Medicare: Payments for Covered Outpatient Drugs Exceed Providers'
Costs (21-SEP-01, GAO-01-1118). 				 
								 
Although physicians and other health care providers acknowledge  
that they can purchase drugs for prices lower than Medicare	 
payments, they contend that they need drug payments in excess of 
their actual costs to compensate for inadequate or nonexistent	 
Medicare payments for administrating the drugs. Physicians are	 
able to obtain Medicare-covered drugs at prices significantly	 
below current Medicare payments, which are set at 95 percent of  
average wholesale prices (AWP). Wholesalers' and group purchasing
organizations' prices that would be generally available to	 
physicians were considerably less than AWPs used to establish the
Medicare payment for these drugs. The difference between these	 
prices and AWP for physician-administered drugs in GAO's sample  
varied by drug. For most physician-administered drugs, the	 
average discount from AWP ranged from 13 percent to 34 percent;  
two physician-administered drugs had discounts of 65 percent and 
86 percent. Other suppliers are also able to purchase drugs at	 
prices that are considerably less than the AWP used to establish 
the applicable Medicare payment. Pharmacy suppliers were	 
predominant billers for 10 of the high-expenditure and		 
high-volume Medicare-covered drugs GAO analyzed. In general,	 
these suppliers provide two types of drugs--drugs administered	 
through durable medical equipment (DME) and covered oral drugs,  
such as certain immunosuppressives. Also, suppliers generally	 
receive a payment from Medicare for DME and supplies. Private and
other public payers use differing payment methods for drugs and  
their administration. Private health plans use their		 
drug-purchase and patient volume to negotiate favorable prices	 
for drugs and physician and supplier services related to	 
supplying or delivering the drugs.  Other public payers also use 
their purchasing volume along with information about actual	 
transaction prices from private payers to lower their drug	 
payments.							 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-01-1118					        
    ACCNO:   A01949						        
  TITLE:     Medicare: Payments for Covered Outpatient Drugs Exceed   
Providers' Costs						 
     DATE:   09/21/2001 
  SUBJECT:   Health care cost control				 
	     Health care programs				 
	     Drugs						 
	     Federal Supply Schedule				 
	     Medicare Program					 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Testimony.                                               **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-01-1118
     
Report to Congressional Committees

United States General Accounting Office

GAO

September 2001 MEDICARE Payments for Covered Outpatient Drugs Exceed
Providers? Cost

GAO- 01- 1118

Page i GAO- 01- 1118 Medicare Drug Payments Letter 1

Results in Brief 4 Background 6 Physicians Can Obtain Drugs Covered Under
Part B for

Significantly Less Than Medicare Payments 11 Medicare Drug Payments Also
Substantially Higher Than Costs for

Pharmacy Supplier- Billed Drugs 16 Private Payers Negotiate Prices, Other
Public Payers Rely on

Market- Based Benchmarks to Establish Payments for Drugs 21 Conclusions 24
Recommendations for Executive Action 26 Agency Comments 26

Appendix I Scope and Methodology 28

Appendix II Comments From the Centers for Medicare and Medicaid Services 30

Tables

Table 1: Medicare Part B Drugs by Share of Total Medicare Drug Spending,
1999 7 Table 2: Medicare Part B Drugs by Share of Total Medicare Units,

1999 8 Table 3: Medicare Part B Drug Spending and Volume by Place of

Service, 1999 9 Table 4: Widely Available Discounts From AWP for
MedicareCovered Drugs Billed Primarily by Physicians, 2001 12 Table 5:
Discounts From AWP Obtained by Physicians Who Billed

Medicare for a Low Volume of Selected Drugs, Compared to Widely Available
Discounts, 2001 14 Table 6: Discounts From AWP for Medicare- Covered Drugs
Billed

Primarily by DME Pharmacy Suppliers, 2001 18 Contents

Page ii GAO- 01- 1118 Medicare Drug Payments Abbreviation

AMP Average Manufacturer Price AWP Average Wholesale Price BBRA Balanced
Budget Refinements Act of 1999 BESS Part B Extract and Summary System BIPA
Medicare, Medicaid, and SCHIP Benefits Improvement and

Protection Act of 2000 CMS Centers for Medicare and Medicaid Services DME
Durable Medical Equipment DOD Department of Defense DOJ Department of
Justice ESRD End- Stage Renal Disease FSS Federal Supply Schedule GPO Group
Purchasing Organization HCFA Health Care Financing Administration HCPCS HCFA
Common Procedure Coding System HHS Health and Human Services NDC National
Drug Code OBRA Omnibus Budget Reconciliation Act OIG Office of the Inspector
General PBM Pharmacy Benefits Manager PHS Public Health Service VA
Department of Veterans? Affairs WAC Wholesale Acquisition Cost

Page 1 GAO- 01- 1118 Medicare Drug Payments

September 21, 2001 Congressional Committees While Medicare does not have a
comprehensive outpatient drug benefit, certain drugs and biologicals 1 are
covered under part B of the program. 2 In general, drugs are covered if they
cannot be self- administered and are related to a physician?s services, such
as cancer chemotherapy, or are provided in conjunction with covered durable
medical equipment (DME), such as inhalation drugs used with a nebulizer. 3
In addition, Medicare covers selected immunizations and certain drugs that
can be selfadministered, such as blood clotting factors and some oral drugs
used in association with cancer treatment and immunosuppressive therapy.
Medicare spending for these drugs totaled almost $4 billion in 1999. 4

Recent reports by the Department of Health and Human Services? (HHS) Office
of Inspector General (OIG), the Department of Justice (DOJ), and the House
Committee on Commerce found that in some cases Medicare?s payments for part
B- covered drugs were significantly higher than physicians? and other
providers? actual costs of acquiring these products. They found that the
average wholesale prices (AWP)- the ?list prices? or

?sticker prices? set by drug manufacturers and used by Medicare to calculate
payment rates- were not representative of the actual costs of these drugs to
providers. The Health Care Financing Administration (HCFA) 5 initiated steps
to reduce Medicare payment for some drugs. The agency issued a program
memorandum in September 2000 that authorized

1 For the remainder of this report, we will refer to part B- covered drugs
and biological products, such as clotting factors used to treat hemophilia,
as ?drugs.?

2 Medicare part B or Supplementary Medical Insurance provides coverage for
certain physician, outpatient hospital, laboratory, and other services to
beneficiaries who pay monthly premiums.

3 A nebulizer is a device driven by a compressed air machine. It allows the
patient to take medicine in the form of a mist (wet aerosol). It consists of
a cup, a mouthpiece attached to a T- shaped part or a mask, and thin,
plastic tubing to connect to the compressed air machine.

4 The Medicare program payment is 80 percent of the total payment, or $3.2
billion; beneficiaries are responsible for the remaining 20 percent. 5 HCFA
was renamed the Centers for Medicare and Medicaid Services in June 2001; we
continue to refer to the agency as HCFA when discussing actions it took
under that name.

United States General Accounting Office Washington, DC 20548

Page 2 GAO- 01- 1118 Medicare Drug Payments

the use of prices obtained in DOJ?s investigations of providers? drug
acquisition costs to set Medicare payment rates for a subset of covered
drugs, which would have lowered Medicare payment for these products.
However, HCFA retracted the original program memorandum in November 2000.

While physicians- particularly oncologists- and other health care providers
such as pharmacy suppliers acknowledge that they can purchase drugs for
prices lower than Medicare payments, they contend that they need drug
payments in excess of their actual costs to compensate for inadequate or
nonexistent Medicare payments for administering the drugs. Specifically,
oncologists argue that Medicare?s physician payment system does not
adequately reimburse them for their costs of administering chemotherapy.
Further, they suggest that some providers, particularly those who purchase
drugs in low volume, may not have access to low drug prices, because the
lowest prices reflect volume discounts and other factors. They assert that
beneficiary access could be impaired if Medicare drug payments are reduced.

