Mark W. Friedberg, M.D., M.P.P., Dana Gelb Safran, Sc.D., Kathryn Coltin, M.P.H., Marguerite Dresser, M.S., and Eric C. Schneider, M.D., M.Sc.
M. W. Friedberg, D. G. Safran, K. Coltin et al., "Paying for Performance in Primary Care: Potential Impact on Practices and Disparities," Health Affairs, May 2010 29(5):926–32.
Under a typical performance-based incentive program, primary care practices that serve high shares of poor and minority patients would receive lower payments than practices serving fewer such patients, according to this Commonwealth Fund–supported study. The finding adds to the growing concern that these payments, while intended to improve quality of care, may exacerbate health care disparities.
Pay-for-performance programs are increasingly used to stimulate health care providers to improve the care they deliver to patients. But even as such incentives become more common, disparities in quality based on patient race, ethnicity, and income status persist. Concerns about whether performance-based payments could undermine care for vulnerable populations and worsen disparities have prompted calls from the Institute of Medicine to monitor the effects of incentives on disparities. To simulate how performance-based payments would be distributed among physician practices serving areas that have low, medium, and high concentrations of minority and poor patients, the authors of this study applied the financial incentive model used by the Medicare Care Management Performance demonstration to primary care practices in Massachusetts.
Performance-based payments may disproportionately flow away from practices caring for minority and poor patients, thereby widening sociodemographic inequalities and potentially exacerbating disparities in the quality of care. The authors note that simulated performance-based payments were relatively small—less than $2,400 annually per physician in a median practice. But if performance-based incentives become a sizeable share of practices' total income, low payments could undermine the financial viability of providers who serve vulnerable populations, resulting in worsening access to primary care for those populations. The authors suggest several policy solutions, including stratifying performance-based incentives by sociodemographic group and targeting grants for practices that care for vulnerable populations.
The authors identified practices providing primary care by using the 2007 statewide physician directory of Massachusetts Health Quality Partners, which is used for statewide quality reporting initiatives. The demographic characteristics of the population served by each practice were identified by matching patients' residential ZIP codes to geocoded census data. To simulate performance-based payment, the authors adapted design specifications from the Medicare Care Management Performance demonstration. They used performance data for preventive and diabetes care, and added an additional cervical cancer screening measure.
To avoid shifting financial resources away from physician practices that care for vulnerable populations, pay-for-performance programs should carefully monitor and address the impact of performance-based incentives on the distribution of resources and health care disparities.