Comparison of the instructional efficacy of Internet-based CME with live interactive CME workshops: a randomized controlled trial.

CONTEXT Despite evidence that a variety of continuing medical education (CME) techniques can foster physician behavioral change, there have been no randomized trials comparing performance outcomes for physicians participating in Internet-based CME with physicians participating in a live CME intervention using approaches documented to be effective. OBJECTIVE To determine if Internet-based CME can produce changes comparable to those produced via live, small-group, interactive CME with respect to physician knowledge and behaviors that have an impact on patient care. DESIGN, SETTING, AND PARTICIPANTS Randomized controlled trial conducted from August 2001 to July 2002. Participants were 97 primary care physicians drawn from 21 practice sites in Houston, Tex, including 7 community health centers and 14 private group practices. A control group of 18 physicians from these same sites received no intervention. INTERVENTIONS Physicians were randomly assigned to an Internet-based CME intervention that could be completed in multiple sessions over 2 weeks, or to a single live, small-group, interactive CME workshop. Both incorporated similar multifaceted instructional approaches demonstrated to be effective in live settings. Content was based on the National Institutes of Health National Cholesterol Education Program--Adult Treatment Panel III guidelines. MAIN OUTCOME MEASURES Knowledge was assessed immediately before the intervention, immediately after the intervention, and 12 weeks later. The percentage of high-risk patients who had appropriate lipid panel screening and pharmacotherapeutic treatment according to guidelines was documented with chart audits conducted over a 5-month period before intervention and a 5-month period after intervention. RESULTS Both interventions produced similar and significant immediate and 12-week knowledge gains, representing large increases in percentage of items correct (pretest to posttest: 31.0% [95% confidence interval {CI}, 27.0%-35.0%]; pretest to 12 weeks: 36.4% [95% CI, 32.2%-40.6%]; P<.001 for all comparisons). Chart audits revealed high baseline screening rates in all study groups (> or =93%) with no significant postintervention change. However, the Internet-based intervention was associated with a significant increase in the percentage of high-risk patients treated with pharmacotherapeutics according to guidelines (preintervention, 85.3%; postintervention, 90.3%; P = .04). CONCLUSIONS Appropriately designed, evidence-based online CME can produce objectively measured changes in behavior as well as sustained gains in knowledge that are comparable or superior to those realized from effective live activities.