   20
             Figure 7: Contribution toward Per-Beneficiary Expenditure Growth
                      between 2004 and 2009, by Orthopedic IMD Procedures                21




             Page i                                    GAO-12-126 Implantable Medical Devices
Abbreviations

AICD              automated implantable cardioverter defibrillator
AHA               American Hospital Association
APC               ambulatory payment classification
CMS               Centers for Medicare & Medicaid Services
CRT-D             cardiac resynchronization therapy defibrillator
DOD               Department of Defense
FFS               fee-for-service
GPO               group purchasing organization
IMD               implantable medical device
MS-DRG            Medicare-severity diagnosis related group
PPS               prospective payment system
VA                Department of Veterans Affairs



This is a work of the U.S. government and is not subject to copyright protection in the
United States. The published product 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.




Page ii                                            GAO-12-126 Implantable Medical Devices
United States Government Accountability Office
Washington, DC 20548




                                   January 13, 2012

                                   The Honorable Max Baucus
                                   Chairman
                                   Committee on Finance
                                   United States Senate

                                   Dear Mr. Chairman:

                                   Implantable medical devices (IMD)—a broad range of items that include
                                   coronary stents, cardiac defibrillators, and hip and knee joint
                                   replacements—can prolong life and improve the quality of life for patients
                                   that receive them. 1 As the Medicare population grows and beneficiaries
                                   expect to live longer, the demand for IMDs is likely to increase steadily.
                                   Therefore, the Centers for Medicare & Medicaid Services (CMS)—the
                                   agency that administers the Medicare program—will continue to be a
                                   major payer for procedures involving IMDs provided in both inpatient and
                                   outpatient hospital settings.

                                   For hospitals, IMDs represent a significant share of their supply costs;
                                   further, for some procedures, the cost of IMDs can be the most expensive
                                   part of an inpatient hospital stay or outpatient procedure. As with other
                                   medical supplies, hospitals buy IMDs at prices negotiated directly with
                                   manufacturers or at prices negotiated by buying intermediaries, known as
                                   group purchasing organizations (GPO). By pooling the purchasing power
                                   of multiple providers, GPOs may be strongly positioned to bargain for
                                   lower prices from manufacturers. 2 However, hospitals face challenges in
                                   obtaining information on IMD prices in the market because comparable
                                   price information across manufacturers generally is not publicly




                                   1
                                    For this report, we define implantable medical devices as artificial devices implanted
                                   entirely within the body that are intended to remain in the body permanently. However,
                                   some of these devices have a limit to their effective life span and will require replacement.
                                   2
                                    See GAO, Group Purchasing Organizations: Services Provided to Customers and
                                   Initiatives Regarding Their Business Practices, GAO-10-738 (Washington, D.C.: Aug. 24,
                                   2010).




                                   Page 1                                              GAO-12-126 Implantable Medical Devices
available. 3, 4 Device manufacturers often require hospitals to sign
purchasing agreements that contain confidentiality clauses restricting
them from revealing to third parties the prices they paid for medical
devices. To the extent that information on IMD prices in the market is not
disclosed, the ability of hospitals to bring comparative price information to
bear in IMD purchasing negotiations and decisions may be limited.
Without such competitive pressure, the prices hospitals pay for IMDs may
be higher than they otherwise would be. 5

In an environment of increasing health care costs, the efficiency of the
IMD market has implications for Medicare cost containment.
Policymakers are concerned that the lack of price transparency inhibits
competition in the device market, leading to higher costs for hospitals,
and ultimately higher spending in Medicare. 6 You asked us to examine
various issues regarding Medicare spending and the prices of implantable
medical devices. In this report, we examined: (1) how Medicare’s hospital
payment systems account for the prices hospitals pay for IMDs, (2) the
spending and utilization trends for procedures involving IMDs provided to
Medicare beneficiaries, and (3) what available information shows about
the prices hospitals pay for IMDs and any factors particular to the IMD
market that influence those prices.

To describe how Medicare’s payment systems account for the prices
hospitals pay for IMDs, we reviewed annual regulations for the inpatient
prospective payment system and outpatient prospective payment system.


3
 Certain information on IMD prices may be publicly available through sources such as the
Department of Veterans Affairs (VA) National Acquisition Center website. However, the
database is not complete and represents price ceilings established under VA’s national
contracts with IMD manufacturers.
4
 GAO has reported on challenges consumers face in obtaining cost information for the
health care services they obtain. See GAO, Health Care Price Transparency: Meaningful
Price Information Is Difficult for Consumers to Obtain Prior to Receiving Care,
GAO-11-791 (Washington, D.C.: Sept. 23, 2011).
5
 See D. Andrew Austin and Jane G. Gravelle, Does Price Transparency Improve Market
Efficiency: Implications of Empirical Evidence in Other Markets for the Health Sector,
RL34101 (Washington, D.C.: Congressional Research Service, Apr. 29, 2008.)
6              th
 During the 110 Congress, Senators Grassley and Specter sponsored legislation that
would require device manufacturers to submit the average and median sales prices of
covered devices to the Secretary of Health and Human Services on a quarterly basis.
             th
S. 2221, 110 Cong. § 2 (as introduced and referred to the S. Comm. on Finance, Oct. 23,
2007). This legislation did not become law.




Page 2                                           GAO-12-126 Implantable Medical Devices
We also interviewed officials from CMS and the Medicare Payment
Advisory Commission.

To identify the Medicare spending and utilization trends for IMD
procedures provided to Medicare FFS beneficiaries, we analyzed the
most recently available Medicare inpatient and outpatient hospital claims
for a 6-year period, from fiscal years 2004 through 2009. 7 Using
procedure and payment codes obtained from CMS and industry sources,
we identified 364 inpatient and 85 outpatient procedure codes that were
generally associated with an IMD. 8 We matched these procedure codes
with 115 inpatient and 30 outpatient “device dependent” payment codes
in each claim. Because our list of IMD-dependent codes represented
procedures performed almost exclusively in a hospital setting, we
excluded procedures performed in other facilities (e.g., ambulatory
surgical centers). We only included services and expenditures when the
IMD-dependent procedure code was the primary procedure associated
with the payment for the claim submitted to Medicare. 9, 10 We examined
the extent to which changes in expenditures were explained by changes
in utilization or average cost. 11 We determined that the claims data we
used were sufficiently reliable for the purposes of our analysis by
performing appropriate electronic data checks.




7
 These data only include payments to hospitals for procedures performed in the 50 states
and the District of Columbia. Unless noted otherwise, we report all years on a fiscal year
(Oct. 1 through Sept. 30) basis.
8
 Because codes used by hospitals to submit claims to Medicare are broadly defined to
cover a set of related services and may or may not include the implantation of a medical
device, we relied on several sources to develop a list of procedure codes for services that
generally involve an IMD—what we refer to as “device-dependent.” To ensure that these
codes met our definition of an IMD, we developed a content analysis using key word
criteria to identify exclusions—procedures involving devices that did not meet our
definition of an IMD—and inclusions—procedures generally involving devices that did
meet our definition. We completed our compilation of inpatient procedure and payment
codes by reviewing CMS payment system final rules.
9
 We use the term “case” to indicate claims where an IMD procedure largely determined
the Medicare payment.
10
    All expenditures are reported in nominal terms and are not adjusted for inflation.
11
  Because inpatient and outpatient services are paid for differently, we examine these
separately.




