Quantifying the effect of perceived gender identity on competency-based performance evaluations of postoperative scar images

Abstract

Objective: Implicit gender bias in competency-based evaluations remains a critical concern in medical education, particularly within surgical training, where assessment outcomes impact trainee progression and patient safety. This study examines whether perceived gender identity influences the evaluations of technical skill in a controlled setting, where expert and non-expert raters assessed postoperative scar images labeled with male faces, female faces, or no identity markers. Methods: Expert raters were recruited from professional surgical societies and non-expert raters through the Prolific crowdwork platform. Using the Scar Cosmesis Assessment and Rating (SCAR) scale—a validated tool for measuring scar quality—the raters were randomized to one of the three identity conditions and evaluated images in the belief that a shown surgeon had performed the wound closure resulting in a postoperative scar shown on the screen. Results: No statistically significant differences were found in SCAR scores between the gender-labeled conditions, indicating an absence of observable bias among both expert raters (n=614) and non-expert raters (n=4,181) when assessing technical skills. There was no evidence of gender bias when completing subgroup analysis based on the demographics of the raters in either population. Conclusions: The consistency of results across expert and non-expert raters suggests that both rating methods can be unbiased and objective within a controlled evaluation environment. The use of a clearly defined task-specific scale could be a potential strategy to combat gender bias in the evaluation of surgical skills. The substantial difference in recruitment times between the two study populations underscores the need for long-term scalable solutions that allow consistent evaluations of technical skills in competency-based medical education settings.