Medicare Part B Drugs:
CMS Data Source for Setting Payments Is Practical but Concerns Remain
GAO-06-971T, Jul 13, 2006
In 2005, the Centers for Medicare & Medicaid Services (CMS), as required by law, began paying for physician-administered Part B drugs using information on the drugs' average sales price (ASP). Subsequently, CMS selected ASP as the basis to pay for a subset of Part B drugs provided at hospital outpatient departments. To calculate ASP, CMS uses price data submitted quarterly by manufacturers. GAO was asked to discuss its work on Medicare payment rates for Part B drugs. This testimony is based on several GAO products: Medicare Hospital Pharmaceuticals: Survey Shows Price Variation and Highlights Data Collection Lessons and Outpatient Rate-Setting Challenges for CMS, GAO-06-372, Apr. 28, 2006; Medicare: Comments on CMS Proposed 2006 Rates for Specified Covered Outpatient Drugs and Radiopharmaceuticals Used in Hospitals, GAO-06-17R, Oct. 31, 2005; and Medicare: Payments for Covered Outpatient Drugs Exceed Providers' Costs, GAO-01-1118, Sept. 21, 2001. Specifically, GAO's statement discusses (1) ASP as a practical and timely data source for use in setting Medicare Part B drug payment rates and (2) components of ASP that are currently unknown and implications for Medicare rate-setting.
In summary, using an ASP-based method to set payment rates for Part B drugs is a practical approach compared with methods based on alternative data sources, for several reasons. First, ASP is based on actual transactions and is a better proxy for providers' acquisition costs than average wholesale price or providers' charges included on claims for payment, neither of which is based on transaction data. Second, ASPs, which manufacturers update quarterly, offer information that is relatively timely for rate-setting purposes. In comparison, rates for other Medicare payment systems are based on data that may be at least 2 years old. Finally, using manufacturers as the data source for prices is preferable to collecting such data from health care providers, as the manufacturers have data systems in place to track prices, whereas health care providers generally do not have systems designed for that purpose. CMS lacks certain information about the composition of ASP that prompted GAO, in commenting on CMS's 2006 proposed payment rates for a subset of Part B drugs, to call ASP "a black box." Significantly, CMS lacks sufficient information on how manufacturers allocate rebates to individual drugs sold in combination with other drugs or other products; this is important, as CMS does not have the detail it needs to validate the reasonableness of the data underlying the reported prices. In addition, CMS does not instruct manufacturers to provide a breakdown of price and volume data by purchaser type--that is, by physicians, hospitals, other health providers, and wholesalers, which purchase drugs for resale to health care providers. As a result, CMS cannot determine how well average price data represent acquisition costs for different purchaser types. In particular, to the extent that some of the sales are to wholesalers that subsequently mark up the manufacturer's price in their sales to providers, the ASP's representation of providers' acquisition costs is weakened. Additionally, a sufficient empirical foundation does not exist for setting the payment rate for Medicare Part B drugs at 6 percent above ASP, further complicating efforts to determine the appropriateness of the rate. Given these information gaps, CMS is not well-positioned to validate the accuracy or appropriateness of its ASP-based payment rates.