Medicare’s Pay-for-Performance Experiment

Pay for performance (P4P), a model of rewarding and punishing health care providers based on the quality of care delivered to patients, was recommended by the Medicare Payment Advisory Commission (MedPAC) in spring 2005.

Data for this study comes from Hospital Compare, a U.S. Department of Health and Human services database containing quality measures from 4,203 hospitals nationwide. Early analysis estimates the financial impact P4P will have on hospitals.

Two different P4P models –Premier and MedPAC – were applied to hospitals after linking quality and payment data for heart attack, heart failure, and pneumonia. The Premier method, currently implemented in the demonstration project, rewards the top 10% of hospitals with a 2% bonus on Medicare payments. Hospitals in the eightieth to ninetieth percentiles earn a 1% bonus. Those hospitals with the poorest quality receive reductions in payments such that those below the tenth percentile lose 2% and those below the twentieth percentile lose 1%. The MedPAC method would reduce all Medicare payments by 1% at the start, subsequently using those funds to provide bonuses for the hospitals with highest quality.

The Premier method results in a net increase of $9.0 million Medicare payments. Whereas the MedPAC method would not alter aggregate Medicare spending but instead result in redistribution of $140 million from the lowest quality to the highest quality hospitals.

Commentary:

These preliminary findings only reflect voluntary reporting of process measures but not outcome measures. Multivariate analysis was not performed and is needed to appreciate the true implications of payment redistribution across regions, rural vs. urban, teaching vs. nonteaching hospitals. Efforts to implement P4P might results in unintended consequences such as continued hospital specialization and continued poor performance among hospitals losing funds under the MedPAC method.

Health Affairs 2006; 25 (1) 148-162.