In a forthcoming report mandated by the Balanced Budget Refinements Act of
1999 (BBRA), 6 we examine the adequacy of Medicare?s physician payments to
oncologists. 7 We conclude that the basic methodology HCFA used to establish
the fees is sound. We found that payments to oncologists are 8 percent
higher than they would have been if the prior charge- based method had been
maintained. We also determined that oncologists? payments relative to their
estimated practice expenses are close to the average for all specialties.
However, we identify modifications to the basic method that substantially
lowered payments for certain services, including most chemotherapy
administration. In our report, we make recommendations to improve Medicare?s
physician payment system.

The Medicare, Medicaid, and SCHIP Benefits Improvement and Protection Act of
2000 (BIPA), 8 required that we determine whether Medicare is reimbursing
physicians and other providers adequately for covered drugs and related
services. It also stipulated that the Medicare program make no

6 P. L. 106- 113, Appendix F, Sec. 213( a), 113 Stat. 1150, 1501A- 350. 7
See our forthcoming report Medicare Physician Fee Schedule: Practice Expense
Payments to Oncologists Indicate Need for Overall Refinements (GAO- 02- 53,
October 2001).

8 P. L. 106- 554, Appendix F, Sec. 429, 114 Stat. 2763, 2763A- 522.

Page 3 GAO- 01- 1118 Medicare Drug Payments

reductions to drug payments until we complete our work. As agreed with the
committees of jurisdiction, our analysis addressed the following questions:
(1) What are physicians? costs of providing drugs relative to applicable
Medicare payments? (2) What are other providers? costs of providing drugs
relative to applicable Medicare payments? (3) What are the methods that
public and private payers use to establish payments for providing these
drugs? This legislation also directed that we take into account the results
of the study of oncology physician payments mandated by BBRA in evaluating
drug acquisition costs relative to their Medicare payments.

To answer these questions, we interviewed officials at the Centers for
Medicare and Medicaid Services (CMS), the Department of Veterans Affairs
(VA), HHS OIG, and DOJ. We also interviewed and solicited information about
drug acquisition costs and costs related to drug administration from
professional associations representing physicians and nonphysician
providers, such as DME or pharmacy suppliers. We used pricing data from
pharmaceutical wholesalers, group purchasing organizations (GPO), specialty
pharmacies, DME pharmacy suppliers, and physicians, including several
physicians we identified as billing for low volumes of Medicare- covered
drugs. We attempted to obtain pricing information directly from additional
major wholesalers and group purchasing organizations, but these entities did
not provide the information we requested by the time we finalized this
report. We interviewed representatives from large managed care organizations
regarding the methods they used to purchase or pay for pharmaceuticals, as
well as the purchase prices they were able to obtain. We also interviewed
staff members at the companies that publish the major price reporting
compendia on average wholesale prices for drugs.

We conducted quantitative analyses using data from HCFA and other sources.
We used 1999 drug claims data from HCFA to determine which covered drugs
accounted for the most spending and volume. For detailed analysis, we
selected the 20 drugs with the highest total Medicare expenditures and the
20 drugs with the highest volume; the results of our analysis indicated that
35 different drugs accounted for 82 percent of total Medicare drug spending
and 95 percent of total drug units. 9 We obtained prices for 31 of the 35
individual drugs. We did our work in accordance with generally accepted
government auditing standards from January

9 Volume for a drug is measured in terms of the number of Medicare- allowed
services.

Page 4 GAO- 01- 1118 Medicare Drug Payments

through August 2001. A detailed discussion of our scope and methodology is
in appendix I.

Physicians are able to obtain Medicare- covered drugs at prices
significantly below current Medicare payments, which are set at 95 percent
of AWP. Wholesalers? and GPOs? prices that would be generally available to
physicians were considerably less than AWPs used to establish the Medicare
payment for these drugs. The difference between these prices and AWP for
physician- administered drugs in our sample varied by drug. For most
physician- administered drugs, the average discount from AWP ranged from 13
percent to 34 percent; two physician- administered drugs had discounts of 65
percent and 86 percent. Our survey of physicians who billed Medicare for low
volumes of drugs used in cancer treatment indicated they received discounts
that were as large as or larger than widely available discounts for 11 of
the 16 products for which they were able to provide price information.
Physicians are reimbursed under the physician fee schedule for the costs of
administering chemotherapy drugs, which account for most of Medicare?s drug
spending. HCFA deviated from the basic methodology for determining practice
expense payments for certain services, including chemotherapy administration
by nonphysicians, which reduced Medicare?s practice expense payment for most
chemotherapy administration services. However, even with this alternative
methodology, oncologists? average practice expense payments in 2001 are 8
percent higher than what they would have been had chargebased payments
continued.

Other suppliers are also able to purchase drugs at prices that are
considerably less than the AWP used to establish the applicable Medicare
payment. Pharmacy suppliers were the predominant billers for 10 of the high-
expenditure and high- volume Medicare- covered drugs we analyzed. In
general, these suppliers provide two types of drugs- drugs administered
through DME and covered oral drugs, such as certain immunosuppressives.
Widely available prices in 2001 reflected average discounts of 78 percent
from the AWP for ipratropium bromide and 85 percent for albuterol, two DME-
delivered inhalation therapy drugs that account for most of Medicare
payments to pharmacy suppliers. Medicare pays a monthly fee to pharmacy
suppliers for dispensing these and certain other respiratory therapy drugs.
Also, suppliers generally receive a payment from Medicare for DME and
supplies. Although there has been no recent analysis of the adequacy of
Medicare?s DME payments, there are indications that the payments may be
above current market rates. Wholesaler and GPO prices for two of the high-
volume oral drugs averaged Results in Brief

Page 5 GAO- 01- 1118 Medicare Drug Payments

14 percent and 77 percent below AWP. Medicare makes no separate payments for
costs associated with supplying or administering oral drugs.

Private and other public payers use differing payment methods for drugs and
their administration. Private health plans use their drug- purchase and
patient volume to negotiate favorable prices for drugs and the physician and
supplier services related to supplying or delivering the drugs. Other public
payers also use their purchasing volume along with information about actual
transaction prices from private payers to lower their drug payments. VA and
certain other government agencies use the Federal Supply Schedule (FSS)
prices for drugs. 10 State Medicaid programs reimburse for drugs using
formulas based on standard price lists but subsequently receive rebates from
the manufacturers calculated using the average manufacturer price (AMP), to
substantially lower their net prices for drugs. 11 Both the FSS and the AMP
are derived from actual market transaction data reported by drug
manufacturers. In limited instances, VA also uses competitive bidding
approaches to obtain lower drug prices, an approach that Medicare has used
in limited demonstration projects.

We recommend that the Administrator of CMS take steps to begin reimbursing
providers for part B- covered drugs and related services at levels
reflecting providers? acquisition costs using information about actual
market transaction prices. We recommend that the CMS Administrator evaluate
expanding competitive bidding approaches to setting payment levels. We also
recommend that the CMS Administrator closely monitor beneficiary access to
covered drugs in light of any changes to reimbursement.

10 FSS prices are available to any federal agency that directly procures
pharmaceuticals, including VA medical centers, the Department of Defense
(DOD), the Bureau of Prisons, the Public Health Service (PHS), and other
designated entities such as the District of Columbia, U. S. territorial
governments, the Indian Health Service, and some state veterans homes.
Manufacturers must also sell brand- name drugs listed on the FSS to four
federal drug purchasers- VA, DOD, PHS, and the Coast Guard- at a price at
least 24 percent lower than the nonfederal average manufacturer price, a
ceiling price that is lower than the FSS price for many drugs.

11 State Medicaid programs generally pay pharmacies a dispensing fee for
each prescription and physicians a fee for administering the drugs.

Page 6 GAO- 01- 1118 Medicare Drug Payments

Although Medicare reimburses providers for roughly 450 unique drugs under
part B, a small number of products accounted for the majority of Medicare
spending and volume. 12 In 1999, the 20 highest expenditure drugs accounted
for 75 percent of total Medicare drug spending and the 20 highest volume
drugs accounted for 93 percent of total units. 13 (See tables 1 and 2.)
Combined, these two groups of drugs yielded 35 unique drugs, accounting for
82 percent of total drug spending and 95 percent of total drug units. 14

12 Each covered drug is identified by an alphanumeric code under the HCFA
Common Procedure Coding System (HCPCS), which specifies the drug name,
method of administration, and dosage.

