ASPE ISSUE BRIEF
`
`OFFICE OF THE ASSISTANT SECRETARY FOR PLANNING AND EVALUATION
` OFFICE OF SCIENCE AND DATA POLICY - U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES
`
`
`JUNE 2014
`
`MEDICARE PART B REIMBURSEMENT
`
`OF PRESCRIPTION DRUGS
`
`Introduction
`
`Certain health care providers, especially oncologists and
`anesthesiologists, purchase drugs that they administer to patients
`as a componentoftheir practices. These drugsinclude sterile
`injectable drugs and some drugs used in conjunction with certain
`durable medical equipment, such as nebulizers. Unlike the many
`other medical supplies that providers routinely buy in the private
`market, insurers often separately reimburse providers for the cost
`of these drugs, rather than expecting providers to cover these
`costs as part of their global reimbursement. The prices of the
`drugs themselves, however, are not subject to price regulation
`under Medicare, and the prices paid by providers to suppliers for
`these products do not depend onthe type of insurance a patient
`has.
`
`Medicare Part B Reimbursementof Drugsprior to the
`Medicare Modernization Act
`
`The Medicare Part B drug payment system is used by Medicare to
`reimbursehealth care providers for the average costs of the drugs
`they administer whenproviding outpatient services to Medicare
`beneficiaries. Reimbursementto individual providers is based ona
`formula computed from national sales data, not on the price paid
`by a specific provider. This reimbursement formula has changed
`over time. Following enactmentof the Balanced Budget Act of
`1997, the Health Care Financing Administration, now the Centers
`for Medicare and Medicaid Services (CMS), required Medicare
`carriers, which process Medicare claims, to base their
`reimbursementfor a covered drug onits average wholesale price
`(AWP) as published in RED BOOK™ orsimilar drug pricing
`publications used by the pharmaceutical industry. Specifically, for
`covered drugs available only from a brand source, reimbursement
`wascalculated as 95 percent of the drug’s AWP.For covered
`drugs available from brand and generic sources, reimbursement
`
` ABOUT THIS ISSUE
`
`BRIEF
`
`This Issue Brief describes how
`Medicare Part B reimburses
`the cost ofprescription drugs
`administered in physician
`offices and hospital
`outpatient settings. It explains
`the changes madeto the
`reimbursementsystem under
`the Medicare Prescription
`Drug, Improvement, and
`Modernization Act of2003,
`summarizes the direct
`consequences of these
`changes, and presents
`analysis of drug price
`variation subsequentto these
`changes.
`
`This Issue Brief was written by
`Former Assistant Secretaryfor
`Planning and Evaluation
`Sherry Glied and Kevin
`Haninger.
`
`Office of the Assistant Secretary
`for Planning and Evaluation
`
`Office of Science and Data Policy
`
`U.S. Departmentof Health and
`HumanServices
`
`Washington, DC 20201
`
`
`
`Page 01 of 09
`
`Exhibit 2155
`
`Mylan v. Regeneron
`IPR2021-00881
`U.S. Pat. 9,254,338
`Exhibit 2155
`
`Exhibit 2155
`Page 01 of 09
`
`

