Time to get serious about pay for performance.

IN 2005, THE COMMONWEALTH FUND, A HIGHLY respected health policy group, commissioned a survey of health care opinion leaders. The leaders identified pay for performance (P4P), the notion that the provision of better care should result in more payment, as the single most promising approach to improving the performance of the US health care system. The enthusiasm behind P4P was driven by a simple concept: people and institutions respond to incentives. If incentives are created that promote quality and efficiency, better care should occur. However, the intervening years have not been kind to the advocates of P4P. Since that 2005 survey, the preponderance of the evidence that has emerged suggests that P4P, at least as currently conceived, is not working. The question, then, is why? Why has this intuitively promising idea that has tremendous face validity failed to deliver measureable improvements in quality and efficiency? This question is particularly salient now, as the Centers for Medicare & Medicaid Services (CMS) embarks on Value-Based Purchasing (VBP), a national, mandatory, wide-reaching P4P program. The VBP program is based largely on the Premier Hospital Quality Incentives Demonstration (HQID), the largest hospitalbased P4P program to date. This program had minimal effects on improving adherence to process measures and no effect on patient outcomes. As the health care system moves forward on P4P, then, what lessons should be drawn from the failure of programs like Premier HQID? This Viewpoint addresses 3 major issues: incentive size, incentive structure, and metric choice.