Audit and feedback versus alternative strategies: effects on professional practice and health care outcomes (Withdrawn Paper. 1998, art no. CD000260)

Background Audit and feedback has been identified as having the potential to change the practice of health care professionals.Objectives To assess the effects of audit and feedback compared with other interventions in changing health professional practice and to assess whether the effectiveness of audit and feedback can be improved by modifying how it is done.Search strategy We searched MEDLINE up to June 1997, the Research and Development Resource Base in Continuing Medical Education, and reference lists of related systematic reviews and articles.Selection criteria Randomised trials of audit and feedback (defined as any summary of clinical performance of health care over a specified period of time) compared with other interventions. The participants were health care providers responsible for patient care.Data collection and analysis Two reviewers independently extracted data and assessed study quality.Main results Twelve studies were included involving more than 2194 physicians. Seven trials with direct comparisons were included. The targeted behaviours were the management of low haemoglobin, the delivery of preventive care services (two studies), the management of high cholesterol, the performance of cervical smears, and the ordering of diagnostic tests (two studies). From the results of four trials, there is little evidence of a measurable effect of adding a complementary intervention such as a local consensus process to audit and feedback compared to audit and feedback alone. Two of three trials that compared audit and feedback to reminders reported that reminders were more effective in improving the delivery of some preventive services.Authors' conclusions It is not possible to recommend a complementary intervention to enhance the effectiveness of audit and feedback. Reminders might be more effective than audit and feedback to improve the delivery of some preventive services but the results are not striking. Few trials have investigated the effect of varying different characteristics of the audit and feedback process. Consideration should be given to testing the effects of modifying important characteristics such as the content, source, timing, recipient and format.