In my paper, I indicated that the recertification databases developed by certifying boards may be of limited use for pay-for-performance programs "some certification programs measure performance at 10-year intervals." I appreciate Dr. Cassel's note that the ABIM's MOC program affords physicians the opportunity to assess their clinical performance regularly throughout a 10-year recertification period. If the recent trend for health plans to recognize these assessments in their reward programs spreads more widely and many more physicians participate, my skepticism in this regard may prove unfounded.

I concur with Dr. Sinsky's view regarding the gulf between promise and practice in health information technology. However, some organizations, such as the Veterans Health Administration and the Montefiore Medical Center, Bronx, New York, have made real progress. Dr. Sinsky questions my focus on individual physicians, because quality is often driven at the systems or organizational level. Because most U.S. physicians practice in solo or very small groups, a primary focus on the clinical performance of individual physicians is warranted.

Dr. Sinsky disagrees with my view that referring additional patients to specialists, such as orthopedists and cardiac surgeons, will increase physician income disproportionately. Although these patients require as many resources as others, my experience as chief executive officer of a large medical center leads me to conclude that extra "marginal" patients add significant financial value once the fixed costs of operating the practice have been covered.

I concur with Dr. Krishnamurthy regarding the critical importance of patient factors in the effectiveness of pay-for-performance efforts. Physicians must not be penalized for performance measures that are determined more by characteristics of the patients they choose to serve than by their efforts.

I agree with Dr. Zwelling on the need for more evidence to support the establishment of standards against which the clinical performance of physicians is measured. A glaring deficiency is the lack of suitable information on the comparative effectiveness of various medications, devices, and procedures. The relevant clinical data base is growing rapidly, and we should use the information currently available rather than waiting for perfection of the evidence.

I do not agree that the lack of additional money makes pay-for-performance a ruse. Where in the logic for pay-for-performance is the promise of increased overall payments for physicians? As I indicated in my paper, pay-for-performance is ultimately about paying higher-performing physicians more and paying lower-performing physicians less.