Page 3                                               GAO-12-126 Implantable Medical Devices
To determine what information is available on the prices hospitals pay for
IMDs, we selected five specific devices used in the five IMD-related
procedures that ranked highest in terms of Medicare spending in 2009.
The five devices were primary total knee implants, primary total hip
implants, coronary drug-eluting stents, automated implantable
cardioverter defibrillators (AICD), and cardiac resynchronization therapy
defibrillators (CRT-D). 12 We asked 60 randomly selected hospitals, 13
6 GPOs, 14 Department of Defense (DOD) medical centers and hospitals,
and the Department of Veterans Affairs (VA) 15 to report fiscal year 2010
pricing information for each of these five IMDs. Of the 60 hospitals,
39 hospitals—or 65 percent—provided IMD pricing data; in some cases,
the information reflected purchases made by the health system with
which the hospital is affiliated or for a point in time during fiscal year 2010.
In addition, 5 of the 6 GPOs, 8 DOD medical centers or hospitals, and the
VA health care system provided IMD pricing data. 16 To compare the
prices that hospitals paid for cardiac IMDs, we limited our analysis to
those hospitals that provided detailed information on both the specific
model and the sale price net of discounts and rebates. Not all hospitals
provided pricing information for each device model in our study. 17 To


12
  In a total joint replacement, “primary” denotes the original implant and differentiates it
from a revision total joint replacement which replaces the original implant.
13
  These hospitals performed at least one implantation procedure for each of the five IMDs
based on Medicare claims data for 2009.
14
  The six GPOs were selected based on their reported 2007 purchasing volume in Health
Industry Distributors Association, Group Purchasing Organization & Integrated Delivery
Network: Market Brief (Alexandria, Va.: July 2009).
15
  We obtained device price data from the VA National Prosthetic Patient Database, which
contains all IMD price information reported by VA medical centers. We did not use data
from the National Acquisition Center because certain IMDs, such as coronary stents, are
not purchased through national contracts. In addition, VA medical centers may obtain
waivers to purchase IMDs outside of national contracts for medical reasons only.
Therefore, the data on the National Acquisition Center website do not reflect the full range
of prices that VA medical centers actually pay for IMDs.
16
  DOD data include IMD pricing information from two of three Navy medical centers, two
Navy hospitals, and four of five Air Force medical centers. Data obtained from the Army
were not representative of the full range of prices hospitals paid and were not used in this
analysis.
17
  The number of hospitals providing sale price net of discounts and rebates for a particular
device model ranged from 3 to 19. In addition, to ensure comparability of prices for device
model types that may or may not be purchased with leads, we removed cases where the
price reported included leads. A lead is the connection between the heart and the power
source for an AICD or CRT-D.




Page 4                                               GAO-12-126 Implantable Medical Devices
                           compare the prices that hospitals paid for orthopedic IMDs, we used
                           examples from the information provided by hospitals and GPOs on behalf
                           of their member hospitals. The data we obtained on cardiac and
                           orthopedic IMD prices are not generalizable. To identify the factors that
                           affect the prices hospitals pay for IMDs, we reviewed relevant literature
                           and interviewed officials from the 6 GPOs from which we requested
                           pricing data, 3 nongovernment hospital systems with the highest
                           revenues for 2009 and 2010, 18 5 large IMD manufacturers that
                           volunteered to participate through a trade association, 2 small hospitals
                           that performed a low volume of procedures involving orthopedic IMDs
                           commonly used among Medicare beneficiaries in 2009, and professional
                           associations representing orthopedic surgeons and cardiologists.

                           We conducted this performance audit from June 2010 to December 2011
                           in accordance with generally accepted government auditing standards.
                           Those standards require that we plan and perform the audit to obtain
                           sufficient, appropriate evidence to provide a reasonable basis for our
                           findings and conclusions based on our audit objectives. We believe that
                           the evidence obtained provides a reasonable basis for our findings and
                           conclusions based on our audit objectives.



Background
Device Features and        Cardiac IMDs include cardiac rhythm management devices (such as
Industry Characteristics   pacemakers, AICDs, CRT-Ds) and coronary stents. A pacemaker
                           monitors the patient’s underlying heart rhythm and delivers an electrical
                           pulse to cause the heart to beat at the desired rate. An AICD is similar to
                           the pacemaker in design, but is capable of delivering a higher energy
                           electrical pulse—called defibrillation shocks—to correct more serious
                           rapid and sustained heart rhythm irregularities. A CRT-D is a combination
                           of a pacemaker and defibrillator. A coronary stent (either bare metal or




                           18
                              The three hospital systems were selected based on the reported 2009 revenues in
                           Modern Healthcare, Big dividends: annual survey of hospital systems shows a strong
                           financial comeback thanks in large part to resurgent investment portfolios (Chicago, Ill.:
                           June 2010) and 2010 revenues in Modern Healthcare, A solid year: annual survey shows
                           health systems posted strong revenue and earnings, but expenses are also rising
                           (Chicago, Ill.: June 2011).




                           Page 5                                             GAO-12-126 Implantable Medical Devices
drug-eluting) is a wire mesh tube used to prop open a blocked coronary
artery. 19

Orthopedic devices include joint implants—mostly commonly hips and
knees—as well as spinal devices used for spinal fusion. Typically, hip and
knee implants have a variety of components and are made up of different
materials, which may be combined into different configurations to make
the total device. The components and configurations of orthopedic
devices can vary by manufacturer. For example, for a hip replacement
with four different components there are different configurations and
different types of materials (metal, plastic, ceramic) that can be used as
well as different ways to secure the implant (cemented or fitted into the
bone with new bone growth to hold the implant in place). See figure 1.

Figure 1: Components of a Total Hip Replacement




The markets for specific cardiac and orthopedic IMDs are generally
dominated by a few large manufacturers. In the cardiac rhythm
management segment, the market is dominated by three U.S.
manufacturers. Also, four manufacturers share the drug-eluting stent




19
  Some stents are coated with drugs that slowly release and help keep the artery open
(drug-eluting stents), while others are not (bare-metal stents).




Page 6                                           GAO-12-126 Implantable Medical Devices
                       market. 20 In the hip and knee replacement orthopedic segment, sales by
                       four manufacturers accounted for about 83 percent of the market share in
                       2009.


IMD Purchasing         Typically, hospitals negotiate IMD prices with manufacturers directly or
Agreements             through GPOs. Contract negotiations between hospitals or GPOs and
                       manufacturers can occur at the national or local level. A custom or local
                       contract establishes a price for products that were not part of the GPO’s
                       overall contract portfolio or achieves a better price for a specific product
                       than what the national GPO contract offered. Contracts can include an
                       agreed-upon price for one or more devices and reflect rebates—an
                       agreement by the manufacturer to return a portion of the payment, which
                       may depend on a hospital’s commitment to purchase a certain volume.

                       In addition, hospitals may engage consulting firms that may act as data
                       intermediaries by gathering price information from clients and sharing
                       benchmarks for IMD costs with them. However, to the extent that such
                       information is derived from purchasing agreements containing
                       confidentiality clauses, these data intermediaries may be potential targets
                       of litigation by device manufacturers. 21


Medicare’s Payment     Medicare has separate inpatient and outpatient prospective payment
Systems for Hospital   systems (PPS) for paying hospitals for services provided to beneficiaries
Services               enrolled in FFS Medicare. Under the inpatient PPS, a hospital generally
                       receives a fixed, predetermined payment amount for a bundle of services
                       provided during a beneficiary’s hospital stay. 22 Inpatient payment rates
                       are based on Medicare-severity diagnosis-related groups (MS-DRG), a
                       system that classifies inpatient stays according to both patients’ clinical
                       conditions (the primary diagnosis along with any secondary illnesses and


                       20
                         One of the four manufacturers of drug-eluting stents plans to exit the market by the end
                       of 2011, leaving only three manufacturers in that market segment.
                       21
                         Researchers examining this issue have noted lawsuits alleging violations of
                       confidentiality clauses by data intermediaries providing price information to hospitals. See
                       Jeffrey C. Lerner, Daniel M. Fox, Todd Nelson, and John B. Reiss, “The Consequence of
                       Secret Prices: The Politics of Physician Preference Items” Health Affairs, vol. 27, no. 6
                       (2008).
                       22
                        Throughout this report, we use the term “stay” to represent a patient’s hospitalization,
                       which CMS and hospitals refer to as a discharge for data-reporting purposes.




