DESCRIPTION (provided by applicant): Evaluating BLS' Pharmaceutical Price Indexes and Impact on CMS' Rx Cost Forecasts The U.S. Bureau of Labor Statistics publishes a monthly Consumer Price Index for Prescription Drugs (CPI-Rx) designed to cover the sum of out-of-pocket costs to consumers plus third party payments to a pharmacy for a representative prescription, and a Producer Price Index for Pharmaceuticals for Human Use - Prescription (PPI-Rx) that measures changes over time in the average selling prices received by domestic manufacturers of prescription pharmaceuticals, excluding imports from other countries and Puerto Rico. The proposed research examines CPI-Rx and PPI-Rx sampling procedures, and the ability of these indexes to track average price changes as brands face generic entry, and an ever smaller share of prescriptions is paid for by cash (uninsured) transactions. Measuring pharmaceutical prices is challenging, in part because when the brand's patents expire, generic manufacturers enter, and as price competition intensifies, prices tend to drop dramatically. Almost 80% of prescriptions are currently dispensed as generics. Legal and regulatory developments in the last decade have complicated price index measurement as generic entry is often constrained in the six months immediately following patent expiration, but is followed by massive generic entry and heightened price competition. We examine whether these price measures are biased upwards. Implementation of Medicare Part D in 2006 increased access and utilization of medicines by seniors and made the federal government a more central payer for prescription drugs. Given the large government role as purchaser, any price mismeasurement has potentially significant consequences for public sector budgeting. An upwards biased CPI-Rx may have contributed to CMS' original and recent overprojections of Medicare Part D spending, since the CPI-Rx serves as an input into CMS's budget projections. In collaboration with economists at the U.S. Bureau of Labor Statistics, we propose to compare BLS' CPI-Rx and PPI-Rx sample composition with data from a much larger set of pharmaceutical transactions tracked by IMS Health's National Prescription Audit (data from about 38,000 pharmacies and 120 mail order firms) and National Sales Perspective (1.5 billion annual transactions involving a large number of manufacturers and wholesalers). Initial research suggests the BLS is relying ever more heavily on price quotes from diminishing cash (uninsured) transactions, whose rates of price growth may differ from insured transactions prices. Price growth of domestically available prescription drugs likely differs from that of drugs
manufactured domestically, due to increased importation of low cost generics. We will examine BLS' procedures, compute average price changes under various simulated procedures, compare price index growth using alternative data sources, consider implications for CMS' spending projections, and work with the BLS in examining how these procedures could be revised, augmented or improved.