Pay-for-performance does not improve care quality in the short term in safety-net settings
Research Activities, October 2010, No. 362
A new study questions the effectiveness of pay-for-performance (P4P) programs in improving the quality of care in safety-net settings that predominantly serve Medicaid and uninsured patients. The researchers found no evidence that P4P, financial incentives to providers to improve care quality, led to substantial quality improvements in the short term in two safety-net settings they examined. For example, incentives for well-child visits led to a significant increase in well-child visits at one of the safety-net sites studied. However, a comparable increase was also noted for nephrology visits, which had not received incentives. Conversely, the lack of incentives did not cause physicians to pay less attention to non-incentivized quality measures, which also increased during the study period.
The physicians surveyed from the two safety-net settings were generally comfortable with P4P as a concept, but less certain about its role in directly motivating quality improvement, the researchers found. Also, the safety-net providers generally agreed that the challenges of meeting the needs of their underserved, complex patients competed for clinicians' time and energy to devote to P4P quality goals.
The study was conducted with two safety-net providers in the northeastern United States. At site A, a teaching hospital's Medicaid managed care plan provided services through a network of community health centers and provided incentives for reaching four quality targets: annual retinal eye exams; annual glycosylated hemoglobin (HbA1c) measurement for patients with diabetes; prescription of controller medications for patients with asthma; and six well-child visits. Site B provided safety-net care through primary care physicians in medical groups owned by the hospital, including three groups that primarily served Medicaid and uninsured patients. The incentive program focused on three quality measures related to diabetes: an annual HbA1c test; an annual low-density lipoprotein check; and an annual foot exam. Data included a survey of provider attitudes among 256 site A physicians and 156 employed by site B, interviews with key leaders at sites A and B, and clinical information on the established quality measures for each site.
The study was funded in part by the Agency for Healthcare Research and Quality (Contract No. 290-02-0006). More details are in "Pay-for-performance in safety net settings: Issues, opportunities, and challenges for the future," by Gary Young, J.D., Ph.D., Mark Meterko, Ph.D., Bert White, D.Min., and others in the March/April 2010 Journal of Healthcare Management 55(2), pp.132-141.