                       Page 7                                              GAO-12-126 Implantable Medical Devices
complications developed during the stay) and the procedures patients
receive. 23 In general, an inpatient MS-DRG payment is calculated using a
base payment amount—a per hospital-stay rate for operating costs that
efficient hospitals would be expected to incur in furnishing covered
inpatient services—multiplied by a MS-DRG relative weight. The MS-
DRG relative weight signifies the average costliness of stays assigned to
that MS-DRG relative to the average costliness of other inpatient stays.
CMS updates MS-DRG weights annually to incorporate any changes in
the cost of inpatient care.

Under the outpatient PPS, generally, the unit of payment is the individual
service. Outpatient PPS rates are based on ambulatory payment
classifications (APC), a system that classifies services based on their
similarity in terms of clinical characteristics and cost. For each APC,
Medicare makes a single bundled payment for the primary service and
any ancillary or supportive services. For example, the APC payment for a
pacemaker implantation procedure represents a bundled payment for the
pacemaker device, routine supplies, and the operating or procedure
room. An APC payment is calculated by multiplying an APC relative
weight by the conversion factor, a dollar amount that translates the
relative weight into dollar amounts. 24 The APC relative weight measures
the resource requirements of the service and is based on the median cost
of services in the APC. 25 CMS updates APC weights annually to account
for any changes in the cost of outpatient care.

CMS calculates inpatient and outpatient payments based on data from
Medicare cost reports and charges from Medicare claims. Each year,
hospitals submit Medicare cost reports to CMS identifying hospitals’
actual costs for services rendered to all patients, not just Medicare
beneficiaries. Upon a Medicare beneficiary’s discharge, hospitals submit




23
  Prior to 2008, CMS assigned inpatient stays to diagnosis-related groups (DRG) which
classified inpatient stays by patient diagnosis and the procedures they received; CMS
replaced DRGs with MS-DRGs to better account for severity of illness.
24
  CMS initially set the conversion factor so that projected total payments would equal the
estimated amount that would have been spent under the old cost-based outpatient
payment system, after correcting for some anomalies in statutory formulas.
25
  CMS scaled all of the APC relative payment weights to APC 601, a midlevel clinic visit,
because it is one of the most frequently performed services.




Page 8                                             GAO-12-126 Implantable Medical Devices
                      Medicare claims to CMS identifying charges for services delivered to the
                      beneficiary—such as procedures involving IMDs.


                      Medicare does not directly purchase IMDs but accounts for hospitals’ IMD
Medicare’s Payment    prices indirectly. To help set prospective payment rates for inpatient and
Systems Indirectly    outpatient hospital procedures, Medicare’s payment systems use data
                      from cost reports and claims. CMS collects data on hospitals’ costs for all
Account for Prices    services and supplies from Medicare cost reports. The cost reports
Hospitals Pay for     capture data on hospitals’ costs for all services and supplies, including
IMDs; Efforts Are     IMDs, but do not separately identify specific IMD costs. Rather, hospitals
                      report costs for all medical supplies in one category, aggregating the
Under Way to Better   costs of low-cost items, such as surgical gloves, with the costs of high-
Account for These     cost items, such as cardiac defibrillators and total knee replacements. On
                      hospitals’ claims submitted for inpatient services, charges are billed by
Prices                categories of service. For example, an inpatient hospital claim for the
                      implantation of an AICD could include charges for an AICD and surgical
                      dressings in the supplies category. Using these aggregated cost data,
                      Medicare’s inpatient and outpatient payments indirectly account for prices
                      hospitals reportedly paid for IMDs.

                      Compared to hospital cost reports and charges on inpatient claims, the
                      data on hospitals’ charges on outpatient claims provide more specific
                      information about hospitals’ charges for individual IMDs. This occurs
                      because of the way Medicare classifies outpatient procedures for claims
                      and payment purposes. Specifically, CMS requires hospitals to include
                      device codes on claims where an appropriate code exists to describe a
                      device utilized in the procedure. Further charges reported on the
                      outpatient claims generally reflect what the hospital reports as the full cost
                      of the IMD.

                      Due to a lag in data, the cost of the newest IMD technology may not be
                      reflected in data used to set payment rates. Although CMS uses the most
                      recently available hospital claims data in order to help set inpatient MS-
                      DRG and outpatient APC payments, the claims reflect data that predate
                      the year for which rates are being set by 2 years. For example, 2009
                      inpatient payments are based on Medicare claims from 2007.

                      To better reflect costs of new technology, Medicare makes additional
                      payments under the inpatient and outpatient payment systems. These
                      additional payments are generally made for 2 to 3 years, until sufficient
                      data on hospitals’ new technology costs can be collected and used in
                      payment rate setting. For inpatient procedures, CMS makes add-on


                      Page 9                                     GAO-12-126 Implantable Medical Devices
payments—made in addition to the MS-DRG payment—for procedures
involving new IMDs that meet certain criteria for newness, clinical benefit,
and cost. 26 For outpatient procedures involving new IMD technology,
CMS makes pass-through payments for eligible new IMDs that can be
used in an existing service, and new technology APC payments for a new
procedure involving an IMD that cannot be adequately described by an
existing APC. Pass-through payments are made in addition to the APC
payment for the procedure involving the new IMD. 27 CMS assigns
services to new technology APCs based on cost information collected on
applications for new technology status. 28

CMS is taking further steps to better account for hospitals’ reported IMD
prices in the data it uses to set inpatient and outpatient prospective
payment rates. Specifically, CMS has modified the way hospitals report
costs of IMDs on Medicare cost reports. CMS created a new cost
category called “Implantable Devices Charged to Patients,” which
requires hospitals to report separately their high-cost IMDs and their other
lower-cost medical supplies. The new cost category is intended to provide
CMS with more accurate information about hospitals’ reported costs
specific to IMDs for use in setting inpatient and outpatient payment rates.
The revised cost report with the new cost category for “Implantable
Devices Charged to Patients” was made available for use for cost
reporting periods beginning on or after May 1, 2009. According to CMS,
data collected with the revised cost reports will likely be available for
setting payment rates for 2013.




26
  CMS determines the add-on payment on a claim-by-claim basis; payment is limited to
50 percent of the hospital’s costs above the standard MS-DRG payment, up to a
maximum of 50 percent of the estimated cost of the new device. For fiscal year 2012,
there is one brand-specific new technology eligible to receive an add-on payment.
27
  Pass-through payments are based on each hospital’s actual reported cost of the new
IMD minus the device offset amount—the portion of each APC payment rate CMS could
reasonably attribute to the cost of the device that the new IMD replaces. As of October
2011, there are three device categories eligible for pass-through payment.
28
  New technology APC payment levels are set at the midpoint range of the cost category;
for example, payment made for New Technology APC 1507 ($500 to $600) is made at
$550. In calendar year 2011, there are three procedures receiving payment through a new
technology APC.




