Implicit gender bias among US resident physicians

Background: The purpose of this study was to characterize implicit gender bias among residents in US Emergency Medicine and OB/GYN residencies. Methods: We conducted a survey of all allopathic Emergency Medicine and OB/GYN residency programs including questions about leadership as well as an implicit association test (IAT) for unconscious gender bias. We used descriptive statistics to analyze the Likert-type survey responses and used standard IAT analysis methods. We conducted univariate and multivariate analyses to identify factors that were associated with implicit bias. We conducted a subgroup analysis of study sites involved in a multi-site intervention study to determine if responses were different in this group. Results: Overall, 74% of the programs had at least one respondent. Out of 14,234 eligible, 1634 respondents completed the survey (11.5%). Of the five sites enrolled in the intervention study, 244 of 359 eligible residents completed the survey (68%). Male residents had a mean IAT score of 0.31 (SD 0.23) and females 0.14 (SD 0.24), both favoring males in leadership roles and the difference was statistically significant (p < 0.01). IAT scores did not differ by postgraduate year (PGY). Multivariable analysis of IAT score and participant demographics confirmed a significant association between female gender and lower IAT score. Explicit bias favoring males in leadership roles was associated with increased implicit bias favoring males in leadership roles (r = 0.1 p < 0.001). Conclusions: We found that gender bias is present among US residents favoring men in leadership positions, this bias differs between male and female residents, and is associated with discipline. Implicit bias did not differ across training years, and is associated with explicit bias.