13 Our analysis of drug acquisition costs excluded four high- volume and
high- expenditure drugs. Specifically, we excluded a code for ?not otherwise
classified antineoplastic drugs,? two antihemophilia clotting factors, and a
radiopharmaceutical. We could not collect acquisition cost data on ?not
otherwise classified antineoplastic drugs? because it does not refer to a
specific product. The clotting factors, which are typically billed by non-
physician suppliers, were not included for two reasons. First, we were
unable to obtain adequate pricing information for these products, a problem
also encountered by HHS OIG in its prior work on drug reimbursement. Second,
these products differ significantly from other pharmaceutical products
discussed in this report. Their source material is collected from human
donors; their manufacturing, storage, and distribution processes differ from
other products; and they are administered to a very small patient
population. The excluded radiopharmaceutical, Technetium TC Sestamibi, is
used by cardiologists in certain diagnostic imaging procedures. Since
Medicare only began requiring data on AWPs to reimburse for
radiopharmaceutical products in 2001, these data are currently being
developed.

14 Units are defined as the number of claims for each drug times the number
of units specified by its HCPCS label. Background

Page 7 GAO- 01- 1118 Medicare Drug Payments

Table 1: Medicare Part B Drugs by Share of Total Medicare Drug Spending,
1999 Drug name Share of total Medicare drug

spending (percentage)

Leuprolide acetate (for depot suspension) 15.1 Epoetin alpha for non- ESRD
use 9. 5 Goserelin acetate implant 7. 9 Ipratropium bromide, unit dose form
6.4 Albuterol, unit dose form 6.3 Paclitaxel 6.2 Carboplatin 2.9 Pamidronate
disodium 2.8 Irinotecan 2.0 Gemcitabine HCl 1.9 Rituximab 1.8 Filgrastim (G-
CSF) 480 mcg 1. 7 Leucovorin calcium 1. 6 Docetaxel 1.5 Factor VIII
(antihemophilic factor, recombinant) 1. 3 Technetium TC Sestamibi 1. 2 Hylan
G- F 20 1.2 Filgrastim (G- CSF) 300 mcg 1. 2 Not otherwise classified
antineoplastic drugs 1.2 Dolasetron mesylate, injection 1. 2 Subtotal, 20
highest- expenditure drugs and biologicals 74.9 All other Medicare- covered
drugs and biologicals 25.1

Total 100.0

Source: GAO analysis of data from the Medicare Part B Extract and Summary
System (BESS).

Page 8 GAO- 01- 1118 Medicare Drug Payments

Table 2: Medicare Part B Drugs by Share of Total Medicare Units, 1999 Drug
name Share of total units (percentage)

Albuterol, unit dose form 65.8 Ipratropium bromide, unit dose form 8.2
Epoetin alpha for non- ESRD use 3. 4 Dolasetron mesylate, injection 3. 3
Albuterol, concentrated form 2.6 Mycophenolate mofetil, oral 1.7 Cromolyn
sodium, unit dose form 1.2 Heparin sodium 1.0 Cyclosporine, oral 0. 9
Ondansetron HCl, injection 0. 8 Tacrolimus, oral 0.7 Prednisone, oral 0.6
Acetylcysteine, unit dose form 0. 5 Botulinum toxin, type A 0. 5
Imiglucerase 0.4 Factor VIII (antihemophilic factor, human) 0.4
Dexamethasone sodium phosphate 0.4 Leucovorin calcium 0. 3 Saline solution,
sterile 0. 3 Granisetron HCl, injection 0. 3 Subtotal, 20 highest- volume
drugs and biologicals 93.3 All other Medicare- covered drugs and biologicals
6. 7

Total 100.0

Note: Units are defined as the number of Medicare- allowed services. Each
drug has a standard unit dosage specified by the HCPCS code.

Source: GAO analysis of data from BESS.

The drugs provided by physicians account for the largest share of Medicare
expenditures for drugs under part B, while billing volume is dominated by
the drugs provided by pharmacy suppliers. Drugs provided in the physician
office setting accounted for over 75 percent of Medicare spending for drugs
in 1999. (See table 3.) Three specialties, hematology/ oncology, medical
oncology, and urology, bill Medicare primarily for drugs used in the
treatment of cancer and represented 80 percent of total Medicare payments to
physicians for drugs. By contrast, pharmacy suppliers accounted for over 80
percent of Medicare drug billing volume and less than 20 percent of
corresponding payments. Two inhalation therapy drugs dominated these home-
administered products, accounting for 88 percent of Medicare volume in the
home setting. When the drug is delivered in a physician?s office, Medicare
makes a separate

Page 9 GAO- 01- 1118 Medicare Drug Payments

additional payment through the physician fee schedule for the physician or
his or her staff administering a drug. When the drug is administered via DME
in the home, Medicare pays separately for DME, the drug, and associated
supplies as well as a monthly dispensing fee for providing nebulizer drugs.

Table 3: Medicare Part B Drug Spending and Volume by Place of Service, 1999
Place of service Total spending

Share of total spending (percentage) Total units

Share of total units (percentage)

Physician office $3,021,662,605 76.2 142,247,564 14.9 Home 727,559,447 18.3
759,461,862 79.7 Other a 218,297,305 5. 5 51,770,956 5. 4

Total $3,967,519,357 100.0 953,480,382 100.0

a ?Other? includes, for example, immunization centers, end- stage renal
disease treatment facilities, and independent laboratories. Note: Medicare?s
payment is 80 percent of the allowed amount and the beneficiaries? share is
20 percent, after they have fulfilled their annual deductible requirements.
Units are defined as the number of Medicare- allowed services, as specified
by the HCPCS code.

Source: GAO analysis of data from BESS.

Medicare bases its reimbursement to physicians and other providers of drugs
on AWP, which is often described as a ?list price,? ?sticker price,? or

?suggested retail price,? reflecting the fact that AWP is not necessarily
the price paid by a purchaser or a consistently low or ?wholesale? price.
AWPs are published for each drug identified by a National Drug Code (NDC).
15 Manufacturers periodically report AWPs for NDCs to publishers of drug
pricing data, such as the Medical Economics Company, Inc., which publishes
the Red Book, or First Data Bank, which compiles the National Drug Data
File. Publishers of AWPs and other drug prices stated that they list the
prices as reported to them by the manufacturers. There is no required
frequency for manufacturers to report AWPs, but publishers said they attempt
to update AWPs at least annually. Medicare carriers, the contractors
responsible for paying part B claims, use published AWPs to

15 NDCs are the universal product identifiers for drugs for human use; the
Food and Drug Administration assigns the first part of the NDC, which
identifies the firm that manufactures, repackages, or distributes a drug.
Each NDC is specific to a chemical entity, dosage form, manufacturer,
strength, and package size. For example, a drug made by one manufacturer, in
one form and strength, but in three package sizes, would have three NDCs.
Medicare Payments for

Drugs Are Based on Published AWPs

Page 10 GAO- 01- 1118 Medicare Drug Payments

determine the Medicare- allowed amount, or payment level, which is 95
percent of AWP for each HCPCS- coded drug. Because one HCPCS code may have
multiple NDCs that match the HCPCS- coded product?s definition, the carriers
may determine the Medicare payment by analyzing multiple NDCs. 16

Pharmaceutical sales and distribution networks can involve multiple entities
and purchasing arrangements that affect the actual acquisition price of the
drug for the end purchaser that supplies it to a Medicare beneficiary.
Physicians and pharmacies can purchase the kinds of drugs covered under
Medicare part B from general or specialty pharmaceutical wholesalers or they
can have direct purchase agreements with manufacturers. Purchasers may
belong to GPOs that pool the purchasing of multiple entities and negotiate
prices with wholesalers or manufacturers. GPOs may negotiate different
prices for a drug for different end users, such as physicians, pharmacies,
or hospitals.

Determining physicians? or other providers? actual acquisition cost of drugs
is complicated by certain practices in the pharmaceutical marketplace that
may result in transaction prices paid at the time of sale that do not
reflect the final net cost to the purchaser. For example, manufacturers or
wholesalers may offer purchasers rebates based on the volume of products
purchased. In addition, manufacturers may establish

?chargeback? arrangements for end purchasers, under which the purchaser
negotiates a price with the manufacturer that is lower than the price the
wholesaler charges for the product. The wholesaler sells the product to the
purchaser for the lower price negotiated with the manufacturer, and then
asks the manufacturer to pay back the difference between the wholesaler?s
price and the negotiated price.

16 For single- source drugs (drugs whose manufacturer is the sole source for
a given product), Medicare?s payment is 95 percent of the drug?s AWP. For
multisource drugs (drugs with generic equivalents or drugs for which there
are two or more competing brandname products), the payment allowance is 95
percent of the lower of (1) the median AWP of all generic forms of the drug
or (2) the lowest brand- name product AWP. Within these guidelines, each
carrier contracting with Medicare to process claims has discretion to
determine which NDCs should be used to calculate the payment rate for each
HCPCS code. This can lead to variation in payment amounts among carriers for
the same HCPCS- coded drug. By October 1, 2002, all payers, including
Medicare, will be required to process and pay drug claims by NDC, rather
than HCPCS code, in compliance with the administrative simplification
provisions of the Health Insurance Portability and Accountability Act of
1996 (P. L. 104- 191). 65 Fed. Reg. 50312, 50370 (to be codified at 42 C. F.
R. 162.1002( c)). Drug Supply Chain

Involves Multiple Parties and Arrangements That Influence the Net Cost to
the End Purchaser

Page 11 GAO- 01- 1118 Medicare Drug Payments

Physicians are able to obtain drugs at prices significantly below current
Medicare reimbursements. The widely available prices available from
wholesalers and GPOs for physician- administered drugs we examined were
considerably less than AWPs used to establish the Medicare payment. For most
of the high- expenditure or high- volume physicianadministered drugs we
studied and for which we obtained price data, the average widely available
discounts from AWP ranged from 13 percent to 34 percent; but two drugs
exhibited considerably larger discounts. Physicians we identified as low-
volume billers for oncology drugs can also purchase drugs for considerably
less than Medicare?s payment. In addition to reimbursement for drugs,
physicians are paid separately for services associated with drug
administration under the Medicare physician fee schedule. In a separate
forthcoming report, we find that Medicare?s basic method for calculating
these payments is sound, but it deviated from this method in calculating
payments for certain services, including chemotherapy administration. 17 We
also find that oncologists? payments relative to their estimated practice
expenses are close to the average for all specialties. In that report, we
make recommendations to improve Medicare physician payments for all
services.

Among the 31 drugs we analyzed, which accounted for the majority of Medicare
spending and volume, 21 were provided almost exclusively by physicians.
Physicians providing drugs to treat cancer- the specialties of hematology/
oncology, medical oncology, and urology- account for most Medicare claims
for 19 of these drugs. 18 Oncologists and groups representing oncologists
told us that oncologists can purchase Medicarecovered drugs for less than
the Medicare payment amount. Price lists from wholesalers and GPOs show that
widely available prices were considerably less than AWPs used to establish
the Medicare payment for 18 of these physician- administered cancer drugs,
with price information unavailable for one other. (See table 4.) For 16 of
the cancer drugs, the average discount from AWP ranged from 13 percent to 34
percent, and two drugs exhibited considerably larger discounts. Prices were
available for only one of the two noncancer drugs generally provided by
physicians, and

17 See our forthcoming report Medicare Physician Fee Schedule: Practice
Expense Payments to Oncologists Indicate Need for Overall Refinements (GAO-
02- 53, October 2001). 18 Of the other high- volume/ high- expenditure drugs
we analyzed, orthopedic surgeons and neurologists were each the primary
billers for a single product. Physicians Can

Obtain Drugs Covered Under Part B for Significantly Less Than Medicare
Payments

Physicians? Drug Acquisition Costs Are Generally Lower Than AWP- Based
Medicare Payment

Page 12 GAO- 01- 1118 Medicare Drug Payments

its average discount from AWP was 18 percent. Certain purchasers may have
access to even greater discounts for certain products than the widely
available discounts we report here.

Table 4: Widely Available Discounts From AWP for Medicare- Covered Drugs
Billed Primarily by Physicians, 2001

Drug name Specialty most

frequently billing for drug Average

AWP a Average widely

available discount from

AWP (percentage) b

Leuprolide acetate (for depot suspension) urology $618.93 17.6 Rituximab
oncology c $478.47 19.2 Goserelin acetate implant urology $469.99 21.9
Docetaxel oncology $313.51 22.0 Filgrastim (G- CSF) 480 mcg oncology $300.40
18.0 d Pamidronate disodium oncology $279.86 16.8 Hylan G- F 20 orthopedic

surgery $225.13 17.7 d Filgrastim (G- CSF) 300mcg oncology $193.62 18.4 d
Paclitaxel oncology $180.57 19.0 Irinotecan oncology $141.32 22.9
Carboplatin oncology $120.48 20.3 Gemcitabine HCl oncology $112.34 21.3
Dolasetron mesylate, injection oncology $45.02 65.0 d Granisetron HCl,
injection oncology $19.52 29.3 Leucovorin calcium oncology $18.44 85.6
Epoetin alpha for non- ESRD use oncology $12.91 15.2 Ondansetron HCl,
injection oncology $6.41 12.8 Botulinum toxin type A neurology $4.86 n/ a e
Imiglucerase oncology $3.95 n/ a e Dexamethasone sodium phosphate oncology
$1.44 14.2 Heparin sodium oncology $0.43 34.4

a ?Average AWP? is the average of AWP of each NDC for that product adjusted
to the HCPCS- defined dosage. b ?Average widely available discount from AWP?
for each drug was calculated by (1) determining the

average widely available price( s) for each NDC for that drug, (2)
determining the percentage difference between the average widely available
price( s) and the AWP for each NDC for that drug, and (3) averaging the
percentage differences for all NDCs for that drug. c ?Oncology? specialty
includes hematology/ oncology and medical oncology.

Page 13 GAO- 01- 1118 Medicare Drug Payments

d ?Average widely available discount from AWP? in 2001 for this drug is
based on a price or prices from a single wholesaler. For these four drugs,
we had 2000 data from two or more sources. Those data showed that the
average widely available discount from AWP in 2000 was 18.8 percent for
Filgrastim (G- CSF) 480 mcg, 17.6 percent for Hylan G- F 20, 19.0 percent
for Filgrastim (G- CSF) 300mcg, and 42.2 percent for Dolasetron mesylate,
injection. e We were unable to obtain wholesaler or GPO prices for these
products.

Source: GAO analysis of data from BESS, the Medical Economics Drug Topics
Red Book CD- ROM vol. 21, and wholesaler and GPO price lists.

While physician practices that purchase large volumes of drugs may have
access to larger discounts and rebates, low- volume providers can also
purchase drugs for markedly less than AWP, and often at additional discounts
below widely available prices. 19 (See table 5.) Our survey of physicians
who billed Medicare for low volumes of cancer drugs indicated they received
discounts that were as large as or larger than widely available discounts
for 11 of the 16 products for which they were able to provide price
information, although these discounts may not be as high as those obtained
by higher- volume purchasers. These practices reported a variety of
arrangements for obtaining these drugs, including contracts with
manufacturers, wholesalers, or GPOs.