Page 10                                           GAO-12-126 Implantable Medical Devices
In addition to more accurately accounting for IMD costs, CMS intends to
link payment to the quality and efficiency of care provided—a move
toward value-based purchasing in Medicare. One of the principles of
value-based purchasing is having good data on performance; currently
such data generally are lacking for specific brands of implantable medical
devices. Most IMDs are introduced as incremental modifications to
existing devices without studies of performance relative to alternative
devices and for specific patient populations. 29 There are registries of
postoperative outcomes for some cardiac devices, such as AICDs, but
none for orthopedic or spinal devices outside of specific organizations
such as Kaiser Permanente. 30, 31 Furthermore, because hospital data
currently are embedded in multiple data systems—such as medical
records, operating room logs, purchasing department records, and billing
systems—it can be difficult to match which device brand was used with a
particular patient. 32




29
  James C. Robinson, “Value-Based Purchasing for Medical Devices,” Health Affairs, vol.
27 no. 6 (2008).
30
  In 2005, CMS issued a Medicare national coverage policy for AICDs that requires
providers implanting these devices for certain clinical conditions to enter implant data into
a clinical registry.
31
  There is a joint registry under development. The American Joint Replacement Registry
is a nonprofit organization for data collection and research on total hip and knee
replacements. This registry is a collaborative effort supported by orthopedic medical
professional societies, hospitals, health insurers, government agencies, and medical
device manufacturers.
32
  Efforts to improve information known about patient safety and medical devices would be
supported by the implementation of unique device identifiers. The Food and Drug
Administration Amendments Act of 2007 requires the promulgation of regulations
establishing a unique device identification system, which will require the devices to bear a
unique identifier. Pub. L. No. 110-85, § 226, 121 Stat. 823, 854 (codified at 21 U.S.C.
§ 360i(f)) (the provision allows for the exemption of particular devices or types of devices).




Page 11                                             GAO-12-126 Implantable Medical Devices
                            From 2004 through 2009, Medicare expenditures for IMD hospital
Medicare                    procedures increased from about $16 billion to $20 billion. While cardiac
Expenditures for IMD        and orthopedic procedures accounted for nearly all IMD-related
                            expenditures, orthopedic procedures accounted for most of the increase
Procedures Increased        in such expenditures during our period of study. Utilization increased at a
from About $16              faster rate for orthopedic devices and explained the majority of changes
Billion to $20 Billion      in expenditures for IMD procedures during the period.

from 2004 through
2009—Driven Largely
by Increased
Utilization of
Orthopedic Devices

Medicare Expenditures for   From 2004 through 2009, expenditures for hospital IMD procedures
IMD Procedures Increased    increased from $16.1 billion to $19.8 billion, an increase of 4.3 percent
from About $16 Billion to   per year—a rate equal to that of Medicare spending for other hospital
                            procedures. Expressed in terms of expenditures per beneficiary—a
$20 Billion between 2004    measure which accounts for changes in the size of Medicare’s FFS
and 2009, with Orthopedic   population—IMD expenditures increased from $444 to $561, an increase
IMD Procedures Driving      of 4.8 percent per year (see fig. 2). Inpatient expenditures for IMD
Most of That Growth         procedures increased from $15.1 billion to $17.0 billion, an increase of
                            2.4 percent per year. Outpatient expenditures for IMD procedures
                            increased from $1.0 billion to $2.9 billion, an increase of 24.1 percent per
                            year. In comparison, Medicare expenditures for all non-IMD procedures
                            increased by 4.3 percent per year, 4.8 percent when measured per
                            beneficiary. Inpatient and outpatient expenditures for these procedures
                            increased by 2.9 and 10.8 percent per year, respectively.




                            Page 12                                    GAO-12-126 Implantable Medical Devices
Figure 2: IMD Expenditures per Medicare FFS Beneficiary, 2004 through 2009




Note: Reported expenditures include Medicare’s payment to the hospital for the set of services
associated with the procedure performed. The other IMD category includes neurological and
outpatient orthopedic IMD procedures.


As figure 2 indicates, orthopedic and cardiac procedures accounted for
nearly all IMD-related Medicare expenditures from 2004 through 2009.
Furthermore, during this period, the share of inpatient orthopedic
expenditures relative to all IMD-related Medicare expenditures increased
from 38 percent to 45 percent. In addition, an increasing share of cardiac
procedures shifted to the outpatient setting. Generally, Medicare pays
hospitals a relatively lower rate for the same procedures delivered in the
outpatient setting than in the inpatient setting. Specifically, as a share of
Medicare’s total IMD expenditures, inpatient cardiac IMD procedures
decreased from 56 percent to 40 percent while outpatient cardiac IMD
procedures increased from 5 to 12 percent.




Page 13                                                 GAO-12-126 Implantable Medical Devices
From 2004 through 2009, orthopedic procedures accounted for most of
the growth in Medicare IMD-related expenditures. Medicare expenditures
for orthopedic IMD procedures increased from $6.1 billion to $9.0 billion,
an increase of 8.1 percent per year. 33, 34 Procedures related to knees,
hips, shoulders, and the spine accounted for nearly all of Medicare’s
orthopedic IMD expenditures in 2009. 35 The average growth rate of
expenditures related to each of these procedure types exceeded that of
non-IMD hospital procedures. 36 Spinal fusion procedures had the highest
growth in per beneficiary expenditures—more than doubling during the
period (see fig. 3).




33
  Medicare expenditure growth rates for orthopedic IMD procedures exceeded that of non-
IMD hospital procedures throughout our period of study.
34
  Revisions, procedures that replace part or all of an IMD, accounted for 8.9 percent of
these expenditures in 2004 and 11.0 percent in 2009, increasing from $0.5 billion to
$0.9 billion.
35
  For example, in 2009, these procedures accounted for 98 percent of the program’s
orthopedic IMD expenditures.
36
  On a per beneficiary basis, expenditures related to hips, shoulders, and the spine grew
the most between 2007 and 2009.




Page 14                                            GAO-12-126 Implantable Medical Devices
Figure 3: Orthopedic IMD Expenditures per Medicare FFS Beneficiary, 2004 through
2009




Note: Reported expenditures include Medicare’s payment to the hospital for the set of services
associated with the procedure performed.


Cardiac procedures accounted for relatively little of the growth in
Medicare IMD-related expenditures from 2004 through 2009. Medicare
expenditures for cardiac IMD procedures increased from $9.8 billion in
2004 to $10.3 billion in 2009, an increase of 1.2 percent per year. In
2009, procedures involving stents, pacemakers, AICDs, and CRT-Ds
accounted for 93 percent of expenditures for cardiac IMD procedures. 37
Over the entire 2004 through 2009 period, only cardiac expenditures
related to AICD procedures increased at a higher annual rate than that of
non-IMD procedures. On a per beneficiary basis, expenditures peaked in
2006 but reached their second lowest point of the period in 2007, largely
resulting from expenditures for coronary stent procedures, which


37
  Other cardiac IMDs included vascular grafts and vena cava filters.




Page 15                                                 GAO-12-126 Implantable Medical Devices
decreased by 15 percent that year (see fig. 4). Furthermore, inpatient
expenditures for cardiac IMD procedures generally decreased over the
2004 through 2009 period, while those expenditures in the lower payment
outpatient setting generally increased. Inpatient expenditures for cardiac
IMD procedures reached their peak of $10.0 billion in 2005 but dropped to
$7.9 billion in 2009, decreasing 5.5 percent per year over that period. In
contrast, outpatient expenditures for cardiac IMD procedures increased
from $0.8 billion in 2004 to $2.4 billion in 2009, an increase of
25.5 percent per year. During that period, expenditures for coronary stent,
AICD, and pacemaker procedures decreased in the inpatient setting but
increased in the outpatient setting. 38




38
  Medicare’s Recovery Audit Contractor program may be associated with both the shift of
cardiac IMD procedures from the inpatient to the outpatient setting and the higher
outpatient expenditure growth rates for IMD procedures compared with non-IMD
procedures. From 2005 to 2008, these audits collected overpayments for inpatient
defibrillator and pacemaker procedures that could have been performed in the outpatient
setting, possibly prompting other hospitals to change their admissions patterns.




