Effects of Peer-Review Groups on Physicians’ Practice

Abstract

Background: Peer-review groups are often considered to be an effective means of improving routine practice at the primary health care level. In contrast, little evidence has been published proving the efficacy of peer-review groups in controlled studies.Objectives: This study evaluates the effects of peer-review groups of primary-care physicians on process quality of diabetes care, especially the screening of retinopathy, nephropathy and peripheral neuropathy.Methods: Prospective controlled study, assessment of office-based documentation of diabetes-related data in a random sample of 25 patients per practice initially and after one year in the intervention and the control group.Intervention: Primary-care physicians from ten practices participated in five sessions of a structured peer-review group on the subject of diabetes. Control: Seven primary-care physicians met once to discuss the results of the first assessment of office-based documentation of diabetes-related data.Results: Records of 239 patients in the intervention group and 164 patients in the control group could be evaluated. Initially, in the patients’ charts results of screening for diabetic foot complications could hardly be detected: In the intervention group a result of pallaesthesia screening with a tuning fork was documented in 0.4% of diabetic patients and in 4.9% in the control group. Results of screening for retinopathy were documented in 8.4 resp. 5.5%, results of screening for microalbuminuria in 1.2 vs. 8.5% per year. After intervention, relevant improvements were observed in the intervention group: Documentation of screening for pallaesthesia increased up to 35.1% (+34.7% absolute), documentation of screening for retinopathy up to 32.2% (+23.8% absolute), whereas the results in the control group remained unchanged.Conclusion: Structured peer-review groups following the described methodology are a means of changing physicians’ behaviour concerning the quality of office-based documentation. Further relevant improvements in process quality would be desirable. Other means (e.g. remuneration combined with control of process quality) should be evaluated.