19 For example, physicians and other purchasers can obtain lower prices if
they make prompt payments.

Page 14 GAO- 01- 1118 Medicare Drug Payments

Table 5: Discounts From AWP Obtained by Physicians Who Billed Medicare for a
Low Volume of Selected Drugs, Compared to Widely Available Discounts, 2001

Drug name Low volume billers?

average discount from AWP (percentage)

Average widely available discount

from AWP a (percentage)

Leuprolide acetate (for depot suspension) 32.8 17.6 Rituximab 15.7 19.2
Goserelin acetate implant 22.3 b 21.9 Docetaxel 22.0 22.0 Filgrastim (G-
CSF) (480 mcg) 22.4 18.0c Pamidronate disodium 18.0 16.8 Filgrastim (G- CSF)
(300 mcg) 21.7 18.4 c Paclitaxel 25.8 19.0 Irinotecan 27.1 22.9 Carboplatin
20.0 b 20.3 Gemcitabine HCl 16.1 21.3 Dolasetron mesylate, injection 62.0
65.0 c Granisetron HCl, injection 28.1 29.3 Leucovorin calcium 90.4 85.6
Epoetin alpha for non- ESRD use 22.1 15.2 Ondansetron HCl, injection 26.4
12.8

a ?Average widely available discount from AWP? for each drug was calculated
by (1) determining the average widely available price( s) for each NDC for
that drug, (2) determining the percentage difference between the average
widely available price( s) and the AWP for each NDC for that drug, and (3)
averaging the percentage differences for all NDCs for that drug. b ?Low-
volume billers? average discount from AWP? for this drug is based on a price
from a single

physician. c ?Average widely available discount from AWP? for this drug is
based on a price or prices from a

single wholesaler. Notes: Out of our sample of 108 physicians, 14 provided
us with acquisition cost data for 16 of the 18 cancer treatment drugs we
examined. An additional 37 physicians belonged to large, hospital- based or
national chain oncology practices that likely had access to widely available
drug price discounts. Fifty- six physicians could not be contacted or
refused to participate. One physician in the sample did not purchase drugs.

Source: GAO telephone survey of a sample of physicians who billed Medicare
for a low volume of cancer drugs in 1999 and AWPs listed in a
contemporaneous wholesaler catalog.

In addition to the AWP- based payment for the drug, physicians receive a
payment based on Medicare?s physician fee schedule for administering the
drug. The physician fee schedule consists of three components: one to
account for physician time and effort to provide the service, one for the
expenses of operating the physician practice (including payments reflecting
the costs of nonphysician personnel involved in the delivery of services),
and one for malpractice expenses. The practice expense Medicare Pays
Physicians

for Drug Administration Through the Physician Fee Schedule

Page 15 GAO- 01- 1118 Medicare Drug Payments

component was developed with the best available physician- reported data on
the actual costs of running their practices. These costs were then allocated
to individual services that are provided and billed for, such as office
visits and medical procedures.

Oncologists have expressed concerns that certain costs necessary to
administer drugs are not fully reimbursed by the practice expense component
of Medicare?s physician fee schedule payment. Their representatives told us
that certain costs were not reflected in the data originally used to set
physician fee schedule payments for drug administration, as changes have
occurred in how services are delivered over time. Oncology representatives
also indicate concern about the representativeness of the data used to
construct the fee schedule. They believe the sample of physician practices
that supplied the practice expense data may have been biased by including
too many hospital- based oncologists or surgical oncologists, which would
understate the practice expenses for chemotherapy services. In addition, the
number of oncology practices in the sample is small, increasing the risk
that the sample information is not representative. 20

We have reported previously that the basic methodology for determining
physicians? practice expense payments is reasonable. 21 In a separate
forthcoming report, we examine the adequacy of these payments for
oncologists. We find that HCFA deviated from the basic methodology for
determining practice expense payments for certain services, including
chemotherapy administration by nonphysicians. The use of the alternative
methodology resulted in a significant reduction in chemotherapy
administration service payments compared to what payments would have been
had they been calculated under the basic methodology. However, even with
this alternative methodology, oncologists? average practice expense payments
in 2001 are 8 percent higher than what they would have been had charge-
based payments continued. Further, under the current practice expense
methodology, oncologists? payments relative to their estimated practice
expenses are close to the average payment for all

20 In another study, also mandated by the Medicare, Medicaid, and SCHIP
Benefits Improvement and Protection Act of 2000 (see P. L. 106- 554,
Appendix F., Sec. 411, 114 Stat. 2763, 2763A- 508), we are examining methods
CMS may use to collect and employ more current information to modify the
physician fee schedule to reflect potential changes in service delivery or
costs and more representative data on practice expenses.

21 See Medicare Physician Payments: Need to Refine Practice Expense Values
During Transition and Long Term (GAO/ HEHS- 99- 30, February 24, 1999).

Page 16 GAO- 01- 1118 Medicare Drug Payments

specialties. In the separate report, we make recommendations to improve the
appropriateness of Medicare physician fee schedule payments for all
services. 22

Pharmacy suppliers can also obtain drugs such as those administered through
the DME benefit at prices far lower than Medicare payment levels. Pharmacy
suppliers were the predominant billers for 10 of the drugs in our sample of
high- expenditure and high- volume drugs; these drugs included inhalation
therapy drugs administered through DME and oral immunosuppressive drugs.
Pharmacy suppliers may also administer infusion drugs, such as chemotherapy
drugs in the home setting. 23 For the inhalation therapy drugs, wholesaler
and GPO prices that would be generally available to these suppliers for
drugs administered through DME were considerably less than AWPs used to
establish the Medicare payment. Wholesaler or GPO price discount information
for two of the oral immunosuppressive drugs in our sample was not available,
but retail prices for these products from Internet pharmacies were generally
below Medicare payment levels. Medicare makes an additional payment for
dispensing drugs used in inhalation therapy, but not for drugs used in
infusion treatments. Further, Medicare makes a separate payment to DME
suppliers for the rental or purchase of the equipment needed to administer
the drug, such as nebulizers or infusion pumps, in addition to the payment
to the pharmacy supplier for the drug itself. There has been no recent
analysis of Medicare?s payments relative to the costs for DME used to
administer drugs.

22 See our forthcoming report Medicare Physician Fee Schedule: Practice
Expense Payments to Oncologists Indicate Need for Overall Refinements (GAO-
02- 53, October 2001). 23 Like inhalation therapy, infusion therapy, such as
therapy for heart failure, can be provided in the home setting using an
infusion pump. Infusion therapy involves injection of drugs over time, most
often intravenously; Medicare covers infusion pumps and necessary drugs and
supplies when it is medically necessary to administer the drug with a
durable infusion pump that regulates the rate of infusion. Infusion drugs
provided by pharmacy suppliers did not appear as high- volume or high-
expenditure products and make up a small percentage of pharmacy suppliers?
aggregate Medicare billing. The 10 drugs on our list provided primarily by
pharmacy suppliers accounted for 91 percent of their Medicare payments and
98 percent of their Medicare volume in 1999. Medicare Drug

Payments Also Substantially Higher Than Costs for Pharmacy SupplierBilled
Drugs

Page 17 GAO- 01- 1118 Medicare Drug Payments

Pharmacy suppliers were the dominant billers for 10 of the highest
expenditure or volume drugs paid for under Medicare part B. Among these 10
drugs were five inhalation therapy drugs and four oral immunosuppressive
drugs. 24 Prices available from wholesalers and GPOs were much less than the
AWPs used to establish the Medicare payment for the inhalation therapy
drugs. (See table 6.) The average discount from AWP in 2001 ranged from 69
percent to 85 percent. These results were consistent with prior studies of
the acquisition costs of similar drugs. 25 Widely available wholesale or GPO
prices for two of the oral immunosuppressive drugs involved average
discounts from AWP of 14 percent and 77 percent. Although we were unable to
obtain 2001 price discount information for two other oral immunosuppressive
drugs, a review of prices available from Internet pharmacies found that
retail prices available for these products were 13 percent and 8 percent
below AWP.

24 Sterile saline is the tenth. 25 HHS OIG, Medicare Reimbursement of
Albuterol (OEI- 03- 00- 00311), June 2000, and Medicare Reimbursement of
Prescription Drugs (OEI- 03- 00- 00310), January 2001. Pharmacy Suppliers?
Drug

Acquisition Costs Are Generally Lower Than AWP- Based Medicare Payment

Page 18 GAO- 01- 1118 Medicare Drug Payments

Table 6: Discounts From AWP for Medicare- Covered Drugs Billed Primarily by
DME Pharmacy Suppliers, 2001

Drug name Mode of administration Average AWP a

Average widely available discount

from AWP (percentage) b

Ipratropium bromide, unit dose form inhalation $3.52 78.0 c Acetylcysteine,
unit dose form inhalation $0.67 71.8 d Albuterol, unit dose form inhalation
$0.50 85.0 Cromolyn sodium, unit dose form inhalation $0.38 69.1 Albuterol,
concentrate inhalation $0.15 n/ a e Tacrolimus oral $3.10 n/ a e
Mycophenolate mofetil oral $2.53 14.2 d Cyclosporine oral $1.55 n/ a e
Prednisone oral $0.15 76.7 c

Saline solution, sterile intravenous, other $0.71 n/ a e a ?Average AWP? is
the average of the AWP of each NDC for that product adjusted to the
HCPCSdefined dosage. b ?Average widely available discount from AWP? for each
drug was calculated by (1) determining the average widely available price(
s) for each NDC for that drug, (2) determining the percentage difference
between the average widely available price( s) and the AWP for each NDC for
that drug, and (3) averaging the percentage differences for all NDCs for
that drug. c ?Average discount from AWP? in 2001 for this drug is based on a
price from a single wholesaler. For

Ipratropium bromide, unit dose form, we had 2000 data from three sources.
Those data showed that the average widely available discount from AWP in
2000 was 66.7 percent. For Prednisone we also had 2000 data from only one
source. d Average discounts from AWP were not available for 2001; average
discounts are from 2000.

e Wholesaler or GPO prices for these products were not available. Source:
GAO analysis of data from BESS, the Medical Economics Drug Topics Red Book
CD- ROM vol. 21, and wholesaler and GPO price lists.

