OFFICE OF INSPECTOR GENERAL MEDICARE PART B DRUG PAYMENTS: IMPACT OF PRICE SUBSTITUTIONS BASED ON 2015 AVERAGE SALES PRICES Suzanne Murrin Deputy Inspector General for Evaluation and Inspections OEI-03-17-00360 September 2017 Report in Brief of September 2017 OEI-03-17-00360 Why OIG Did This Review Medicare Part B Drug Payments: Impact of Price When Congress established average sales price (ASP) as the basis for Medicare Substitutions Based on 2015 Average Sales Prices Part B drug reimbursement, it also What OIG Found provided a mechanism for monitoring market prices and limiting potentially  CMS lowered Part B reimbursement for 13 drugs on the basis of 2015 data. excessive payment amounts. The Social Security Act mandates that the Office of  CMS’s price-substitution policy saved Medicare and its beneficiaries Inspector General (OIG) compare ASPs $5.4 million over 1 year based on 2015 data. with average manufacturer prices (AMPs). If OIG finds that the ASP for a  Medicare and its beneficiaries could have saved up to an additional drug exceeds the AMP by a certain $17 million over 1 year if CMS implemented a more expansive percentage (currently 5 percent), the Act price-substitution policy that, for example, allowed substitution for drugs directs the Secretary of Health and that exceeded the 5-percent threshold in a single quarter. Human Services to substitute the ASP-based payment amount with a lower calculated rate. Through regulation, the Exhibit 1: Results of the Medicare Part B Centers for Medicare & Medicaid Price-Substitution Policy Services (CMS) outlined that it would make this substitution only if the ASP for a drug exceeds the AMP by 5 percent in the 2 previous quarters or 3 of the previous 4 quarters. Over the last decade, OIG has produced annual reports aggregating the results of our mandated quarterly ASP-to-AMP comparisons. This annual report quantifies the savings to Medicare and its beneficiaries that are a direct result of CMS’s price-substitution policy based on 2015 average sales prices, and this report also offers recommendations for Source: OIG analysis of ASP and AMP data from 2015 achieving additional savings. What OIG Recommends How OIG Did This Review Because of the potential for savings to Medicare beneficiaries and the program, To determine the effects of the price-substitution policy, we determined OIG recommends that CMS expand the price-substitution policy. CMS did not the difference between ASP-based concur with the recommendation. However, CMS stated that as additional data payment and AMP-based payment for becomes available and as it continues to gain experience with the each drug with a price substitution. We price-substitution policy, it will consider further changes as necessary. OIG then applied this difference to the recognizes that CMS, in setting policy for payment substitution, needs to Medicare utilization for each of these achieve an important balance between safeguarding access to drugs and drugs. To account for a 3-quarter lag between the reporting of pricing data ensuring that Medicare and its beneficiaries do not overpay for drugs. To and the application of price substitutions, provide greater flexibility and achieve this continued balance, any future we used drug utilization data for the expansion of the payment-substitution policy could contain a provision that fourth quarter of 2015 through the third would prevent a price substitution when there are indications that the quarter of 2016 to calculate the savings substitution amount is below the provider acquisition costs. based on 2015 ASP data. Full report can be found at http://oig.hhs.gov/oei/reports/oei-03-17-00360.asp BACKGROUND When Congress established average sales price (ASP) as the Drug Pricing Terms basis for Medicare Part B drug reimbursement, it also provided a mechanism for monitoring market prices and adjusting Manufacturer’s Average Sales Price ASP-based payments in certain situations. Specifically, the (ASP) Social Security Act (the Act) mandates that the Office of In general, the manufacturer’s ASP Inspector General (OIG) compare ASPs with average for a unit of drug that is sold is manufacturer prices (AMPs).1 If OIG finds that the ASP for a defined as the manufacturer’s sales drug exceeds the AMP by a certain percentage (currently of a drug to all purchasers in the 5 percent), the