Thursday, October 7, 2010
Do Group Purchasing Organizations Achieve the Best Prices for Member Hospitals? An Empirical Analysis of Aftermarket Transactions
Posted by D. Daniel Sokol
Robert E. Litan (Kauffman Foundation, Brookings Institution) and Hal Singer Navigant Economics) discuss Do Group Purchasing Organizations Achieve the Best Prices for Member Hospitals? An Empirical Analysis of Aftermarket Transactions.
ABSTRACT: Group purchasing organizations (GPOs) were formed with the purpose of securing the best prices for medical devices for their member hospitals. They ostensibly do so by, among other things, conducting auctions for the right to supply a particular device to hundreds of hospitals at once. Until they were exposed to Congressional scrutiny of their practices, GPOs often structured these supply contracts as “sole source”—that is, the winning bidder would have exclusive access to the GPO’s member hospitals to supply the device. In the past few years, GPOs have restructured their contracts to allow “multisource” contracting among multiple device makers. Whether this has resulted in competitive prices for member hospitals is the subject of the current study. Even though competition for the GPO contract may have been marginally reduced by permitting multiple firms (rather than single firm) to supply the device in question, the GPO auction process should still generate competitive bids in theory. Because GPOs are still compensated with a share of the revenues generated under the contracts, however, they may be incentivized to maintain some monopoly pricing power for the winning bidder or bidders. After all, a fixed percentage (typically twotothree percent, but as high as 18 percent) of revenues from a monopoly concession is more valuable than the same percentage of a competitive concession. To measure the extent of this potential competitive distortion, we analyzed a database of approximately 8,100 aftermarket transactions, in which the winning GPO price was put up for bid after the initial GPO auction. The transactions data suggest that, when exposed to competition in the aftermarket, hospitals were able to achieve average savings of approximately 10 to 14 percent across the entire database (2001 through 2010) and a savings of 15 percent on average for 2010 data. Indeed, in over half of all auctions in the transactions database, incumbent device makers on the GPO contract were induced to lower their own prices for the same product to the same hospital, and did so by approximately 7 percent on average. Our results are inconsistent with the hypothesis that GPOs secure the best prices for their member hospitals. One clear policy implication of this study is to modify the incentives that limit the intended procompetitive objectives of GPOs—namely, by changing the method of compensation of GPOs to reduce conflicts of interest, which could be achieved by reinstating the application of the existing Medicare antikickback statute of the 1986 Social Security Act, thereby prohibiting vendors from paying GPOs. This exemption has allowed GPOs to retain an equity interest (or its functional equivalent) in their contracts with those whom they are to negotiate for lower prices. So long as GPOs are compensated via an equity interest in the concession, they have an inherent conflict that limits their ability to negotiate the best prices for their member hospitals and those hospitals (and their payors, including the federal government) will likely continue to overpay for medical devices. Based on the results from our empirical analysis, we conservatively estimate that changing the incentive structure by reapplying the antikickback statutes would reduce private U.S. health care expenditures by roughly $25 billion annually, and would reduce federal health care spending by roughly $11.5 billion annually.