The effect of a 13-hour curriculum to improve residents' teaching skills: a randomized trial.

Background: Although resident physicians often teach, few trials have tested interventions to improve residents' teaching skills. A pilot trial in 2001-2002 found that 13 trained resident teachers taught better than did untrained control residents.

Objective: To determine whether a longitudinal residents-as-teachers curriculum improves residents' teaching skills.

Design: Randomized, controlled trial from May 2001 to February 2002 (pilot trial) and March 2002 to April 2003.

Setting: 4 generalist residencies affiliated with an urban academic medical center.

Participants: 62 second-year residents: 23 in the 2001-2002 pilot trial and 39 more in 2002-2003; 27 of the 39 participants were medicine residents required to learn teaching skills.

Intervention: A 13-hour curriculum in which residents practiced teaching and received feedback during 1-hour small-group sessions taught twice monthly for 6 months.

Measurements: A 3.5-hour, 8-station, objective structured teaching examination that was enacted and rated by 50 medical students before and after the intervention. Two trained, blinded raters independently assessed each station (inter-rater reliability, 0.75).

Results: In the combined results for 2001-2003, the intervention group (n = 33) and control group (n = 29) were similar in sex, specialty, and academic performance. On a 1 to 5 Likert scale, intervention residents outscored controls on overall improvement score (post-test-pretest difference, 0.74 vs. 0.07; difference between intervention and control groups, 0.68 [95% CI, 0.55 to 0.81]; P < 0.001) by a magnitude of 2.8 standard deviations and on all 8 individual stations. The intervention residents improved 28.5% overall, whereas the scores of control residents did not increase significantly (2.7%). In 2002-2003, 19 intervention residents similarly outscored 19 controls (post-test-pretest difference, 0.83 vs. 0.14; difference between intervention and control groups, 0.69 [CI, 0.53 to 0.84]; P < 0.001). Twenty-seven medicine residents required to learn teaching skills achieved scores similar to those of volunteers.

Limitations: The study was conducted at a single institution. No "real life" assessment with which to compare the results of the objective structured teaching examination was available.

Conclusions: Generalist residents randomly assigned to receive a 13-hour longitudinal residents-as-teachers curriculum consistently showed improved teaching skills, as judged by medical student raters. Residents required to participate improved as much as volunteers did.