Act directs the Secretary of Health and Human United States in a calendar quarter Services (after being notified by OIG) to substitute the payment divided by the total number of units amount with the lesser of the widely available market price of the drug sold by the manufacturer (if any) or 103 percent of the AMP.2, 3 in that same quarter. Average Manufacturer Price (AMP) Payments for Prescription Drugs Under Medicare In general, AMP is defined as the Part B average price paid to the Medicare Part B covers a limited number of outpatient manufacturer for the drug in the prescription drugs. These drugs are usually administered in a United States by (1) wholesalers for physician’s office or other outpatient setting and include, for drugs distributed to retail community example, drugs used to treat cancer. To obtain reimbursement pharmacies and (2) retail community for Part B drugs, health care providers submit claims to pharmacies that purchase drugs Medicare contractors using Healthcare Common Procedure directly from the manufacturer. Coding System (HCPCS) codes. (Hereinafter in this report, we Wholesale Acquisition Cost (WAC) refer to HCPCS codes as “drugs.”4) WAC is defined as the manufacturer's CMS calculates the payment amount for these drugs using list price for a drug to wholesalers or information provided by manufacturers. Certain direct purchasers. It does not include manufacturers must provide CMS with the ASP and volume of discounts or rebates. sales for each of their national drug codes (NDCs) on a National Drug Code (NDC) quarterly basis.5, 6 CMS then calculates an ASP-based payment An NDC is a code used to identify a amount for the drug, which includes all of the NDCs associated drug based on its manufacturer, with the drug.7 Under the ASP pricing methodology, the product, and package size. Medicare reimbursement for most Part B drugs is equal to Healthcare Common Procedure 106 percent of the volume-weighted ASP for the drug.8 Coding System (HCPCS) Code However, under sequestration legislation, the payment amount A HCPCS code is a standardized billing for most drugs is reduced by 2 percent.9 code that is used primarily to identify Quarterly reimbursement amounts are not based on current products, supplies, and other quarter data because there is a 2-quarter lag between the sales services. A HCPCS code specifies the period for which ASPs are reported and the effective date of name and the amount of the drug and may represent one or more the reimbursement amounts. For example, manufacturers’ NDCs. ASPs from the first quarter of 2015 were used to establish reimbursement amounts for the third quarter of 2015. Medicare Part B Drug Payments: Impact of Price Substitutions Based on 2015 Average Sales Prices 1 OEI-03-17-00360 Manufacturer Reporting of AMPs In addition to providing quarterly ASPs, certain manufacturers must provide CMS quarterly with the AMP for each of their NDCs.10, 11 The AMP is generally calculated as a weighted average of prices for all of a manufacturer’s package sizes of a drug and is reported for the lowest identifiable quantity of the drug, e.g., 1 milliliter, one tablet, one capsule. AMP-Based Price Substitutions Through regulation, CMS established the criteria under which it would implement a price substitution for a drug. CMS may substitute 103 percent of the AMP for the ASP-based reimbursement amount when OIG identifies a drug that exceeds the 5-percent threshold in the 2 previous quarters or 3 of the previous 4 quarters.12 CMS implemented the AMP substitution policy in April 2013. Because CMS believes that comparisons based on partial AMP data may not adequately reflect market trends, the agency will consider lowering reimbursement amounts only when corresponding AMP data is available for each of the NDCs used to determine the published reimbursement amount for a drug.13 To prevent the price-substitution policy from inadvertently raising Medicare reimbursement amounts, CMS does not substitute prices when the substituted amount is greater than the ASP-based payment amount calculated for the quarter in which the price substitution takes effect.14 CMS also does not substitute prices when a drug is identified by the Food and Drug Administration (FDA) as being in short supply.15 Price substitutions take effect in the quarter after OIG shares the results of its most