`

`was the lesser amount of 95 percent of the median AWP for generic sources or 95 percent of the AWP
`for the brand source.1
`Beginning in 1997, several organizations, including the U.S. Department of Health and Human Services
`(HHS) Office of Inspector General,2,3,4,5 the Medicare Payment Advisory Commission (MedPAC),6 the U.S.
`Government Accountability Office,7 and the Congressional Research Service,8 identified two main
`problems with this reimbursement system. First, the Balanced Budget Act of 1997 did not define AWP,
`and most analysts found that the figures used were inflated relative to actual prices paid, lacked uniform
`reporting criteria, and could not be verified. The lack of standardization also resulted in local Medicare
`carriers using different AWPs for the same drug code, even though the drug payment system was a
`national formula that did not provide for differential reimbursement based on geography. Second,
`because of rebates and other discounts, the published AWPs used by Medicare carriers to calculate
`reimbursement were substantially higher than the actual acquisition prices available to providers who
`billed for these drugs. While Medicare paid 95 percent of the AWP, most of these drugs were available
`to providers for 66 percent to 87 percent of the AWP, with some drugs available for considerably less.6
`As a result, Medicare paid providers roughly a billion dollars more than acquisition costs annually for
`Part B drugs, and Medicare beneficiaries, who were responsible for a 20 percent copayment, paid
`hundreds of millions of dollars more annually than if payment rates reflected actual acquisition costs.2,3
`
`Medicare Part B Reimbursement of Drugs under the Medicare Modernization Act
`The Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (MMA) intended to
`reduce overpayments for drugs administered in physician offices and hospital outpatient settings by
`basing the reimbursement formula on a more readily verifiable and market-based price measure. The
`reimbursement changes, which took full effect in January 2005, tied reimbursement more closely to
`health care providers’ acquisition costs by paying for a drug’s average sales price (ASP) plus a 6 percent
`margin to cover overhead costs for drugs administered in physician offices9 or plus an annually updated
`margin (currently 6 percent) for separately payable drugs administered in hospital outpatient settings.10
`By law, a drug’s ASP is defined as the volume-weighted average manufacturer sales price net of all
`rebates, discounts, and other price concessions to U.S. purchasers, excluding sales that are exempt from
`
`
`1 42 C.F.R. § 405.517.
`2 U.S. Department of Health and Human Services, Office of Inspector General, Excessive Medicare Payments for
`Prescription Drugs, Pub. No. OEI-03-97-00290, 1997.
`3 U.S. Department of Health and Human Services, Office of Inspector General, Medicare Reimbursement of
`Prescription Drugs, Pub. No. OEI-03-00-00310, 2001.
`4 U.S. Department of Health and Human Services, Office of Inspector General, Excessive Medicare Reimbursement
`for Albuterol, Pub. No. OEI-03-01-00410, 2002.
`5 U.S. Department of Health and Human Services, Office of Inspector General, Excessive Medicare Reimbursement
`for Iipratropium Bromide, Pub. No. OEI-03-01-00411, 2002.
`6 Medicare Payment Advisory Commission, Chapter 9: Medicare Payments for Outpatient Drugs under Part B,
`Report to the Congress: Variation and Innovation in Medicare, 2003.
`7 U.S. Government Accountability Office, Medicare: Payments for Covered Outpatient Drugs Exceed Providers’
`Cost, Pub. No. GAO-01-1118, 2001.
`8 Congressional Research Service, Medicare: Payments for Covered Prescription Drugs, Pub. No. RL31419, 2002.
`9 42 C.F.R. § 405 as amended.
`10 42 C.F.R. § 419 as amended. Separately payable drugs are those that are not packaged within an ambulatory
`payment classification group because their average cost per day of treatment exceeds $80. The Medicare
`Hospital Outpatient Prospective Payment System (OPPS) has typically reimbursed these drugs at ASP plus a 4 to
`6 percent margin.
`
`
`
`
`
`
`ASPE ISSUE BRIEF | 2
`
`
`
`Exhibit 2155
`Page 02 of 09
`
`

