Randomized controlled trials of continuing medical education: what makes them most effective?

INTRODUCTION: It is essential that professional standards of excellence are demonstrated in the continuing medical education (CME) curriculum and research. METHODS: This review examines 20 randomized controlled trial (RCT) studies in CME and their effect on physician performance and/or patient health care outcomes. A systematic evaluation of the 20 RCT articles was performed. The investigators of the trials were interviewed using a standardized interview schedule. Citations from science and social science publications were compiled to obtain an unobtrusive measure of the influence of the trials. RESULTS: Investigators were most often motivated to build on earlier research of others, their own earlier research, or a combination of others' earlier research and their own. The most effective educational strategies used multiple interventions, two-way communications, printed and graphic materials in person, and locally respected health personnel as educators. Statistically significant findings more often related to physician performance than to patient health care outcomes. The most effective studies were the ones in which the educational methods were cost effective, findings could be generalized to other physician groups, the studies were implemented elsewhere in multisite health care and health-related programs and had the most citations. Investigators interviewed about their RCTs provided advice for future directions of CME curriculum development and research. DISCUSSION: CME program directors should determine what physicians need to learn, should reach out to nonparticipating physicians, and should focus on relevant problem areas. These problem areas should be ones in which it is possible to make changes, particularly in patient health care outcomes.