An investigation into gender bias in the evaluation of orthopedic trainee arthroscopic skills

Abstract

Background: Women surgeons receive lower compensation, hold fewer academic positions, and hold fewer leadership positions than men, particularly in orthopedic surgery. Gender bias at the trainee level has been demonstrated in various surgical subspecialties, but there is a lack of information on gender bias within the orthopedic training environment. This study investigated whether implicit gender bias is present in the subjective evaluation of orthopedic trainee arthroscopic skills.

Methods: After institutional review board approval, a web-based survey was sent to American Shoulder and Elbow Surgeons (ASES) members via the society's email listserve. Study participants were informed that the study was being done to develop a systematic evaluation method for trainees. The survey randomized participants to view and evaluate a prefellowship and a postfellowship video of "Rachel" (she/her) or "Charles" (he/him) performing a 15-point diagnostic shoulder arthroscopy. The videos for Rachel and Charles were identical except for the pronouns used in the video. Participants evaluated the trainee's skill level using questions from the Arthroscopic Surgical Skill Evaluation Tool (ASSET). Blinded and deidentified additional comments regarding the trainee's skill were classified as positive, negative, or neutral. Statistical analyses were used to compare scores and comments between Rachel and Charles.

Results: Of 1115 active ASES members, 181 ASES members started the survey and 106 watched both videos and were included in the analysis. Of the 106 participants completing the survey, 96 (91%) were men and 10 (9%) were women with a median (interquartile range) age of 44 (38-51). A teaching role was reported by 84 of 106 participants (79%). There was no significant difference between prefellowship scores (P = .87) or between postfellowship scores (P = .84) for the woman and man fellow. The numbers of comments classified as positive, negative, or neutral were not significantly different between the man and woman fellow (P = .19). Participants in teaching roles gave significantly lower scores to both fellows at both time points (P = .04), and participants who had fellow trainees were more likely to give negative comments to both fellows (P = .02).

Discussion: Trainee gender did not influence the ratings and comments participants gave for trainee arthroscopic skills, suggesting that gender bias may not play a major role in the evaluation of arthroscopic skill during orthopedic training.

Keywords: Bias; arthroscopy; education; evaluation; gender; shoulder; surgical training.