We used several sources to identify candidate private- and public-hospital P4P programs to construct the universe of hospital P4P programs:
- The published literature on P4P programs (e.g., CMS PHQID).
- The Med-Vantage annual survey of P4P programs (2006) and a review of our candidate list by Med-Vantage staff who had conducted the annual survey.
- Information provided by research and policy staff within leading professional organizations, including the Association of Health Insurance Plans (AHIP), AHA, Blue Cross/Blue Shield Association (BCBSA), and Joint Commission.
- The Leapfrog Compendium of incentive and reward programs (Leapfrog Group, 2007).
- A Lexis/Nexis search of major U.S. newspapers, a broad Google-based Internet search, and a search of relevant trade journals.4
- The knowledge accumulated by RAND project staff who have been directly involved in evaluating a number of P4P demonstrations, and
- Input from the project’s Technical Expert Panel (TEP), some of whose members currently operate or are involved with P4P programs.
From this scan, we identified 41 candidate organizations thought to sponsor hospital P4P programs. We then cataloged the 41 programs by a range of characteristics (e.g., type of sponsor, geographic region, type of insurance product) and selected a subset of hospital P4P program sponsors for discussions. During the selection process, we attempted to include a broad cross-section of programs that would encompass the range of variation in program design and operation. The goal of pursuing this strategy, as contrasted with a pure random sample, was to provide a rich base of information for consideration by ASPE and CMS.
The characteristics we sought to balance in our purposive approach to sampling were:
- The inclusion of a broad array of sponsor types, such as single organization sponsors, multi-stakeholder coalitions, private- versus public-sector sponsors.
- The inclusion of different types of insurance products, such as health maintenance organizations (HMOs), preferred provider organizations (PPOs), point of service (POS), administrative services only (ASO), Medicare, and Medicaid.
- The programs needed to cover various geographic areas of the country because of the variation in market characteristics that could affect design.
From the 41 programs, we selected 31 organizations and requested their participation in the discussions. We held discussions with 27 of the 31 organizations between August and December 2006. Of the four organizations that did not participate, one had no hospital P4P program, one declined to participate, one never replied, and for one we were unable to establish correct contact information.
The numerical statistics presented in the following sections reflect 23 of the 27 organizations. The four organizations excluded from our tabulations were in the planning stages of designing a P4P program or were the national plan office that delegated operation of P4P programs to the local plan. We did, however, include information gathered from our conversations with these four organizations in our descriptive summaries.