Medicare?s coverage of most drugs provided by pharmacy suppliers is
established through the DME benefit; Medicare also covers selected
immunizations and certain drugs that can be self- administered, such as
blood clotting factors and some oral drugs used in association with cancer
treatment and immunosuppressive therapy. The DME benefit covers equipment,
such as nebulizers used to administer inhalation therapy drugs for treating
respiratory conditions or infusion pumps used to administer Medicare?s
Payments to

Pharmacy Suppliers for Dispensing Drugs Vary by Class of Drug

Page 19 GAO- 01- 1118 Medicare Drug Payments

intravenous drugs, and the supplies and drugs used with it. Medicare also
covers and makes a separate payment for drugs and supplies used in
conjunction with covered DME. 26 Medicare uses a fee schedule to pay for
most DME. DME fees, which are state specific, are based on historical
average supplier charges that are adjusted for inflation over time and are
subject to national minimum and maximum payment limits. Medicare pays for
certain DME on a monthly basis, subject to a maximum number of months,
depending upon whether the equipment is rented or purchased. 27

Medicare pays a dispensing fee in conjunction with inhalation therapy drugs
used in nebulizers, which account for the majority of Medicare volume and
spending in the home setting. In 2000, these dispensing fees amounted to
over $15 million for 3 million billed dispensing services. 28 Pharmacy
suppliers in aggregate received the bulk of their Medicare payments for
dispensing the inhalation therapy drugs we analyzed. Unlike many private
payers and most state Medicaid programs, Medicare does not pay a dispensing
fee to pharmacists or other providers who supply oral drugs. Finally,
Medicare neither covers nor reimburses pharmacy suppliers for the costs of
clinical services related to providing infusion therapy drugs. We did not
analyze the costs of infusion therapy drugs provided in the home setting
because they do not account for a substantial share of Medicare drug
spending or volume. DME pharmacy suppliers argued that they provide home
infusion services to beneficiaries who are too frail to travel to their
physicians? offices for care and need these services in the home. However,
such beneficiaries would likely qualify for skilled nursing care under
Medicare?s home health benefit- designed to

26 The DME benefit allows for coverage of DME itself plus ?supplies,?
including drugs that

?must be put directly into the equipment in order to achieve the therapeutic
benefit of the durable medical equipment or to assure the proper functioning
of the equipment.? (See Medicare Carriers Manual, Part 3, Chapter II-
Coverage and Limitations, section 2100.5,

Rev. 1564/ pp. 2- 45.) 27 Medicare will make a maximum of 15 monthly
payments if the equipment is rented and 13 monthly payments if it is
purchased. There are no limits, however, on how long Medicare will pay for
one category of nebulizers that requires frequent servicing as long as the
equipment is medically necessary.

28 Pharmacy suppliers bill Medicare for each month of providing nebulizer
drugs to a beneficiary, regardless of the amount of the drug billed, using
HCPCS code E0590. The average dispensing fee is currently $5. 00 per month
of service.

Page 20 GAO- 01- 1118 Medicare Drug Payments

provide clinical care with adequate supervision- which could cover such
service costs. 29

Pharmacy suppliers and their representatives said that the profit on the
Medicare drug payment is needed to compensate them for their costs related
to inhalation and infusion therapy that are not explicitly covered by
Medicare. According to these suppliers, these costs include purchasing
drugs; all clinical, administrative, and other labor associated with
delivering the drugs; billing and collection; facility and employee
accreditation; licensing and certifications; and providing printed patient
education materials. They said they are currently able to provide these
services to Medicare beneficiaries because of their margin on the Medicare
drug payments. Suppliers of inhalation and infusion therapies that we spoke
with were unable to quantify the costs of these services specific to the
provision of drugs or to the different courses of therapy they provide
relative to Medicare?s total payment, and we did not identify any other such
relevant analysis. Suppliers and their representatives reported to us that
the suppliers? acquisition cost of the drug is only a ?nominal? portion of
the total cost of providing infusion and inhalation therapy to Medicare
patients. Specifically, they indicated that drugs represented roughly a
quarter of the total cost of inhalation therapy, and approximately 40
percent of the cost of infusion therapy, excluding the cost of DME.

In this work, we did not analyze suppliers? costs of supplying DME relative
to Medicare?s payment to assess whether those payments also include a
profit. However, a prior GAO report indicated two underlying problems with
Medicare?s fees for DME that could lead to inappropriately high payments. 30
First, Medicare does not know what specific products it pays for when its
claims administration contractors process claims for DME. DME products with
similar functions or purposes are grouped together, with Medicare paying the
same amount for any product in the group. These groups of DME products often
represent a broad range of product types, quality, and market prices, but
claims only indicate a product?s group and not the specific item. The second
underlying problem with

29 Medicare?s home health care benefit enables certain beneficiaries with
post- acute- care needs and chronic conditions to receive care in their
homes. For beneficiaries who qualify, Medicare will pay for skilled nursing,
therapy, and home health aide visits under the home health benefit. See
Medicare Home Health Care: Prospective Payment System Could Reverse Recent
Declines in Spending (GAO- HEHS- 00- 176, September 8, 2000).

30 See Medicare: Need to Overhaul Costly Payment System for Medical
Equipment and Supplies (GAO/ HEHS- 98- 102, May 12, 1998).

Page 21 GAO- 01- 1118 Medicare Drug Payments

Medicare?s DME payment system is that the fee schedule allowances for DME
are often out of line with current market prices. Until recently, the
process for adjusting DME payments was very cumbersome. Medicare contractor
representatives told us that they do not change the fee schedule allowances
to reflect market prices for DME; rather the fees are typically only
adjusted for inflation annually. As a result, payment levels may not reflect
changes in technology and other factors that could significantly affect
market prices.

Private and public payers use a variety of mechanisms to determine payments
for drugs and their administration. Although varied, these mechanisms can
generally be categorized as negotiated or contracted prices and fee
schedules based on reference or benchmark prices. Private payers can
negotiate with pharmacies, other suppliers, and manufacturers, either
directly or through agents, to secure discounted prices and rebates that are
often based on volume or market share. They or their agents also may
negotiate with providers and suppliers to establish payments for services
related to providing drugs. Some public programs, such as VA, can purchase
drugs at FSS prices, which are linked to market prices. State Medicaid
programs often use fee schedules based on AWP or wholesale acquisition cost
(WAC) to reimburse providers for outpatient drugs provided under their
prescription drug benefit, but receive statutorily guaranteed manufacturers?
rebates on drugs that are derived from actual transaction data. The rebates
are based in part on AMP. In contrast to Medicare?s payment for drugs, both
the FSS and the AMP are determined under a statutory methodology and are
based on actual market prices that can be verified. In some instances, VA
and Medicare have also achieved savings by setting prices through
competition.