`

`Medicaid “best price” calculations and sales to other federal purchasers.*? Manufacturers are required
`to provide CMSwith the quarterly sales price and volumeofsales for each covered drug by National
`Drug Code (NDC) within 30 days of the end of the quarter. Because multiple manufacturers may produce
`the same drug, CMScrosswalks NDCs for the same drug using the Healthcare CommonProcedure
`Coding System (HCPCS). CMS then calculates a volume-weighted ASP for each HCPCS code, which
`becomesthebasis for the reimbursementrate for the following quarter. Given the time needed to
`submit and process sales data, the current reimbursementrate always reflects a drug’s ASP from two
`quartersprior. Figure 1 showsthe timeline for establishing Medicare Part B reimbursementrates.
`
`Following the MMAchange, mostprivate payers adopted the Medicare Part B drug payment system.
`Since providers do not buy drugs separately for differently insured patients, the Medicare database of
`ASPswasthebasis for private payer reimbursements, although someprivate payers paid higher or
`lower surcharges (compared to the 6 percent) than Medicare paid.’”*? More recently, however, one of
`the nation’s largest private payers eliminated this payment mechanism altogether and began bundling
`paymentsfor a total course of chemotherapy, with the goal of separating oncologists’ income from their
`drug selection.“ Under this payment arrangement, chemotherapydrugsare treatedlike all other
`medical supplies and products — the providerreceives no special reimbursementfortheir cost.
`
`Figure 1. Timeline for Setting of Medicare Part B Reimbursement Rates
`
`Manufacturers havesales data
`prior to submitting to CMS.
`
`CMS has sales data
`prior to publishing
`reimbursement rates.
`
`same drug.
`
`A manufacturer’s average sales price (ASP)
`and volume sold of a given drug is calculated
`by the manufacturer every quarter and
`submitted to CMS within 30 days of the end
`of the quarter.
`
`manufacturers of the
`
`442 U.S.C. § 1395w-3a(c).
`?2 Mullen P, “The Arrival of Average Sales Price,” Biotechnology Healthcare 4:48-53, 2007.
`43 Arkansas BlueCross and BlueShield, Provider’s News: March 2005.Availableat:
`http://www.arkansasbluecross.com/doclib/publications/march 2005.pdf.
`44 Appleby J, “A New WayTo Pay For Chemotherapy: Major Insurer Would Pay Oncologists A Set Fee For Certain
`Cancers,” Kaiser Health News: October20, 2010. Available at:
`http://www.kaiserhealthnews.org/stories/2010/october/20/chemotherapy.aspx.
`
`ASPEISSUEBRIEF| 3
`
`Exhibit 2155
`
`Page 03 of 09
`
`Exhibit 2155
`Page 03 of 09
`
`

`

`Direct Impacts of the Medicare Modernization Act
`Following the MMA changes to the Medicare Part B drug payment system, MedPAC issued two
`Congressionally-mandated reports, which found that health care providers could still purchase most
`covered drugs at prices below the Medicare Part B reimbursement rate.15,16
`As had been expected, the change in the Medicare Part B drug payment system had immediate impacts
`on the revenues of the affected providers. There is also some evidence that physicians changed their
`prescribing patterns in response to the lower reimbursement rates, either by providing more services or
`by substituting more profitable services for less profitable ones. MedPAC studied how physician
`specialties responded to the MMA reimbursement change.15,16 Overall, MedPAC found, oncologists and
`rheumatologists responded to the payment change by providing more services, urologists provided
`fewer services, and infectious disease specialists shifted some services back to hospital settings where
`drugs are typically purchased by the hospital. A more recent study found that the percentage of lung
`cancer patients who received chemotherapy within one month of diagnosis increased by 2.4 percentage
`points shortly after the January 2005 payment change.17
`We reviewed stock analyst reports and assessments by the drug manufacturers at the time of the MMA
`change. None of the reports at the time anticipated any effect of the reimbursement mechanism change
`on the prices paid to manufacturers.
`
`Drug Price Variation subsequent to the Medicare Modernization Act
`The MMA reimbursement change did not alter the process through which providers negotiated with
`drug manufacturers over the price of drugs and would not be expected to have any effect on price
`volatility. Medicare Part B reimbursement rates do not establish future prices – they are based on prices
`previously obtained in the market. Thus, shortages (or surpluses) of a given drug will lead to price
`increases (or decreases), just as had been the case before the MMA. As had been the case under the
`prior reimbursement mechanism, the two-quarter delay in updating reimbursement rates means that
`when the market price of a drug rises, reimbursement rates lag prices; conversely, when the market
`price of a drug falls, reimbursement rates exceed prices. Over the period since the MMA, drug prices
`have generally fallen,18 so that reimbursement rates, based on lagged prices, have generally been more
`than 6 percent above the prevailing ASP. However, some analysts have suggested that if providers are
`more sensitive to losses than to gains, manufacturers may be reluctant to raise drug prices by more than
`6 percent within a quarter since providers might face losses on purchases during the period before
`Medicare updates the reimbursement rate to reflect a higher ASP.
`To examine the possibility that Medicare Part B reimbursement rules lead to low price volatility and
`discourage price increases above 6 percent, we analyzed quarterly ASPs and Medicare Part B
`reimbursement rates for covered drugs from 2005 to 2013 using ASP Drug Pricing Files and Part B
`National Summary Data Files provided by CMS.19 For each HCPCS code (n = 735), we calculated the
`
`
`15 Medicare Payment Advisory Commission, Report to the Congress: Effects of Medicare Payment Changes on
`Oncology Services, 2006.
`16 Medicare Payment Advisory Commission, Report to the Congress: Impact of Changes in Medicare Payments for
`Part B Drugs, 2007.
`17 Jacobson M, Earle CC, Price M, and Newhouse JP, How Medicare’s Payment Cuts for Cancer Chemotherapy
`Drugs Changed Patterns of Treatment, Health Affairs 29(7): 1394-1402, 2010.
`18 U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation,
`Economic Analysis of the Causes of Drug Shortages, 2011.
`19 We cannot directly test whether the MMA reduced price volatility because data prior to 2005 are not available.
`
`
`
`
`
`
`ASPE ISSUE BRIEF | 4
`
`
`
`Exhibit 2155
`Page 04 of 09
`
`

