Bias in Surgical Residency Evaluations: A Scoping Review

Given widespread disparities in the surgical workforce and the advent of competency-based training models that rely on objective evaluations of resident performance, this review aims to describe the landscape of bias in the evaluation methods of residents in surgical training programs in the United States.

A scoping review was conducted within PubMed, Embase, Web of Science, and ERIC in May 2022, without a date restriction. Studies were screened and reviewed in duplicate by 3 reviewers. Data were described descriptively.

English-language studies conducted in the United States that assessed bias in the evaluation of surgical residents were included.

The search yielded 1641 studies, of which 53 met inclusion criteria. Of the included studies, 26 (49.1%) were retrospective cohort studies, 25 (47.2%) were cross-sectional studies, and 2 (3.8%) were prospective cohort studies. The majority included general surgery residents (n = 30, 56.6%) and nonstandardized examination modalities (n = 38, 71.7%), such as video-based skills evaluations (n = 5, 13.2%). The most common performance metric evaluated was operative skill (n = 22, 41.5%). Overall, the majority of studies demonstrated bias (n = 38, 73.6%) and most investigated gender bias (n = 46, 86.8%). Most studies reported disadvantages for female trainees regarding standardized examinations (80.0%), self-evaluations (73.7%), and program-level evaluations (71.4%). Four studies (7.6%) assessed racial bias, of which all reported disadvantages for trainees underrepresented in surgery.

Evaluation methods for surgery residents may be prone to bias, particularly with regard to female trainees. Research is warranted regarding other implicit and explicit biases, such as racial bias, as well as for nongeneral surgery subspecialties.