1001 — Do Financial Incentives for Guideline Adherence Improve Care of Hypertension in the VA Primary Care Setting? A Multi-Site Randomized Trial
Petersen LA, Simpson K, Pietz K, Urech T, Lutschg MZ, Hysong SJ, and Profit J, Houston VA HSR&D Center of Excellence and Baylor College of Medicine; Petzel R, Under Secretary for Health, Department of Veterans Affairs; Woodard LD, Houston VA HSR&D Center of Excellence and Baylor College of Medicine;
Few data exist regarding the effectiveness of pay-for-performance programs in improving quality of care. Using a cluster randomized controlled trial, we evaluated the effects of financial incentives to health care personnel for delivering guideline-recommended hypertension care.
We enrolled 83 primary care physicians and 45 non-physician staff members from 12 VA primary care clinics. Participants at each hospital clinic were cluster-randomized to one of four study arms: (1) physician-level incentives; (2) practice group-level incentives; (3) combined physician- and group-level incentives; or (4) control. Prior to intervention commencement, participants received an educational session on guideline-recommended hypertension care. At the end of each of five 4-month periods, all participants received feedback on the proportion of their patients receiving guideline-recommended antihypertensive medications and the proportion with controlled blood pressure (BP) or who received an appropriate clinical response to elevated BP. Intervention arm participants received payments commensurate with their performance. We performed a repeated measures, complete case analysis with longitudinal mixed models using data from the 77 physicians who participated in at least two performance periods.
Compared to control arm physicians, the rate of change in the proportion of patients achieving BP control or receiving an appropriate response to uncontrolled BP was 1.62 percentage points greater per period for physicians in the combined arm (p = .022) and 2.47 percentage points greater per period for physicians in the individual arm (p <.001). In other words, a typical study physician with a panel size of 1,000 patients would be expected to have about 65 additional patients in the combined arm or 99 additional patients in the individual arm achieving this measure after one year of exposure to the intervention. The average total payment for a physician participant was $2,672 in the individual arm and $4,270 in the combined arm.
Patients of physicians who received modest financial incentives in addition to audit and feedback for delivering guideline-recommended hypertension care had greater rates of improvement in the composite measure, BP control or appropriate response to uncontrolled BP, than patients of physicians who received audit and feedback only.
If properly designed, performance-based health care payment models could improve the quality of care.