Page 16                                          GAO-12-126 Implantable Medical Devices
                             Figure 4: Cardiac IMD Expenditures per Medicare FFS Beneficiary, 2004 through
                             2009




                             Note: Reported expenditures include Medicare’s payment to the hospital for the set of services
                             associated with the procedure performed.




Utilization Increased at a   From 2004 through 2009, utilization of IMD procedures increased from
Faster Rate for Orthopedic   about 1.3 million cases to about 1.6 million cases, an increase of
Devices Compared with        3.8 percent per year. 39 Per 10,000 beneficiaries, the utilization of IMD
                             cases increased from 357 to 440, or an increase of 4.3 percent per year.
Cardiac Devices and
                             During our period of study, inpatient IMD utilization increased from
Explained Most of the        1.1 million cases to 1.2 million cases, an increase of 1.4 percent per year.
Changes in IMD-related       In contrast, outpatient IMD utilization increased from about 180,000 cases
Medicare Expenditures        to about 363,000 cases, an increase of 15.0 percent per year.
from 2004 through 2009

                             39
                               We use the term “case” to indicate a claim where an IMD procedure drove the Medicare
                             payment.




                             Page 17                                                 GAO-12-126 Implantable Medical Devices
Utilization increased at a faster rate for orthopedic devices compared with
cardiac devices. From 2004 through 2009, orthopedic IMD procedure
utilization increased from about 556,000 cases to about 697,000 cases,
yielding an annual growth rate of 4.6 percent per year or 5.1 percent
when measured per beneficiary. While utilization per beneficiary of
orthopedic IMD procedures related to knees and hips were highest, 40 the
fastest increase in utilization of orthopedic IMD procedures came from
those related to spinal fusions, increasing at a rate of 12.1 percent per
year (see fig. 5). 41 In contrast, utilization of cardiac IMD procedures
increased from about 705,000 cases to about 797,000 cases, yielding an
annual growth rate of 2.5 percent per year, 3.0 percent when measured
per beneficiary. Not all cardiac procedures followed a steady increase in
utilization. For example, between 2006 and 2007, utilization associated
with stent procedures decreased by 9.3 percent and increased thereafter
(see fig. 6). 42 Inpatient utilization of cardiac IMD procedures decreased
from 154 to 140 cases per 10,000 beneficiaries, while such utilization in
the outpatient setting increased from 41 to 85 cases per 10,000
beneficiaries during the period. Per-beneficiary utilization related to
cardiac stents, AICDs, and pacemakers all decreased in the inpatient
setting and increased in the outpatient setting.




40
  During our period of study, utilization of knee and hip replacements grew at annual rates
of 5.5 and 2.2 percent, respectively.
41
  The percentage of spinal fusions performed with either a cage or bone morphogenetic
protein increased throughout the period, going from 42 percent in 2004 to 60 percent in
2009. The percentage of spinal fusions that used both a cage and bone morphogenetic
protein increased from 7 percent in 2004 to 17 percent in 2009.
42
  A 2007 New England Journal of Medicine study that found treating Stable Coronary
Disease with stent-related procedures did not produce improvements compared to other
types of treatment may be a factor in the decrease in stent utilization. See William Boden,
et al., “Optimal Medical Therapy With or Without PCI for Stable Coronary Disease,” The
New England Journal of Medicine, vol. 356, no. 15 (2007).




Page 18                                            GAO-12-126 Implantable Medical Devices
Figure 5: Orthopedic IMD Utilization per 10,000 Medicare FFS Beneficiaries, 2004
through 2009




Page 19                                       GAO-12-126 Implantable Medical Devices
Figure 6: Cardiac IMD Utilization per 10,000 Medicare FFS Beneficiaries, 2004
through 2009




Between 2004 and 2009, the overall increase in per-beneficiary Medicare
expenditures for IMD procedures resulted more from increased utilization
than from an increased average payment per claim. Specifically, our
analysis shows that utilization contributed to 67 percent of the increase in
such inpatient expenditures. Utilization and average payment growth
rates varied by year during the study period. The largest rate of increase
in inpatient per-beneficiary utilization occurred from 2007 to 2008. During
this time period, CMS adjusted inpatient discharge payment rates to
better distinguish beneficiaries with comorbidities and complications, 43
which was followed by the only decrease of average inpatient
expenditures. The average IMD-related inpatient payment decreased by


43
  CMS transitioned to MS-DRGs, replacing DRGs as the method for grouping patient
stays in determining payments. MS-DRGs better capture severity of illness differences
among patients.




Page 20                                          GAO-12-126 Implantable Medical Devices
                                       5.4 percent while per-beneficiary utilization of those same procedures
                                       increased by 6.5 percent. Similarly, utilization contributed to 69 percent of
                                       the increase in per-beneficiary Medicare expenditures for outpatient IMD
                                       procedures.

                                       Among orthopedic IMD procedures, utilization contributed to the
                                       majority—62 percent—of the overall growth in per-beneficiary Medicare
                                       expenditures between 2004 and 2009. Utilization contributed to most of
                                       the increased expenditures for three of the four main types of orthopedic
                                       IMD procedures (see fig. 7). Four of the five orthopedic IMD procedures
                                       with the largest expenditures in 2009 had growth that resulted more from
                                       increased utilization than an increased average payment. 44

Figure 7: Contribution toward Per-Beneficiary Expenditure Growth between 2004 and 2009, by Orthopedic IMD Procedures




                                       Note: Reported expenditures include Medicare’s payment to the hospital for the device as well as the
                                       set of services associated with the procedure performed.


                                       The extent to which utilization drove changes in per-beneficiary Medicare
                                       expenditures for cardiac IMD procedures varied by hospital setting.
                                       Decreased utilization contributed to 94 percent of the decrease of such
                                       inpatient expenditures between 2004 and 2009. The decrease in
                                       utilization accounted for all of the decrease in per-beneficiary inpatient


                                       44
                                         These five procedures were total knee replacement, total hip replacement, lumbar and
                                       lumbosacral (spinal) fusion with a posterior technique, partial hip replacement, and total
                                       shoulder replacement. They represent 77 percent of the Medicare expenditures for
                                       orthopedic IMD procedures in that year. Per-beneficiary utilization of the partial hip
                                       replacement procedure decreased during our period of study, and therefore, all of its
                                       increase in per-beneficiary expenditures resulted from an increased average payment per
                                       claim.




                                       Page 21                                                 GAO-12-126 Implantable Medical Devices
                              expenditures related to pacemakers. Decreased utilization contributed to
                              61 percent and 52 percent of the decrease in per-beneficiary inpatient
                              expenditures related to AICDs and coronary stents, respectively. In
                              comparison, increased utilization contributed to 67 percent of the increase
                              in per-beneficiary outpatient expenditures for cardiac IMD procedures.
                              Increased utilization contributed to 79 percent and 93 percent of the
                              increase in such expenditures related to pacemakers and AICDs,
                              respectively. Furthermore, increased utilization explained 59 percent of
                              the increase in per-beneficiary outpatient expenditures related to coronary
                              stents.


                              Complete and comparable information on the prices hospitals paid for the
Information Available         various models of IMDs was limited, as reflected in responses to our data
on the Prices                 request. The price information that was provided showed substantial
                              variation in the prices hospitals paid for the same type of device. A
Hospitals Paid for            particular factor that influences the prices hospitals pay for IMDs is
Selected IMDs Shows           hospitals’ ability to manage relationships with physicians and
Substantial Variation;        manufacturers.