recent pricing comparison and remain in effect for one quarter.16 Because of the 2-quarter lag between the ASP reporting period and the effective date of reimbursement amounts, and the additional quarter that is necessary for OIG to complete its pricing comparison, there is a 3-quarter lag between the ASP reporting period and the effective date of the price substitutions. As shown in Exhibit 2, price substitutions that took effect in the fourth quarter of 2015 were based on comparisons of ASPs and AMPs from the first quarter of 2015. Exhibit 2: Timeline for AMP-Based Price Substitutions in 2015  Manufacturers collect ASPs and AMPs from the first quarter of 2015 for their drugs sold during that First quarter Quarter 2015 Manufacturers send ASPs and AMPs from the first quarter of 2015 to CMS by April 30, 2015 Second  CMS sends first quarter 2015 ASP and AMP data to OIG by end of June 2015 Quarter 2015  OIG identifies the drugs that meet the price-substitution criteria and provides them to CMS by August 15, 2015 Third Quarter  CMS publishes Part B drug reimbursement rates for the fourth quarter of 2015, including price substitution for drugs that met the criteria based on data from the first quarter of 2015 2015  CMS uses AMP-based reimbursements for Part B drugs that met criteria based on data from the first Fourth quarter of 2015 Quarter 2015 Medicare Part B Drug Payments: Impact of Price Substitutions Based on 2015 Average Sales Prices 2 OEI-03-17-00360 OIG Monitoring of ASPs and AMPs To comply with its statutory mandate, OIG has provided CMS with pricing comparisons since the January 2005 implementation of the ASP reimbursement methodology for Part B drugs. OIG issued six annual public reports for the years prior to CMS’s April 2013 implementation of the AMP price-substitution policy. These reports estimated that Medicare and its beneficiaries would have saved $35 million if CMS implemented the AMP price substitutions. OIG’s 2013 annual report was the first to provide annual savings that were a direct result of CMS’s price-substitution policy. CMS’s price substitutions based on 2013 and 2014 data saved Medicare and its beneficiaries an estimated $37 million. RESULTS CMS’s price-substitution policy saved Medicare and its beneficiaries $5.4 million over 1 year CMS initiated price substitutions for 13 drugs based on 2015 data.17 Price substitutions for these drugs saved Medicare and its beneficiaries $5.4 million over the 1-year period between the fourth quarter of 2015 and the third quarter of 2016. Exhibit 3 lists the 13 drugs, the quarter(s) during which the price substitution(s) occurred, and the savings. Exhibit 3: Drugs With Price Substitutions Based on 2015 Data Quarter(s) in Which Price Substitutions Occurred Drug Description Fourth First Second Third Savings Quarter Quarter Quarter Quarter 2015 2016 2016 2016 J1110 Dihydroergotamine mesylt injection   $9,236 J1450 Fluconazole  $154 J1570 Ganciclovir sodium injection   $47,709 J1740 Ibandronate sodium injection  $017 J2150 Mannitol injection  $4,968 J2675 Progesterone injection     $279 J3415 Pyridoxine HCl injection  $10,620 J7520 Sirolimus oral    $5,270,778 J9178 Epirubicin HCl injection  $1,387 J9200 Floxuridine injection    $19,874 J9209 Mesna injection    $40,070 Q0166 Granisetron HCl oral  $1,089 Q0167 Dronabinol oral    $16,433 Total $5,422,597 Source: OIG analysis of ASP and AMP data from 2015 Medicare Part B Drug Payments: Impact of Price Substitutions Based on 2015 Average Sales Prices 3 OEI-03-17-00360 Expanding the price-substitution criteria could have generated up to $17 million in additional savings for Medicare and its beneficiaries CMS has maintained a cautious approach to price substitutions. However, this cautious approach may restrict the Government’s ability to limit potentially excessive payment amounts based on ASPs. If CMS had expanded its price-substitution criteria to include certain other Part B drugs in 2015, Medicare and its beneficiaries could have saved up to an additional $17 million over 1 year. Millions could be saved by expanding the substitution criteria to include drugs that exceeded the 5-percent threshold in a single quarter. Nineteen drugs with complete AMP data exceeded this threshold in at least 1 quarter of 2015 but were not eligible for price substitution in that quarter because they did not meet