Large private health plans or their agents typically negotiate with
providers the terms and conditions of payments for drugs and their delivery
or administration. In general, private health plans may negotiate with
preferred or network providers to establish payments for physicianprovided
drugs. 31 Managed care plans reported that the terms of their provider
contracts vary considerably. Under some contracts, the plan may

31 Private health plans may designate a group of preferred or network
providers to deliver services to plan beneficiaries for specified payments.
With some health plans, beneficiaries may obtain services from another
provider but may then be liable for the difference between what that
provider charges and what the plan pays. Private Payers

Negotiate Prices, Other Public Payers Rely on Market- Based Benchmarks to
Establish Payments for Drugs

Private Sector Purchasers Use Contracting and Market Power to Purchase Drugs

Page 22 GAO- 01- 1118 Medicare Drug Payments

pay the physician a specified fee for the drug and its administration. Under
others, the plan may pay participating physicians a capitated rate, placing
them at financial risk for costs of a range of services, including
physiciansupplied drugs and their administration.

A large health- plan- owned pharmacy supplier reported a somewhat different
arrangement. It buys injectible drugs directly from manufacturers and
wholesalers and supplies them to the plan?s physicians. The plan contracts
to pay the physicians separately for administering the drugs. The plan?s
pharmacy benefits manager told us that the plan became a more aggressive
purchaser of physician- provided drugs when it appeared that physicians were
being overpaid for these products under an AWP- based reimbursement system.
For drugs administered through DME, representatives of infusion and
inhalation therapy providers indicated that while some plans may pay an all-
inclusive global fee for these therapies, the most common arrangement is for
private health plans to make per diem payments for pharmacy services and
delivery and separate payments for the drug and nursing services, which,
under certain conditions, Medicare covers under the home health benefit.

Private health plans commonly contract with pharmacy benefits managers (PBM)
to manage their outpatient drugs dispensed through a retail or mail order
pharmacy. PBMs negotiate contracts with retail pharmacies, which usually
include discounted prices for the drugs and fees for dispensing the drugs.
PBMs may obtain rebates from drug manufacturers. These discounts or rebates
are often based on the volume of individual drugs purchased and may be
independent of the arrangements between plans and the third parties
supplying drugs to beneficiaries such as physicians or pharmacies. Plans
often use formularies, lists of preferred drugs that may have lower
beneficiary cost sharing, as a means of increasing use of selected drugs. 32

32 Formularies are lists of prescription drugs, grouped by therapeutic
class, that health plans or insurers prefer and may encourage physicians to
prescribe and beneficiaries to use. A particular product may be included on
a formulary because of its medical value or because a favorable price was
negotiated with the manufacturer.

Page 23 GAO- 01- 1118 Medicare Drug Payments

The FSS for pharmaceuticals, administered by VA, is a list of products and
their prices available to federal entities that purchase prescription drugs.
Manufacturers must agree to supply drugs at these prices in order to have
their products covered and paid for by Medicaid programs. The FSS price is
intended to equal or better the price that the manufacturer offers its most-
favored nonfederal customer under comparable terms and conditions. 33 To
determine the most- favored customer price, manufacturers provide VA
information on price discounts and rebates offered to domestic customers and
the terms and conditions involved, 34 such as length of contract periods and
ordering and delivery practices. 35 The information used to determine the
FSS price is reviewed by VA?s OIG. Agencies using the FSS generally provide
drugs directly to beneficiaries through their own pharmacies and facilities.

Medicaid programs act as third- party payers, like Medicare, in reimbursing
physicians and pharmacies for drugs. Most Medicaid programs reimburse
pharmacies using formulas based on a percentage discount from AWP plus a
dispensing fee or a percentage markup over WAC plus a dispensing fee. 36 WAC
is the list price a wholesaler pays to a manufacturer, but it does not
include discounts that may affect the net price. WAC is not defined in law,
and like AWP is determined by the manufacturer. Medicaid programs receive
rebates based on net market prices. Under a provision of the Social Security
Act added by the Omnibus Budget Reconciliation Act of 1990 (OBRA), Medicaid
programs receive a rebate from manufacturers based on either AMP or the
manufacturer?s ?best price? to a private purchaser. 37 AMP is defined in
federal law as the average price (including cash discounts and other price
reductions) paid to drug manufacturers by

33 48 C. F. R. Sec. 538.270. 34 48 C. F. R. Sec. 538.270. 35 Because the
terms and conditions of commercial sales vary, there may be legitimate
reasons why the government does not always obtain the most- favored customer
price. Hence, under the General Services Administration regulations, VA may
accept a higher price if it determines that (1) the price offered to the
government is fair and reasonable and (2) awarding the contract is otherwise
in the best interest of the government.

36 Medicaid dispensing fees ranged from $2 to over $6 per prescription in
2000. PBMs at private health plans can negotiate even lower dispensing fees
that plans pay pharmacies through their ability to restrict their pharmacy
networks.

37 Section 1927 of the Social Security Act, added by OBRA 1990, P. L. 101-
508, Section 4401, 104 Stat. 1388, 1388- 143( 1990) (classified to 42 U. S.
C. Sec. 1396r- 8). Other Public Programs Use

Market- Based Reference Prices and Competition to Establish Payments or
Rebates

Page 24 GAO- 01- 1118 Medicare Drug Payments

U. S. wholesalers for drugs distributed to the ?retail class of trade.? 38
AMP is thus a measure of actual transaction prices between wholesalers and
manufacturers. ?Best price? is also defined in federal law as the lowest
price (including cash discounts and other price reductions) available from
the manufacturer to any U. S. wholesaler, retailer, provider, health
maintenance organization, nonprofit entity, or government entity, with some
exceptions. Manufacturers currently report AMPs and ?best prices? to CMS in
order to participate in Medicaid. CMS can survey manufacturers? sales
information to ensure that AMP and ?best price? computations are correct.
However, the Social Security Act requires that CMS maintain the
confidentiality of this pricing information and it is therefore not publicly
available.

Public payers have made several attempts to use competition to obtain more
favorable prices for drugs. VA has used a competitive bidding process to
obtain national contracts for selected drugs at prices that are even lower
than FSS prices. VA identified drugs regarded as therapeutically equivalent
and sought contracts from manufacturers for low prices in exchange for VA
including only one drug of each type on its national formulary. VA contract
prices in 2000 averaged 33 percent lower than corresponding FSS prices.

The Balanced Budget Act of 1997 authorized HCFA to conduct several limited-
scale demonstration projects to evaluate competitive bidding?s applicability
to the Medicare program. In one of these demonstration projects currently
under way in San Antonio, Texas, suppliers bid to provide inhalation drugs,
such as albuterol, to Medicare beneficiaries. While Medicare normally allows
any qualified provider to participate in the program, under the
demonstration only those 11 bidders that were selected can participate. In
exchange for restricting their choice of providers, beneficiaries have no
liability for differences between what suppliers charge and what Medicare
pays. Preliminary CMS information suggests savings of approximately 26
percent on these inhalation drugs in the San Antonio competitive bidding
demonstration.

Medicare?s method for establishing drug payment levels is flawed. In tying
its payment to AWP, a price that may be neither an average nor what
wholesalers charge, Medicare has been paying much more than providers?

38 Section 1927( k)( 1) of the Social Security Act (classified to 42 U. S.
C. Sec. 1396r- 8( k)( 1)). Conclusions

Page 25 GAO- 01- 1118 Medicare Drug Payments

likely acquisition costs. Medicare?s AWP- based methodology does not
incorporate information on actual transaction prices, such as that used by
VA in establishing FSS prices or the AMPs used to calculate Medicaid drug
rebates.

Our findings strongly suggest that Medicare should revise its drug payment
policies. Payments for the drugs themselves should closely parallel market
prices that providers pay to acquire drugs. The program needs to use
information on actual market prices net of rebates and discounts, similar to
information currently available to VA to establish the FSS and to HHS to
determine the Medicaid rebates. Although the widely available prices we
report here are often substantially below Medicare?s payment, certain
providers may be able to obtain even greater discounts. However, the lowest
prices, used in setting the FSS prices and Medicaid rebates, may not be
available to all Medicare providers for all Medicare- covered drugs. To
better ensure that beneficiary access is not compromised, in setting payment
levels it is important to be mindful that providers? ability to secure
discounts likely varies.

