risk adjustment for differences in patient mix. Such comparisons, too, require richer clinical information than is currently available in most administrative datasets.
In the Medicare program, the federal government has taken some steps consistent with its purchaser role by facilitating disclosure of comparative quality data in the public domain. In 1998, the National Medicare Education Program—an initiative to educate beneficiaries about Medicare health care options—was launched. Under this program, CMS makes available on the World Wide Web limited comparative quality data for M+C plans from CAHPS and HEDIS to help beneficiaries select an M+C plan. For the nearly 87 percent of beneficiaries enrolled in Medicare FFS, the primary decisions to be made are whether to shift from FFS to an M+C plan and what clinician to select. Current information does not permit a comparison to support the former decision, because most performance data are available only for M+C plans. Few if any performance data are available to help beneficiaries choose a doctor or other clinician.
CMS provides beneficiaries with comparative data on kidney dialysis centers, as required by the Balanced Budget Act of 1998. CMS funded the development of clinical practice measures, based on the practice guidelines of the National Kidney Foundation’s Dialysis Outcome Quality Initiative and awarded the development contract to Pro-West (Centers for Medicare and Medicaid Services, 2001c). The measures were developed collaboratively with providers, and dialysis facilities were given the opportunity to review their data prior to public release (American Association of Kidney Patients, 2001). There is a strong commitment to public disclosure, and the CMS website provides a rating of dialysis centers as average, below average, or above average (Centers for Medicare and Medicaid Services, 2002a). CMS recently announced its intent to make similar comparative quality information available on nursing homes. Data from a pilot project conducted in six states (Colorado, Florida, Maryland, Ohio, Rhode Island, and Washington) using the Minimum Data Set measures were recently released (Centers for Medicare and Medicaid Services, 2001b).
At present, CMS has very limited authority to link payment to performance for traditional Medicare, other than through demonstration projects designed to test alternative purchasing approaches (MedPAC, 1999). For example, under the Centers of Excellence demonstration, Medicare contracts selectively with a limited number of hospitals or other organizations to provide comprehensive services for specific procedures (e.g., heart transplants, total joint replacement procedures) under a bundled payment scheme (Centers for Medicare and Medicaid Services, 2002b). Providers compete for these contracts on the basis of quality, as well as other factors, such as geographic accessibility, organizational ca-