CMS’s duration criteria, i.e., did not exceed the threshold in the 2 previous quarters or 3 of the previous 4 quarters.18 If the 19 drugs had been eligible for price reductions on the basis of data from a single quarter only, Medicare and its beneficiaries could have saved up to an additional $10.5 million between the fourth quarter of 2015 and the third quarter of 2016. Previously, CMS has expressed concern that price substitutions based on results from a single quarter may represent 1 aberrant quarter of pricing rather than a market trend.19 However, price discrepancies for the majority of the 19 drugs do not appear to have resulted from isolated fluctuations. According to 2014 and 2015 data, 14 of these 19 drugs exceeded the 5-percent threshold more than once over the 2-year period.20 Over the 2-year period, 7 of the 19 exceeded the threshold in 2 of the 8 quarters, and another 7 drugs exceeded the threshold three or more times in the 8 quarters. If CMS would prefer to employ a more cautious approach than substitution based on a single quarter of data, it could expand its price-substitution criteria to include drugs that exceed the 5-percent threshold in 2 of the previous 6 quarters. Under this approach, 9 drugs would have been eligible for price substitutions. Medicare and its beneficiaries could have saved an estimated $4.3 million on these 9 drugs. CMS also could expand the price-substitution criteria by using a proxy for missing AMP values. Currently, CMS will consider lowering reimbursement amounts only when corresponding AMP data is available for each of the NDCs used to determine the published reimbursement amount for a drug. We used wholesale acquisition costs (WACs) as the proxy for missing AMP values. According to our analysis of 2015 data, WACs function as a conservative proxy for missing AMP values because at least 94 percent of WACs were higher than or equal to AMPs each quarter. WACs are commonly higher than AMPs. Therefore, substituting WACs for AMPs affords a more conservative calculation because the difference between the ASP and the WAC would most likely be less than the difference between the ASP and the missing AMP. We identified two drugs that would have met CMS’s substitution criteria if WACs had been used in place of the missing AMP data.21 Between the fourth quarter of 2015 and third quarter of 2016, Medicare and its beneficiaries would have saved $6.4 million on these two drugs if the price-substitution criteria had been expanded to allow WACs to be used for missing AMP values. CONCLUSION AND RECOMMENDATION Under the current price-substitution policy, 13 drugs were subject to reimbursement reductions on the basis of data from 2015, saving Medicare and its beneficiaries $5.4 million between the fourth quarter Medicare Part B Drug Payments: Impact of Price Substitutions Based on 2015 Average Sales Prices 4 OEI-03-17-00360 of 2015 and the third quarter of 2016. CMS could achieve even greater savings for Medicare and its beneficiaries by expanding its criteria for AMP-based price substitutions. OIG has previously recommended that CMS expand the price-substitution criteria. CMS indicated that it does not concur with expanding the price-substitution policy and continues to believe that more experience with this policy is needed before it can be expanded. CMS also expressed concern that expanding price-substitution criteria may impede physician and beneficiary access to drugs. OIG agrees that access to prescription drugs should always be considered when contemplating pricing policies and supports current safeguards to prevent substitutions for drugs that are identified by FDA as being in short supply. However, OIG continues to believe that CMS can achieve a better balance between safeguarding access to drugs and ensuring that Medicare and its beneficiaries do not overpay for drugs. To provide greater flexibility and achieve this continued balance, any future expansion of the payment-substitution policy could contain a provision that would prevent a price substitution when there are indications that the substitution amount would be below provider acquisition costs. Therefore, we continue to recommend that CMS:  Expand the price-substitution policy To more effectively limit excessive payment amounts based on ASPs and to generate greater savings for Medicare and its beneficiaries, CMS