Physicians and other providers acknowledge that payments for drugs are
higher than their costs, but contend that profits on the drugs compensate
for what they regard as underpayments for their administration. It should be
a principle of Medicare payment policy to pay for each service appropriately
and not to rely on potential overpayments for some services to offset
potential inadequate payments for complementary services. Consequently,
separate payments for administration and delivery should be made, with those
payments appropriately reflecting the variation associated with how the
drugs are provided. Hence, different methods of determining drug delivery
and administration payments may be necessary for different types of drugs.

Some benefits of competition for securing selected drugs under certain
circumstances have been demonstrated by VA and by Medicare in the San
Antonio demonstration. Medicare?s experience is rather limited and involves
a very small number of drugs. Nevertheless, how the use of competitive
bidding might be expanded without compromising beneficiary access is worth
exploring.

Page 26 GAO- 01- 1118 Medicare Drug Payments

In order to improve the accuracy of Medicare payments for drugs and related
services, we are recommending that the Administrator of CMS take the
following actions.

 Establish Medicare payment levels for part- B prescription drugs and their
delivery and administration that are more closely related to their costs.
Payments for drugs should be set at levels that reflect actual market
transaction prices and the likely acquisition cost to providers. To
accomplish this, the Administrator should consider how information on market
transactions already available to HHS or VA may be used as a benchmark for
Medicare payment levels. If the Administrator determines that legislative
action would be required to use such information in setting Medicare
reimbursements, he should seek this action immediately.

 Examine the benefits and risks of expanding the current competitive
bidding demonstration projects for drugs covered under part B.

 Institute a process to monitor access to Medicare part B- covered drugs to
ensure that payment changes do not negatively affect access for particular
drugs, or groups of beneficiaries or in certain geographic areas.

CMS noted that our findings confirmed the results of studies by the HHS OIG
that indicated that Medicare payments for drugs are substantially higher
than their actual acquisition costs. CMS agreed that Medicare should
appropriately pay for both part B- covered drugs and the services required
to furnish them. CMS?s comments appear in their entirety as appendix II.

We are sending copies of this report to the Administrator of CMS and
interested congressional committees. We will make copies available to others
on request.

If you or your staffs have any questions about this report, please call me
at (202) 512- 7119 or James Mathews at (202) 512- 9427. Major contributors
to this report were Ginny Hsieh, Dina Kirschenbaum, Kathryn Linehan, and
Theresa Thompson.

Laura A. Dummit Director, Health Care- Medicare Payment Issues
Recommendations for

Executive Action Agency Comments

Page 27 GAO- 01- 1118 Medicare Drug Payments

List of Congressional Committees The Honorable Max Baucus Chairman The
Honorable Charles E. Grassley, Jr. Ranking Minority Member Committee on
Finance United States Senate

The Honorable Bill Thomas Chairman The Honorable Charles B. Rangel Ranking
Minority Member Committee on Ways and Means House of Representatives

The Honorable W. J. ?Billy? Tauzin Chairman The Honorable John D. Dingell
Ranking Minority Member Committee on Energy and Commerce House of
Representatives

Appendix I: Scope and Methodology Page 28 GAO- 01- 1118 Medicare Drug
Payments

In conducting this study, we interviewed officials at CMS, VA, HHS OIG, and
DOJ. We also interviewed and solicited information about drug acquisition
costs and costs related to drug administration from professional
associations representing physicians and other providers with a major stake
in our evaluation. We requested price data on specific drugs from major
pharmaceutical wholesalers, oncology specialty wholesalers, and GPOs, none
of which responded to our request by the time we finalized this report. We
also requested price data from specialty pharmacies, DME pharmacy suppliers,
a national oncology clinic chain, and physicians, including physicians we
identified as billing for low volumes of Medicare- covered drugs. We
interviewed representatives from large managed care organizations and their
agents regarding the methods they used to purchase or pay for
pharmaceuticals as well as the purchase prices they were able to obtain. We
also interviewed staff at the companies that publish major price reporting
compendia that collect and report average wholesale prices for drugs. We
conducted quantitative analyses using data from CMS and other sources. We
did our work in accordance with generally accepted government auditing
standards from January through September 2001.

To determine the drugs and biological products that accounted for the most
Medicare spending and volume, we used 1999 data from BESS. We also used BESS
data to determine spending and volume by part B carrier, place of service,
and specialty. All analyses using BESS data excluded data on services
supplied in Puerto Rico and the U. S. Virgin Islands and payments made on
behalf of the Railroad Retirement Board.

We excluded four HCPCS- coded products from our analysis because of problems
obtaining accurate pricing information or, in one case, because the code was
not specific enough to link to a given product. From the highvolume/ high-
expenditure HCPCS codes that were chosen for our study, we evaluated pricing
information for the NDCs that were used to determine reimbursement rates for
that HCPCS. We asked the Medicare part B carriers with the highest allowed
drug charges in each region and the DME regional carriers to provide the
NDCs they used to calculate reimbursement for 31 HCPCS codes. Because
carriers may use somewhat different groups of NDCs, we included all of the
NDCs in our analysis of prices for each HCPCS we anlayzed.

We obtained widely available drug prices for selected HCPCS for 2000 and
2001 from several wholesalers? and GPOs? price lists, which we obtained from
a government agency and from providers. The AWPs were obtained Appendix I:
Scope and Methodology

Appendix I: Scope and Methodology Page 29 GAO- 01- 1118 Medicare Drug
Payments

from the 2000 Drug Topics? Red Book and the 2001 Drug Topics? Red Book CD-
ROM by NDC for each year.

To identify physicians who billed for low volumes of multiple drugs, we
examined all Medicare part B drug claims from 1999 and identified 1,115
physicians who met our definition of a low- volume Medicare biller for
cancer drugs in 1999, and selected a random sample of 108 physicians. We
defined low- volume biller as a physician who billed for between the 10th
and 25th percentile of total allowed units for three or more of the
following drugs: Dexamethasone sodium phosphate, Leuprolide acetate,
Rituximab, Goserelin acetate, Docetaxel, Filgrastim (480 mcg), Pamidronate
disodium, Filgrastim (300 mcg), Paclitaxel, Irinotecan, Carboplatin,
Gemcitabine HCl, Dolasetron mesylate, Granisetron HCl, Leucovorin calcium,
Epoetin alpha for non- ESRD use, and Ondansetron HCl. We then merged the
claims file physician sample with the Unique Physician Identification Number
directory to get names and addresses of providers. We received 74 prices for
NDCs for 16 HCPCS- coded drugs from 14 low- volume billers. An additional 37
physicians were hospitalbased, operated out of large practices, or worked
for large nationally owned chain oncology practices. We could not contact 36
physicians for whom we could not find correct telephone numbers. Twenty
practices declined to provide us with information or did not provide
information in time to be included in this report. One practice indicated it
did not purchase drugs.

Appendix II: Comments From the Centers for Medicare and Medicaid Services

Page 30 GAO- 01- 1118 Medicare Drug Payments

Appendix II: Comments From the Centers for Medicare and Medicaid Services

(290024)

The first copy of each GAO report is free. Additional copies of reports are
$2 each. A check or money order should be made out to the Superintendent of
Documents. VISA and MasterCard credit cards are also accepted.

Orders for 100 or more copies to be mailed to a single address are
discounted 25 percent.

Orders by mail:

U. S. General Accounting Office P. O. Box 37050 Washington, DC 20013

Orders by visiting:

Room 1100 700 4 th St., NW (corner of 4 th and G Sts. NW) Washington, DC
20013

Orders by phone:

(202) 512- 6000 fax: (202) 512- 6061 TDD (202) 512- 2537

Each day, GAO issues a list of newly available reports and testimony. To
receive facsimile copies of the daily list or any list from the past 30
days, please call (202) 512- 6000 using a touchtone phone. A recorded menu
will provide information on how to obtain these lists.

Orders by Internet

For information on how to access GAO reports on the Internet, send an email
message with ?info? in the body to:

Info@ www. gao. gov or visit GAO?s World Wide Web home page at: http:// www.
gao. gov

Contact one:

 Web site: http:// www. gao. gov/ fraudnet/ fraudnet. htm

 E- mail: fraudnet@ gao. gov

 1- 800- 424- 5454 (automated answering system) Ordering Information

To Report Fraud, Waste, and Abuse in Federal Programs
*** End of document. ***