Posted on June 12, 2001 at 4:30 p.m.

BIRMINGHAM, AL — When doctors are given feedback that allows them to compare their performance to the performance of the top physicians in their field, patient care improves significantly, according to a UAB study in the June 13 issue of the Journal of the American Medical Association.

Catarina Kiefe, M.D., Ph.D., and colleagues with the University of Alabama at Birmingham conducted a randomized controlled trial to evaluate the effectiveness of using achievable benchmarks to enhance typical physician performance feedback and improve care. The Alabama physicians used in the study are part of the ongoing Ambulatory Care Quality Improvement Project designed to improve quality of care for diabetic patients by improving physician practice patterns.

“Our hope is that if we are able to improve physician performance using this benchmark system, it will follow that patient outcomes can be improved,” said Kiefe, professor of medicine in the division of preventive medicine and director of the UAB Center for Outcomes and Effectiveness Research and Education. “Because the quality measures we studied are backed by evidence and are applicable to a high proportion of diabetic patients, using the achievable benchmark in populationwide initiatives could benefit a substantial number of patients.”

Researchers examined nearly 3,000 medical records of 70 Alabama physicians caring for patients with diabetes mellitus. Using the medical records, the researchers determined how often the doctors followed guidelines of treatment set by the American Diabetic Association known to improve outcomes for diabetics. Specifically, they determined percentages for each doctor in five areas of performance: measurement of long-term blood sugar control, measurement of serum cholesterol, measurement of triglycerides, performance of an in-office foot exam and administration of a flu vaccine.

They then determined a benchmark in each of the five areas by taking the average, in essence, of the top 10 percent of doctors who performed each therapy the most often. The benchmarks are termed ABCs, or achievable benchmarks of care.

All 70 doctors were then randomly assigned to one of two groups. Thirty-five physicians, designated as the comparison group, received an improvement intervention that included chart review and individual performance feedback. The other 35 physicians, designated as the experimental group, received identical intervention plus information on the ABCs in each area.

The researchers found that the group receiving the benchmark information showed significant improvement over the group that only received the standard feedback.

“The pre-intervention numbers were quite alarming, with doctors only performing the treatments and tests an average of only 30 to 40 percent each,” Kiefe said. “But the achievable benchmarks, the performance by the top 10 percent, were in the high 80 to upper 90 percent range.”

However, following intervention, the percentages increased significantly. Patients of the doctors who received achievable benchmarks had significantly higher adjusted odds of receiving appropriate care at follow-up compared with patients of the other physicians. Kiefe said the “underlying theory is that viewing personal performance within the context of peer performance is a powerful motivator for change.”

Kiefe said the achievable benchmark tool used in this study on diabetics can be easily translated to other areas of disease. “The model is easy to use, achievable benchmarks are easy to calculate and it is essentially free,” she said. The method is already being used now as a model in other areas of disease such as heart disease.