Notwithstanding the difficulties involved in overhauling the current payment system and the gaps and uncertainties in the existing evidence base for pay for performance, the committee believes pay for performance could be a viable tool when used in tandem with other performance improvement strategies, such as public reporting and technical assistance, to transform the health care delivery system. The ultimate purpose of a performance-based payment system is to stimulate behavioral and organizational change within the provider community in ways that will foster performance improvement and improve the value of health care services. The committee does not presume that these changes will be easy or that savings will automatically be generated. It also remains cognizant of the unintended consequences that could arise, but believes the consequences of inaction pose a greater threat.

The lack of evidence associating pay for performance with improvements in clinical quality, patient-centeredness, efficiency, and, most important, outcomes of care suggests that caution will be necessary as Medicare proceeds. The consequences of aligning incentives to promote higher-quality health care are largely unknown. For instance, providers might avoid treating high-risk, nonadherent, or other types of patients who could jeopardize their performance scores. Payment rewards might encourage a system in which providers would work to improve only on measures for which they were paid, and as a consequence, reduce their quality of care on unrewarded measures instead of trying to improve care on a more global level. From the patient’s perspective, payment incentives could undermine the revered physician–patient relationships, which are based largely on trust. A pay-for-performance program could also increase competition within the health care enterprise and thereby impede knowledge transfer among providers. Recognizing the need for some kind of payment reform and the promise of pay for performance as suggested by early experiments in the private sector, it will be necessary to balance actions taken toward implementation of the approach with due caution. This report therefore emphasizes the need to introduce pay for performance within a comprehensive learning system through a multiphase approach that addresses significant variations in clinical conditions and health care settings, and encompasses an evaluation strategy for deriving insights from the experience gained in early stages of implementation.


This report sets forth a vision for how pay for performance could be implemented to best help achieve the goals of improving health care quality and patient outcomes. While the report outlines options, designers should

The National Academies of Sciences, Engineering, and Medicine
500 Fifth St. N.W. | Washington, D.C. 20001

Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement