Internal Medicine Resident Evaluations of Attending Physicians

This study's purpose was to assess the prevalence and characteristics of highly and poorly rated attending physicians as well as the evaluation's reliability and validity. Medicine residents evaluated attending physicians (2013-2024) on eight teaching domains (clarity, supervision, skills, decision-making, expectations, feedback, approachability and overall effectiveness). Overall teaching effectiveness was dichotomized into "highly rated" (scores of 5) and "poorly rated" (scores of ≤3). We assessed correlates of highly and poorly rated attending physicians using generalized latent and linear and mixed methods. Validity and reliability were assessed using factor analysis, Cronbach's alpha and intra-class correlation coefficients. Among 20,150 evaluations of 668 attending physicians by 814 residents, 67% were rated highly (n = 12,801) and 9% were poorly rated (n = 1,754). Highly rated attendings explained decisions (OR: 5.1, 95% CI: 3.l-7.0), were clear (OR: 3.2, 95% CI: 2.4-4.4), approachable (OR: 2.9, 95% CI: 2.2-3.9) and demonstrated skills (OR: 3.2, 95% CI: 2.4-4.3). Poorly rated attending physicians received multiple poor ratings (74%) and lower scores on all teaching domains. Ratings were reliable and consistent, but had low levels of agreement by residents. We conclude that attending physicians were commonly highly rated. While uncommon, poor ratings are better discriminators of teaching quality than highly rated ones.