`

`percentage change in ASP from the previous quarter and determined how frequently ASP changed by
`various specified magnitudes (e.g., 6 percent). We also calculated the coefficient of variation of the
`quarterly ASPs per HCPCS code to standardize how much a single drug’s ASP varied over time and then
`compared the degree of variation over time across all Part B drugs. The coefficient of variation is the
`ratio of the standard deviation to the mean.
`The empirical results provide strong evidence against the claim that manufacturers do not raise drug
`prices by more than 6 percent within a quarter. ASPs for nearly three-quarters (73 percent) of HCPCS
`codes increased by more than 6 percent from the previous quarter at least once between 2005 and
`2013. Table 1 reports the distribution of how frequently ASP increased by more than 6 percent from the
`previous quarter.
`Table 1. Distribution of How Frequently Quarterly ASP Increased by More Than 6 Percent per HCPCS Code
`Percentage of Quarters in which
`Number of
`Percentage of
`ASP Increased by More Than 6 Percent
`HCPCS Codes
`HCPCS Codes
`None
`202
`27.5%
`0% to 5%
`45
`6.1%
`5% to 10%
`77
`10.5%
`10% to 15%
`104
`14.1%
`15% to 20%
`56
`7.6%
`20% to 25%
`97
`13.2%
`25% to 30%
`47
`6.4%
`30% to 35%
`51
`6.9%
`More than 35%
`56
`7.6%
`Total
`735
`100%
`
`
`Increases in ASP of this magnitude are not uncommon. On average, a HCPCS code experienced a greater
`than 6 percent increase in ASP during 15 percent of all quarters, with many HCPCS codes experiencing
`increases of this magnitude more often. Moreover, this analysis underestimates how often individual
`manufacturers raised drug prices by more than 6 percent since ASP averages different manufacturers’
`prices for the same drug. Figure 2 graphs how frequently quarterly ASP changed by various magnitudes
`ranging from decreases of more than 20 percent to increases of more than 20 percent. The graph shows
`a declining distribution of progressively larger changes in quarterly ASP with no concentration at 6
`percent.
`
`
`
`
`
`
`
`
`
`ASPE ISSUE BRIEF | 5
`
`
`
`Exhibit 2155
`Page 05 of 09
`
`

