Effects of training in direct observation of medical residents' clinical competence: a randomized trial.

BACKGROUND Faculty observation of residents and students performing clinical skills is essential for reliable and valid evaluation of trainees. OBJECTIVE To evaluate the efficacy of a new multifaceted method of faculty development called direct observation of competence training. DESIGN Controlled trial of faculty from 16 internal medicine residency programs using a cluster randomization design. SETTING Academic medical centers. PARTICIPANTS 40 internal medicine teaching faculty members: 17 in the intervention group and 23 in the control group. MEASUREMENTS Changes in faculty comfort performing direct observation, faculty satisfaction with workshop, and changes in faculty rating behaviors 8 months after completing the training. INTERVENTION The direct observation of competence workshop combines didactic mini-lectures, interactive small group and videotape evaluation exercises, and evaluation skill practice with standardized residents and patients. RESULTS 37 faculty members (16 in the intervention group and 21 in the control group) completed the study. Most of the faculty in the intervention group (14 [88%]) reported that they felt significantly more comfortable performing direct observation compared with control group faculty (4 [19%]) (P = 0.04), and all intervention faculty rated the training as outstanding. For 9 videotaped clinical encounters, intervention group faculty were more stringent than controls in their evaluations of medical interviewing, physical examination, and counseling; differences in ratings for medical interviewing and physical examination remained statistically significant even after adjustment for baseline rating behavior. LIMITATIONS The study involved a limited number of residency programs, and faculty did not rate the performance of actual residents. CONCLUSION Direct observation of competence training, a new multifaceted approach to faculty development, leads to meaningful changes in rating behaviors and in faculty comfort with evaluation of clinical skills.