Examining the influence of gender on medical students' decision making.

Gender bias, described among practicing physicians, has rarely been examined in medical students. The current study examined the influence of gender bias on medical students' clinical decision making. We experimentally manipulated patient gender in 27 written clinical vignettes embedded in the United States Medical Licensing Examination (USMLE) Step 2 examination (a multiple-choice test of clinical decision making). Female and male patient versions of selected test cases were created within three categories: (1) diseases with previously established evidence of gender bias in the diagnosis or management of the disease, (2) diseases with a higher prevalence in a specific gender, and (3) diseases with similar prevalence in both genders and without evidence of gender bias in the literature. Among the 3059 students who wrote the USMLE Step 2 examination in August 1998, there were small but significant differences in performance on the 12 gender bias cases. Students performed worse for the female patient version of the cases compared with the male patient version of the cases (mean of 55.8% correct for female cases compared with 57.7% correct for male cases) (p < 0. 01). Our data suggest that students were variably influenced by gender bias in their investigation and management of patients in a written test of clinical decision making.