Comparison of faculty, peer, self, and nurse assessment of obstetrics and gynecology residents.

This project was undertaken to determine if the addition of peer, self, and nurse evaluators would enhance faculty assessment of resident performance. An evaluation form was developed that contained 16 items under three major headings: clinical competency, interpersonal skills, and overall assessment. These forms were distributed to the 16 residents in our program, and the residents were asked to complete an evaluation form on themselves and their 15 peers. Evaluation forms were also distributed to all 14 clinical faculty members and to six members of the Obstetrics and Gynecology nursing staff. Intraclass correlation coefficients were determined for each rating group, and using the attending ratings as the gold standard, Pearson correlation coefficients were calculated to determine correlations between groups of raters. The reliability of faculty ratings was good to excellent, whereas resident ratings showed excellent reliability for all three measured variables. Ratings provided by the nurses were the least reliable of the groups rating the residents. Compared with attending ratings, the peer ratings showed significant positive linear relationships in all categories evaluated. There was a low degree of correlation between attending and self-evaluations and attending and nurse evaluations. This small study supports the use of peer evaluations in addition to attending evaluations for Obstetrics and Gynecology residents in training. It also demonstrates that our residents may benefit from doing self-evaluations to improve their ability to honestly appraise their clinical and interpersonal skills.