should consider broadening its price-substitution criteria to include at least some additional drugs. A more expansive policy might include drugs with complete AMP data that exceed the 5-percent threshold in a single quarter. However, CMS also could consider a more modest expansion of the policy that better captures drugs that repeatedly exceed the threshold. For example, CMS could expand the criteria to include drugs with complete AMP data that exceed the 5-percent threshold in 2 of 6 quarters. AGENCY COMMENTS AND OFFICE OF INSPECTOR GENERAL RESPONSE CMS did not concur with our recommendation. However, CMS stated that as additional data becomes available and as it continues to gain experience with the price-substitution policy, it will consider further changes as necessary. CMS believes the current policy safeguards—which identify drugs that exceed the 5-percent threshold for 2 consecutive quarters or 3 of 4 quarters—identify situations in which AMP consistently exceeds ASP. OIG continues to believe that expanding the policy can achieve a balance between safeguarding access to drugs and ensuring that Medicare and its beneficiaries do not overpay for drugs. Our examination of 2015 data shows that if the policy had been expanded to include the 19 drugs that exceeded the 5-percent threshold in a single quarter, up to an additional $10.5 million could have been saved by beneficiaries and the program. The majority of the 19 drugs we identified for these potential savings exceeded the threshold multiple times over a 2-year period. Expanding the policy to capture drugs that exceed the threshold in a single quarter could increase the savings to beneficiaries and the program and still ensure access to drugs. Medicare Part B Drug Payments: Impact of Price Substitutions Based on 2015 Average Sales Prices 5 OEI-03-17-00360 To help ensure that CMS has sufficient information for its consideration regarding the price-substitution policy, OIG will continue to provide CMS with the results from our quarterly pricing comparisons, along with annual reports on the impact of the price-substitution policy. For the full text of CMS’s comments, see Appendix B. Medicare Part B Drug Payments: Impact of Price Substitutions Based on 2015 Average Sales Prices 6 OEI-03-17-00360 APPENDIX A: METHODOLOGY Data Collection and Analysis We obtained NDC-level ASP data, AMP data, and WAC data for Part B drugs from CMS for 2015. We also obtained ASP-based reimbursement amounts and Part B drug utilization for the quarters in which price substitutions occurred, i.e., the fourth quarter of 2015 through the third quarter of 2016. In addition, we obtained the drugs that had price substitutions based on data from 2015. For each quarter of 2015, we calculated the volume-weighted AMP for drugs consistent with CMS’s methodology for calculating volume-weighted ASPs. We then compared the volume-weighted ASPs and AMPs and identified all drugs with ASPs that exceeded the AMPs by at least 5 percent. We also identified drugs that exceeded the 5-percent threshold but did not meet CMS’s duration criteria for price substitution, i.e., they did not exceed the threshold in the 2 previous quarters or 3 of the previous 4 quarters. In addition, we identified drugs with missing AMP values, i.e., drugs that had AMP data for only some of the NDCs that CMS uses to calculate the volume-weighted ASP needed to determine the published reimbursement amount. To calculate the savings associated with price substitutions or potential price substitutions that could be made by expanding the policy, we first reduced AMP-based and ASP-based reimbursement amounts (103 percent of the volume-weighted AMP and 106 percent of the volume-weighted ASP, respectively) by the 2-percent reduction required by sequestration legislation. We then subtracted the AMP-based reimbursement amount from the ASP-based reimbursement amount for the quarter in which the price substitution occurred22 and multiplied the difference by the Part B utilization for each drug in the respective quarter that the price substitution occurred. For drugs with missing AMP data, we followed the same steps to calculate savings, but we substituted the manufacturer-reported WAC in place of each missing AMP value. Limitations We did not verify the accuracy of manufacturer-reported ASP and AMP data, nor did we verify