`

`Figure 2. Frequency and Magnitude of Changes in Quarterly ASP
`
`30
`
`25
`
`20
`
`15
`
`10
`
`05
`
`Percentage of all Quarters
`
`10 12 14 16 18 20+
`8
`6
`4
`2
`0
`-2
`-4
`-6
`-20+ -18 -16 -14 -12 -10 -8
`Percentage Change in ASP from the Previous Quarter
`
`Overall, the quarterly ASPs for many drugs experienced considerable variation between 2005 and 2012.
`The average coefficient of variation of quarterly ASPs per HCPCS code was 0.22, with a range between 0
`and 1.9. Table 2 reports the distribution of the coefficient of variation of the quarterly ASPs per HCPCS
`code.
`Table 2. Distribution of the Coefficient of Variation of Quarterly ASPs per HCPCS Code
`Coefficient of Variation of
`Number of
`Percentage of
`Quarterly ASPs per HCPCS Code
`HCPCS Codes
`HCPCS Codes
`Zero
`27
`3.7%
`0 to 0.1
`264
`36.0%
`0.1 to 0.2
`159
`21.6%
`0.2 to 0.3
`105
`14.3%
`0.3 to 0.4
`61
`8.3%
`0.4 to 0.5
`40
`5.4%
`0.5 to 0.6
`28
`3.8%
`0.6 to 0.7
`13
`1.8%
`0.7 to 0.8
`9
`1.2%
`0.8 to 0.9
`4
`0.5%
`0.9 to 1.0
`10
`1.4%
`More than 1.0
`15
`2.0%
`Total
`735
`100%
`
`
`
`
`
`
`
`
`Figure 3 graphs the quarterly ASPs and Medicare Part B reimbursement rates for three selected drugs to
`illustrate the variation in ASP.
`
`
`
`
`
`
`
`
`
`ASPE ISSUE BRIEF | 6
`
`
`
`Exhibit 2155
`Page 06 of 09
`
`

`

`Figure 3. Average Sales Prices and Medicare Part B Reimbursement Rates for Selected Drugs
`
`$12
`
`$10
`
`$8
`
`$6
`
`$2
`
`$o
`
`$40
`
`$35
`
`$30
`
`$25
`
`$20
`
`$15
`
`$10
`
`$5
`
`$o
`
`m AverageSales Price
`
`= Medicare Part B Payment Rate
`
`CEFOTAXIME SODIUM INJECTION (1000 MG)
`
`007€D
`
`007vO
`
`S002TO
`
`S002ZO
`
`
`
`S002€D
`
`S007vO
`
`9002TO
`
`9002ZO
`
`
`
`9007€D
`
`9007vO
`
`£002TO
`
`4002ZO
`
`
`
`4002€0
`
`2002vO
`
`8002TO
`
`8002ZO
`
`
`
`8007€D
`
`8007vO
`
`6002TO
`
`6002ZO
`
`
`
`6002€D
`
`6007vO
`
`OT0zTO
`
`OT0zZO
`
`O10ZED
`
`O10zvO
`
`TLOZTO
`
`TLOZZO
`
`TLOzEO
`
`TLOZvO
`
`7102TO
`
`7107ZO
`
`
`
`7107€O
`
`7107vO
`
`€10ZTO
`
`€10ZZO
`
`€10Z€O
`
`€10ZvO
`
`o=a=Fud==<Yco2aOo=2xa
`
`002ED
`
`007vO
`
`9002TO
`
`9002ZO
`
`
`
`9002€D
`
`9002vO
`
`£002TO
`
`4002ZO
`
`
`
`2002€D
`
`2002vO
`
`8002TO
`
`8002ZO
`
`
`
`8002€D
`
`8002vO
`
`6002TO
`
`6002ZO
`
`6002€D
`
`$8
`
`$7
`
`$6
`
`$5
`
`$3
`
`$2
`
`$1
`
`$o
`
`S002TO
`
`S002ZO
`
`
`
`S007€D
`
`S007vO
`
`
`
`6002vO
`
`OT0ZTO
`
`OT0zZO
`
`OT0ZED
`
`010ZvO
`
`TLOzTO
`
`TLOZZO
`
`TLOZED
`
`TLOZvO
`
`2102TO
`
`2107ZO
`
`
`
`7107€O
`
`2107vO
`
`€10ZTO
`
`€10ZZO
`
`€10ZED
`
`€10ZvO
`
`DOXORUBICIN HCL INJECTION (10 MG)
`
`
`
`v00z€0
`
`v00zvO
`
`$002TO
`
`S002ZO
`
`
`
`S002€0
`
`S002vO
`
`9002TO
`
`9002ZO
`
`
`
`9002€0
`
`9002vO
`
`£002TO
`
`4002ZO
`
`
`
`4002€0
`
`2002vO
`
`8002TO
`
`8002ZO
`
`
`
`8002€0
`
`8002vO
`
`6002TO
`
`6002ZO
`
`
`
`6002€0
`
`6002vO
`
`OTOzTO
`
`OTOZZO
`
`
`
`OTOZ€D
`
`OTOZvO
`
`TTOZTO
`
`TLOZZO
`
`
`
`TLOZ€0
`
`TLOZvO
`
`zT0zTO
`
`zT0ZZO
`
`
`
`ZT0Z€O
`
`zT02vO
`
`€T0zTO
`
`€T0ZZO
`
`€T0Z€O
`
`€T0ZvO
`
`ASPEISSUEBRIEF| 7
`
`Exhibit 2155
`
`Page 07 of 09
`
`Exhibit 2155
`Page 07 of 09
`
`
`
`
`
`