Hospital-Physician
Relationships Are a
Particular Influence
on Prices Paid

Various Factors Limit the     Detailed information on both the specific model purchased and the sale
Availability of Detailed      price net of discounts and rebates is needed to accurately compare the
Information on the Prices     amount hospitals paid for various types of IMDs. 45 However, not all 39
                              hospitals, 5 GPOs, the VA health care system, and 8 DOD medical
Hospitals Paid for Selected
                              centers or hospitals responding to our request provided us with such
IMDs                          detailed information for three types of cardiac IMDs and two types of
                              orthopedic IMDs purchased during fiscal year 2010. Only 31 hospitals,
                              1 GPO, and 1 DOD medical center provided detailed information on
                              cardiac device prices, and only 14 hospitals,1 DOD medical center, and
                              1 DOD hospital provided detailed information on orthopedic device prices


                              45
                                The size of rebates and discounts may differ based on the volume of devices purchased
                              and the extent to which a hospital purchases a predetermined number of a certain device
                              from a particular manufacturer.




                              Page 22                                         GAO-12-126 Implantable Medical Devices
                                        (see table 1). 46 Information from the remaining respondents did not
                                        adequately reference the IMD by model, provide the sale price net of
                                        discounts and rebates, or both.

Table 1: Type of Detailed Information Reported on Cardiac and Orthopedic IMDs, Fiscal Year 2010

                                                                                    Price data provided indicated
                                                                      Both specific
                                                                    model and sale                           Sale price net of
                                          Number of          price net of discounts           Specific        discounts and            Other
                                                                                                                                           d
IMD          Respondent                 respondents                     and rebates         model only           rebates only          data
Cardiac      Hospitals                                 39                            31                 8                    0                 0
             GPOs                                       5                              1                4                    0                 0
             DOD Medical Centers or                     3                              2                1                    0                 0
                      a
             Hospitals
             VA Health Care System                      1                              0                0                    0                 1
                        b
Orthopedic   Hospitals                                 38                            14                 2                  20                  2
                    c
             GPOs                                       2                              0                2                    0                 0
             DOD Medical Centers or                     8                              2                1                    1                 4
             Hospitals
             VA Health Care System                      1                              0                0                    0                 1
                                        Source: GAO.

                                        Note: In some cases, hospitals and DOD medical centers or hospitals reported detailed pricing or
                                        model information for some devices, but not others. We considered an entity to have reported
                                        detailed information on specific model or sale price net of discounts and rebates if we could
                                        accurately identify the model or price—including all price discounts and rebates—for at least one of
                                        the devices reported by the entity.
                                        a
                                         Three of the eight DOD medical centers or hospitals that reported IMD pricing information included
                                        pricing information for cardiac devices. Some of these medical centers or hospitals indicated that
                                        cardiac procedures were consolidated in certain medical centers, while orthopedic procedures were
                                        performed more widely. In addition to the eight Air Force and Navy medical centers or hospitals that
                                        provided IMD pricing information, the Army also provided information for the highest volume
                                        purchaser of each type of implantable medical device. However, these data were excluded from our
                                        analysis because they may not represent the Army’s range of prices—both high and low—that other
                                        hospitals were able to provide.
                                        b
                                            One hospital did not report information on orthopedic device prices.
                                        c
                                         Information provided by GPOs reflected the experience of their member hospitals. Although we did
                                        not request model numbers for orthopedic devices from GPOs, two GPOs provided this information.




                                        46
                                          Eight Air Force and Navy medical centers or hospitals provided IMD price information.
                                        The Army also provided information for the highest volume purchaser of each implantable
                                        medical device model. However, these data were excluded from our analysis because
                                        they may not represent the Army’s range of prices—both high and low—that other
                                        hospitals were able to provide.




                                        Page 23                                                     GAO-12-126 Implantable Medical Devices
d
 The other data category includes respondents that provided information on cardiac or orthopedic
IMD prices, but which did not include information on the specific model or sale price net of discounts
and rebates for the device.


The respondents included in our study identified several reasons why
detailed information on the model-specific sale price net of discounts and
rebates is limited. First, many respondents indicated that the price
information they provided for at least one device did not account for all
discounts and rebates obtained. This issue was cited by 21 hospitals,
4 GPOs, and 1 DOD medical center. Officials from 1 hospital that
reported receiving a rebate from a manufacturer for meeting certain
volume targets stated that it would be difficult and time-consuming to
determine the proportion of the rebate to attribute to each product.
Officials from other hospitals and a DOD medical center noted that,
because discounts were negotiated at the hospital system level, they
lacked information on the discount amount. Similarly, 4 of the 5 GPOs
told us the prices they provided from their member hospitals did not
account for discount and rebate amounts because member hospitals did
not necessarily report that information. 47

Second, a number of entities stated that the majority of contracts between
hospitals and manufacturers include confidentiality clauses that generally
limit hospitals from sharing price information with third parties.
Specifically, all of the five manufacturers, five GPOs, three hospital
systems, and one of the two small hospitals we interviewed reported that
all or almost all of their contracts included language that restricted
information disclosure to some extent. For example, some confidentiality
clauses allow hospitals to share pricing information with other hospitals
that are part of their system, and others do not. Among our respondents,
four hospitals and one GPO stated that confidentiality clauses limited
them from providing us with detailed pricing information for devices from
certain manufacturers. Another GPO indicated that confidentiality clauses
precluded some member hospitals from reporting pricing information to
them. The extent to which confidentiality clauses may have affected
responses to our request for pricing information is uncertain.




47
  Officials from one GPO stated that even though their price data do not take rebates or
other discounts into consideration, the data provide a valuable and reasonably accurate
understanding of the range of prices being paid in the marketplace.




Page 24                                                   GAO-12-126 Implantable Medical Devices
                              Finally, some respondents cited data retrieval or quality issues as limiting
                              their ability to provide detailed price information. One hospital official
                              reported that the components used for total knees and hips are recorded
                              on paper invoices rather than in an electronic data system and that
                              additional staff would be required to retrieve the hand-written component
                              information. We also found data quality issues with the pricing information
                              provided by VA and DOD. Specifically, in the VA database, product
                              identifiers were entered in an inconsistent fashion or did not match,
                              making it difficult to aggregate the data in a meaningful way. Among the
                              issues identified with pricing information submitted by DOD, one hospital
                              reported it could not calculate the price for primary total knees, and
                              instead provided us with a range of prices per manufacturer. In addition, a
                              DOD medical center only provided estimates rather than actual prices it
                              paid for primary total knees.

                              Even when respondents report detailed information on the prices
                              hospitals pay for IMDs, differences in how hospitals purchase the parts or
                              components needed for certain procedures may limit the comparability of
                              the information. Specifically, the parts or components needed for a certain
                              cardiac and orthopedic procedure may be purchased together or
                              separately. For example, some hospitals purchase AICDs and CRT-Ds
                              with leads, while other hospitals purchase the leads separately from the
                              device. Given these differences, prices for IMDs that include leads—
                              which can cost up to several thousand dollars—cannot be meaningfully
                              compared to prices that exclude leads. Similarly, some hospitals
                              purchase the components used for a primary total knee or primary hip
                              implant together, while other hospitals purchase each component
                              separately. As a result, some hospitals reported one product number that
                              reflects multiple components, while others reported separate product
                              numbers for each component used, making it difficult to compare the
                              prices for devices purchased as whole to devices purchased component
                              by component.


Information from Study        Data from those respondents that provided detailed information on the
Respondents Indicates         specific model purchased and sale price net of discounts and rebates
Substantial Price Variation   indicated substantial variation in the prices hospitals paid for the same
                              device model. Among the 31 hospitals that provided detailed price
for Some IMDs                 information on cardiac IMDs, we found that for AICDs and CRT-Ds the
                              difference between the reported lowest and highest model-specific prices




                              Page 25                                   GAO-12-126 Implantable Medical Devices
                                            was several thousand dollars. 48 For example, the difference between the
                                            lowest and highest price hospitals reported paying for a particular AICD
                                            model was $6,844. The difference between the highest and lowest price
                                            reported for another AICD model was $8,723. The price differences for
                                            the remaining two AICD models in our study fell in between $6,844 and
                                            $8,723. (See table 2.) The median prices across the four AICD models
                                            ranged from $16,445 to $19,007. 49 The other models of cardiac devices
                                            for which we collected pricing data also showed substantial variation in
                                            prices.