the underlying methodology used by manufacturers to calculate ASPs and AMPs. We also did not verify the accuracy of CMS’s calculations of Part B drug reimbursement amounts. Manufacturers are required to submit their quarterly ASP and AMP data to CMS within 30 days after the close of the quarter. We did not determine whether manufacturers provided any updated data to CMS at a later date. Standards This study was conducted in accordance with the Quality Standards for Inspection and Evaluation issued by the Council of the Inspectors General on Integrity and Efficiency. Medicare Part B Drug Payments: Impact of Price Substitutions Based on 2015 Average Sales Prices 7 OEI-03-17-00360 APPENDIX B: AGENCY COMMENTS Medicare Part B Drug Payments: Impact of Price Substitutions Based on 2015 Average Sales Prices 8 OEI-03-17-00360 ENDNOTES 1 Section 1847A(d)(2)(B) of the Act. 2 Section 1847A(d)(3) of the Act. 3 Pursuant to § 1847A(d)(3)(B)(ii) of the Act, the threshold percentage has been maintained at 5 percent. 4 A HCPCS code for a drug defines the drug name and the amount of the drug represented by the HCPCS code but does not specify the manufacturer or package size. 5 Section 1927(b)(3) of the Act. 6 An NDC is a drug code that represents a specific manufacturer, product, and package size. 7 Section 1847A(c) of the Act. Certain types of sales are exempted from ASP, and ASP is net of any price concessions (with limited exceptions). 8 Section 1847A(b)(1) of the Act. Medicare beneficiaries are responsible for 20 percent of this amount in the form of coinsurance. 9 Part B claims dated on or after April 1, 2013, incur a reduction in payment in accordance with the Budget Control Act of 2011 and the American Taxpayer Relief Act of 2012 (see CMS Medicare FFS Provider e-News, Mandatory Payment Reductions in the Medicare Fee-for-Service (FFS) Program –“Sequestration,” March 8, 2013). Under this mandatory payment reduction, the payment rate for most Part B drugs is reduced by 2 percent. This reduction does not apply to the coinsurance portion of the Medicare allowed amount for Part B drugs. 10 Section 1927(b)(3) of the Act. 11 Section 1927(k)(1) of the Act. 12 42 CFR § 414.904(d)(3). 13 Ibid. 14 Ibid. 15 Ibid. 16 Ibid. 17 In the October 2015 ASP Pricing File available on CMS’s website, CMS reported an AMP price substitution for code J1740. However, the actual price that was provided in that file and used for reimbursement was the ASP plus 6 percent, not the substitution price of 103 percent of AMP. Therefore, we did not calculate any savings for code J1740. 18 These 19 drugs were not identified by FDA as being in short supply and did not have AMP-based substitution amounts that were greater than the ASP-based reimbursement amounts in the quarters during which the substitutions would have occurred. One of these drugs did not have any allowed Part B utilization during the reviewed period; therefore, the estimated savings for this drug was $0. 19 76 Fed. Reg. 73026, 73288 (Nov. 28, 2011). 20 This analysis is based on pricing comparison results for the 2-year period between the first quarter of 2014 and the last quarter of 2015. 21 WAC is defined in § 1847A(c)(6)(B) of the Act. The WAC is the manufacturer’s list price for the drug or biological to wholesalers or direct purchasers in the United States for the most recent month for which the information is available. 22 AMP-based price substitutions based on data from the first through fourth quarters of 2015 were applied in the fourth quarter of 2015 through the third quarter of 2016, respectively. Medicare Part B Drug Payments: Impact of Price Substitutions Based on 2015 Average Sales Prices 9 OEI-03-17-00360 ACKNOWLEDGMENTS Conswelia McCourt served as the team leader for this study. Office of Evaluation and Inspections staff who provided support include Joe Chiarenzelli, Althea Hosein, and Christine Moritz. This report was prepared under the direction of Linda Ragone, Regional Inspector General for Evaluation and Inspections in the Philadelphia regional office, and Edward Burley, Deputy Regional Inspector General. To obtain additional information concerning this report or to obtain copies, contact the Office of Public Affairs at Public.Affairs@oig.hhs.gov. Medicare Part B Drug Payments: Impact of Price Substitutions Based on 2015 Average Sales Prices 10 OEI-03-17-00360