`

`Other Factors Affecting Drug Prices
`Another, and much larger, impact on the Medicare Part B reimbursement received by health care
`providers has been the shift from branded drugs to generic drugs. For example, between 2002 and
`2013, while the overall market for sterile injectable drugs increased by 39 percent, the number of units
`sold by generic drug manufacturers increased by 57 percent.20 This increase in sales volume has been
`accompanied by an even greater increase in sales revenue. According to IMS Health sales data, the total
`sales revenue for generic sterile injectable products grew from $2.0 billion in 2002 to $7.8 billion in
`2013, a 200 percent increase in real dollars after deflating prices with the CPI-U. The shift from branded
`drugs to generic drugs has also affected other dosage forms covered by Medicare Part B.21,22
`This shift had very large effects on both total reimbursements and on the dollar value of the 6 percent
`margin received by providers, since providers would often have been receiving reimbursements based
`on the branded drug price while acquiring generic drugs.22
`
`Conclusion
`In response to widespread manipulation and overpayments associated with the previous AWP-based
`Medicare Part B drug payment system, the MMA tied reimbursement more closely to health care
`providers’ acquisition costs by paying for a drug’s market price. Under this acquisition process, Medicare
`has no price-setting power – reimbursement rates lag rather than lead market prices. Empirical analysis
`of quarterly ASPs and reimbursement rates for covered drugs shows that market prices vary
`considerably over time and occasionally experience sharp spikes. In general, however, previous studies
`have found that most providers could still purchase most covered drugs at or below the reimbursement
`rate.
`
`
`
`
`
`20 IMS Health, IMS National Sales Perspective™, Data extracted May 2014.
`21 IMS Institute for Healthcare Informatics, Medicine Use and Shifting Costs of Healthcare: A Review of the Use of
`Medicines in the U.S. in 2013, 2014.
`22 U.S. Department of Health and Human Services, Office of Inspector General, Medicare Payments for Newly
`Available Generic Drugs, Pub. No. OEI-03-09-00510, 2011.
`
`
`
`
`
`
`ASPE ISSUE BRIEF | 8
`
`
`
`Exhibit 2155
`Page 08 of 09
`
`

`

`
`
`
`DEPARTMENT OF HEALTH
`& HUMAN SERVICES
`
`Office of the Secretary
`Washington, DC
`
`
`
`
`
`
`
`
`OFFICIAL BUSINESS
`Penalty for Private Use $300
`
`
`
`
`
`
`
`Exhibit 2155
`Page 09 of 09
`
`