Integrating pay for performance with educational strategies to improve diabetes care

Improving quality at the point of care, the practice site, has become the focal point of many health quality initiatives. Practice sites vary greatly in their levels of knowledge, comfort, and willingness to embark upon quality improvement activities. The objective of this study was to improve consistency of adherence to diabetes evidence-based guidelines; to engage physicians in critical review of their practice patterns around care of diabetic patients; and to change office systems to improve care. The survey used the Diabetes Chart Review Tool for 9 quality improvement cycles at 6-month intervals. The participants were adult primary care physicians with a minimum of 170 commercial members (n = 170-331 physicians, depending on cycle). Participating physicians received a random sample of 15 patients with diabetes for whom they review their medical records to complete a diabetes questionnaire. The survey was scored and physician-specific detail and summary reports were generated. Reports were reviewed with physicians by health plan representatives. The survey was monetary incentive participation-based in early cycles, with performance-based incentives added after the third year. The main outcome measure was the rate of cases meeting specific diabetes process and outcome measures and composite adherence to guideline score. The results were a participation rate above 84% for eligible physicians (32,069 chart reviews), and steady improvement in all process and outcome measures. Adherence to diabetes clinical guidelines shows statistically significant improvement (Student's t test, P < 0.001, mean difference -1.8, confidence interval 1.9-1.7) from baseline. The program achieved significant improvement in comprehensive diabetes care at the physician practice site level. Success is attributed to engagement of physicians, actionable reports, office-based education, written action plans, and alignment with our internal disease management.