Information within residency monthly evaluation forms at two institutions

Abstract

Periodic review of resident performance is an important aspect of residency training. Amongst allopathic residency programs, it is expected that the performance of resident physicians which can be grouped based on the ACGME core competencies, be assessed so as to allow for effective feedback and continuous improvement. Review of monthly evaluation forms for residents in the core ACGME programs at Marshall University and the University of Toledo demonstrated a wide spread in the number of Likert questions that faculty were asked to complete. This number ranged from a low of 7 in Surgery to a high of 65 in Psychiatry (both Marshall Programs). Correlation and network analysis were performed on these data. High degrees of correlations were noted between answers to questions (controlled for each resident) on these forms at both institutions. In other words, although evaluation scores varied tremendously amongst the different residents in all the programs studied, scores addressing different competencies tended to be very similar for the same resident, especially in some of the programs which were studied. Network analysis suggested that there were clusters of questions that produced essentially the same answer for a given resident, and these clusters were bigger in some of the different residency program assessment forms. This seemed to be more the rule in the residency programs with large numbers of Likert questions. The authors suggest that reducing the number of monthly questions used to address the core competencies in some programs may be possible without substantial loss of information.

Keywords

• correlation
• network analysis
• competencies
• milestones
• acgme