2007 Robert and Alma Moreton lecture: pay for performance: more than rearranging the deck chairs?

Over the past 30 years, research exploring the causes and consequences of geographic variations in practice has called into question widely held assumptions about the relationship between spending and quality. Two-fold differences in spending are observed across U.S. regions that are not due to differences in illness or to the prices charged by providers. Rather, higher spending is due primarily to greater use of "supply-sensitive" services: the frequency of visits to physicians and referrals to specialists, the amount of time similar patients spend in the hospital, and the frequency of imaging, tests, and minor procedures. The paradox, however, is that greater use of these services has been shown to be associated with lower quality, no gain in survival, and worse physician and patient-reported quality of care. It may be possible, therefore, to lower spending while improving quality. But this will require addressing the underlying causes of the variations: overuse of discretionary services in a fee-for-service system that ensures that physicians stay busy and that existing capacity remains fully deployed. These findings point to 3 strategies that will be required for pay for performance to achieve its potential: fostering local organizational accountability for the overall quality and costs of care--and for the capacity of the local delivery system; adoption of comprehensive longitudinal performance measures--to reassure the public that lower spending is compatible with higher quality care; and fundamental reform of the payment system.