Table 2: Median Prices and Differences between the Lowest and Highest Prices Hospitals Paid for Cardiac Devices, Fiscal
Year 2010

                                                                                                                               Range of differences
                                                                                                                                between the lowest
                                                                                                           Range of median      and highest model-
Type of device                                                            Number of models                           prices          specific price
Drug-eluting coronary stents                                                                          5     $1,700 to $1,800           $309 to $828
                                                           a
Automated implantable cardioverter defibrillator (AICD)                                               4   $16,445 to $19,007       $6,844 to $8,723
                                                           a
Cardiac resynchronization therapy defibrillator (CRT-D)                                               5   $19,370 to $22,603       $3,035 to $9,247
                                            Source: GAO analysis of hospital-reported pricing data.

                                            Notes: Results are based on implantable medical device pricing data submitted by 31 hospitals and
                                            include all discounts and rebates the hospital reported receiving. We limited our analysis to the most
                                            common devices for which these hospitals provided pricing data. Not all hospitals provided pricing
                                            information for each device model in our study; the number of hospitals providing sale price net of
                                            discounts and rebates for a particular device ranged from 3 to 19. Hospitals that contracted with
                                            GPOs may have obtained other financial benefits not reflected in the net sale price, such as a portion
                                            of the administrative fees GPOs received from manufacturers.
                                            a
                                             AICD or CRT-D prices that included leads were omitted from this analysis.




                                            48
                                              Two DOD medical centers reported prices for cardiac IMDs that were generally greater
                                            than the prices paid by the 31 hospitals that provided detailed price and model
                                            information. One DOD medical center reported detailed pricing information for all three
                                            types of cardiac devices, including four drug-eluting cardiac stent models, one AICD
                                            model, and one CRT-D model. For four of the six device models, the price paid by the
                                            medical center exceeded the highest price hospitals paid for the same device model by as
                                            much as $3,530. For the other two device models, the prices paid by the medical center
                                            were less than $100 below the highest price paid by those hospitals. The other DOD
                                            medical center reported a price for one drug-eluting cardiac stent model that was $220
                                            more than the highest price paid by any of the hospitals that provided detailed pricing
                                            information for the same device.
                                            49
                                              It is possible that hospitals that contract with GPOs receive other financial benefits not
                                            reflected in sales price, such as a portion of the administrative fees GPOs receive from
                                            manufacturers.




                                            Page 26                                                           GAO-12-126 Implantable Medical Devices
Further analysis of the pricing data shows that the variation in prices paid
by hospitals for IMDs cannot be attributed exclusively to a few hospitals
that were outliers in terms of the prices they paid. Among the nine IMD
models included in our study for which at least 10 hospitals reported
prices, the prices at the 90th percentile exceed those at the 10th
percentile by about 14 percent to over 40 percent, depending on the
specific IMD model. On average, the price paid by the hospitals at the
90th percentile for the nine IMD models was about 31 percent greater
than the price paid by the hospitals at the 10th percentile. (See text box
for a hypothetical example of how hospitals can be affected by price
variation for IMDs.)

Hypothetical Example of How Price Variation for IMDs Can Impact Hospitals
The variation in prices hospitals pay for the same device can have a substantial impact
on the share of a hospital’s Medicare payment that is available to cover other expenses
and profit. For example, assume that one hospital pays about $14,500 for a specific
AICD model and another hospital pays about $21,500 for the same device model—a
difference of about $7,000. In addition, assume that both hospitals pay an additional
$3,000 for leads. In 2010, the MS-DRG payment for the implantation of an AICD was
                         a
approximately $32,500. Therefore, the hospital that paid $14,500 for the AICD model
would have a little less than half of its MS-DRG payment to allocate to other costs
associated with the IMD procedure, overhead, and profit. In contrast, the hospital that
paid $21,500 for the same device would have only about a quarter of its MS-DRG
payment available for these items.
a
 The MS-DRG payment amount may vary depending on geography and other factors.


It was more difficult to compare prices for orthopedic IMDs because of the
greater variation among device configurations. However, the data on
orthopedic IMDs reported by hospitals and GPOs—which may not
capture all discounts and rebates—provided some evidence of substantial
price variation. For instance, one hospital reported spending about $4,500
for a specific primary total hip construct in 2010. In comparison, a GPO
provided information showing that one of its members paid about $8,000
for the same device construct, or 78 percent more. Similarly, a GPO
provided data on two of its member hospitals that purchased the same
primary total knee construct. One hospital paid about $5,200 for the




Page 27                                           GAO-12-126 Implantable Medical Devices
                             primary total knee construct, while the other hospital paid about $9,500,
                             or 83 percent more. 50


Hospitals’ Ability to        The influence of physicians on hospitals’ IMD purchasing decisions is a
Manage Relationships with    factor particular to the IMD market that affects prices hospitals pay. While
Physicians Is a Particular   physicians generally are not involved in price negotiations, they often
                             express strong preferences for certain manufacturers and models of
Factor That Influences       IMDs, known as physician preference items. 51 Physicians may seek
Prices Paid for IMDs         certain products in which they have high confidence in terms of their
                             value to patient care and reliability, after promotion by device makers, or
                             for other reasons noted below. To the extent that different physicians in
                             the same hospital have preferences for IMDs made by different
                             manufacturers, it may be difficult for the hospital to obtain volume
                             discounts by consolidating its purchases from particular manufacturers.
                             That is, individual hospitals or hospital systems may have less bargaining
                             power relative to the few cardiac and orthopedic IMD manufacturers. In
                             contrast, if the physician-hospital relationship is strong, the GPO can
                             consider purchasing from multiple IMD suppliers and thus, gain leverage
                             in negotiations. This could result in lower prices for hospitals.

                             Hospitals face a number of obstacles when it comes to working with
                             physicians to consider cost when making IMD decisions. First,
                             confidentiality clauses included in IMD contracts may bar hospitals from
                             sharing price information with some physicians, making it difficult to get
                             physicians the information they need to consider cost when making
                             decisions about devices. According to one GPO, some hospitals
                             restricted by confidentiality clauses have resorted to using colored



                             50
                               GPOs provided information on the sale prices its members reported paying for specific
                             IMDs, but did not confirm that the reported price was net of all discounts and rebates. Two
                             hospitals that provided information on the prices they paid for orthopedic devices also
                             provided information on the discounts and rebates they received from specific
                             manufacturers during our study period. The highest discount and rebate amount reported
                             by these hospitals averaged about 5 percent for one device model. If we applied these
                             discounts and rebates to the higher of the two prices in our orthopedic device examples,
                             the difference between the prices would still be substantial—about 69 percent for the hip
                             example and about 74 percent for the knee example.
                             51
                               Physicians tend to use devices developed by one particular manufacturer, often due to
                             having trained in a hospital that used that manufacturer’s products or past experience with
                             a manufacturer’s products, and may be reluctant to switch manufacturers and learn how to
                             implant another manufacturer’s device using a new set of instruments.




                             Page 28                                            GAO-12-126 Implantable Medical Devices
               stickers to indicate to physicians which devices are the high, medium, and
               low-cost options. Second, physician-hospital relationships regarding
               preference items frequently must contend with strong physician-
               manufacturer relationships. According to another GPO, physicians rely on
               manufacturer representatives to provide technical support during
               procedures involving IMDs, including setting up the operating room,
               consulting with the physician about the procedure, and programming
               devices, such as AICDs and CRT-Ds. In addition, some physicians might
               be loyal to certain manufacturers with whom they have consulting or
               professional relationships.

               Some GPOs noted that, despite the financial impact of physician
               preference, hospitals are likely to accommodate physicians’ IMD
               preferences in order to retain patient referrals to their facilities. One GPO
               commented that hospitals that ask physicians to be sensitive to IMD
               prices have encountered push back from some physicians. Another GPO
               cited a market where hospitals’ attempts at cost-savings led some
               physicians to migrate away from their facilities. A third GPO said that
               because physicians are highly dependent on their relationship with
               manufacturer representatives who provide the physicians with technical
               support, they may not be willing to switch device models. An association
               representing orthopedic surgeons told us that if physicians are not
               involved in the negotiation process, it is difficult to encourage them to
               switch devices. If manufacturers determine that a physician is unwilling to
               switch device models, they can be more aggressive in negotiations, which
               could result in higher prices for hospitals.

               In addition to hospitals’ ability to manage relationships with physicians,
               device manufacturers we interviewed noted several other factors that
               influence the prices hospitals pay for IMDs. They pointed to marketplace
               dynamics—the degree of competition within a local market and the
               market power of hospitals purchasing the devices—as key influences.
               Additionally, they noted that the support offered by manufacturers, such
               as device servicing agreements and training, as well as the terms and
               length of the contract itself, play a role in price negotiations. Finally, the
               manufacturers told us that the extent to which changes in device
               technology improve patient care affects what hospitals pay for IMDs.


               The lack of price transparency for the IMDs we examined makes it difficult
Concluding     to know whether hospitals are achieving the best device prices. This lack
Observations   of price transparency may have implications for Medicare because
               excess or unnecessary IMD costs that hospitals incur may be passed on


               Page 29                                     GAO-12-126 Implantable Medical Devices
                      to the Medicare program. In 2009, Medicare spent over $19 billion for
                      hospital procedures involving IMDs. A substantial portion of this amount
                      may be attributable to the cost of the devices themselves, but exactly how
                      much is unknown, in part, because hospitals purchase the IMDs and
                      Medicare does not track IMD prices or how much individual hospitals pay
                      for them.

                      Although Medicare’s payment approach provides hospitals with an
                      incentive to seek the best price on IMDs, hospitals may vary in their ability
                      to achieve the best price because of limited price information and
                      bargaining power. While we were able to obtain detailed IMD pricing data
                      from 31 of the 60 hospitals we contacted, the effort revealed the
                      challenges in compiling and analyzing meaningful price information even
                      from this relatively small number of hospitals. Furthermore, we observed
                      substantial variation in the prices that these 31 hospitals paid for cardiac
                      devices. Some hospitals paid several thousand dollars more than other
                      hospitals paid for the exact same device produced by the same
                      manufacturer. These data suggest that some hospitals have substantially
                      less bargaining power with the small group of companies that
                      manufacture particular IMD devices and consequently face challenges in
                      obtaining more favorable prices.

                      Physician preferences for particular manufacturer’s devices and models
                      may further complicate hospitals’ bargaining power. Such preferences
                      may shape hospitals’ purchasing decisions and limit their ability to obtain
                      volume discounts from device manufacturers. Moreover, many device
                      manufacturers require confidentiality clauses that prohibit hospitals from
                      disclosing their negotiated prices with third parties, which may include
                      physicians. A hospital that is constrained in sharing price data with its
                      physicians loses an opportunity to enlist their assistance in the hospital’s
                      efforts to be a prudent purchaser of IMDs.


                      The Department of Health and Human Services and VA reviewed a draft
Agency and Industry   of this report and provided technical comments, which we incorporated as
Comments              appropriate. DOD also reviewed a draft of this report and had no
                      comments.

                      Representatives from AdvaMed, the trade association representing
                      device manufacturers, reviewed relevant portions of the draft and
                      commented that there are many other factors—in addition to the hospital-
                      physician relationship—that influence IMD prices. AdvaMed
                      representatives noted that confidentiality clauses are not unique to the


                      Page 30                                    GAO-12-126 Implantable Medical Devices
IMD market. AdvaMed representatives also provided technical comments,
which we incorporated as appropriate.

A representative from the American Hospital Association (AHA) also
reviewed relevant portions of the draft and stated that confidentiality
clauses that restrict hospitals from informing physicians about IMD prices
inhibit hospitals from fully integrating care and making informed, cost-
conscious decisions. Furthermore, the AHA representative noted that
wide variation in IMD prices reported by hospitals shows that much of
device price is driven by negotiations; therefore, lack of IMD price
transparency puts hospitals at a disadvantage. The AHA representative
also provided technical comments, which we incorporated as appropriate.


As we agreed with your offices, unless you publicly announce the
contents of this report earlier, we plan no further distribution of it until
30 days from its date. We are sending copies of this report to the
Secretary of Health and Human Services, the Secretary of Defense, the
Secretary of Veterans Affairs and other interested parties. The report will
also be available at no charge on our website at http://www.gao.gov.

If you or your staffs have any questions about this report, please contact
me at (202) 512-7114 or cosgrovej@gao.gov. 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 I.

Sincerely yours,




James C. Cosgrove
Director, Health Care




Page 31                                    GAO-12-126 Implantable Medical Devices
Appendix I: GAO Contact and Staff
                  Appendix I: GAO Contact and Staff
                  Acknowledgments



Acknowledgments

                  James C. Cosgrove, (202) 512-7114 or cosgrovej@gao.gov
GAO Contact
                  In addition to the contact named above, individuals making key
Acknowledgments   contributions to this report include Jessica Farb, Assistant Director;
                  Gregory Giusto; Luis Serna III; and Ann Tynan. George Bogart and
                  Krister P. Friday also provided valuable assistance.




(290857)
                  Page 32                                    GAO-12-126 Implantable Medical Devices
                      The Government Accountability Office, the audit, evaluation, and
GAO’s Mission         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   The fastest and easiest way to obtain copies of GAO documents at no
                      cost is through GAO’s website (www.gao.gov). Each weekday afternoon,
GAO Reports and       GAO posts on its website newly released reports, testimony, and
Testimony             correspondence. To have GAO e-mail you a list of newly posted products,
                      go to www.gao.gov and select “E-mail Updates.”

Order by Phone        The price of each GAO publication reflects GAO’s actual cost of
                      production and distribution and depends on the number of pages in the
                      publication and whether the publication is printed in color or black and
                      white. Pricing and ordering information is posted on GAO’s website,
                      http://www.gao.gov/ordering.htm.
                      Place orders by calling (202) 512-6000, toll free (866) 801-7077, or
                      TDD (202) 512-2537.
                      Orders may be paid for using American Express, Discover Card,
                      MasterCard, Visa, check, or money order. Call for additional information.

Connect with GAO      Connect with GAO on Facebook, Flickr, Twitter, and YouTube.
                      Subscribe to our RSS Feeds or E-mail Updates. Listen to our Podcasts.
                      Visit GAO on the web at www.gao.gov.

To Report Fraud,      Contact:

Waste, and Abuse in   Website: www.gao.gov/fraudnet/fraudnet.htm
                      E-mail: fraudnet@gao.gov
Federal Programs      Automated answering system: (800) 424-5454 or (202) 512-7470

Congressional         Katherine Siggerud, Managing Director, siggerudk@gao.gov, (202) 512-
                      4400, U.S. Government Accountability Office, 441 G Street NW, Room
Relations             7125, Washington, DC 20548

Public Affairs        Chuck Young, Managing Director, youngc1@gao.gov, (202) 512-4800
                      U.S. Government Accountability Office, 441 G Street NW, Room 7149
                      Washington, DC 20548




                        Please Print on Recycled Paper.