Analysis of the 1990-2007 neurosurgery residency match: does applicant gender affect neurosurgery match outcome?

OBJECTIVE With nearly half of graduating US medical students being female, it is imperative to understand why females typically make up less than 20% of the neurosurgery applicant pool, a number that has changed very slowly over the past several decades. Organized neurosurgery has strongly indicated the desire to overcome the underrepresentation of women, and it is critical to explore whether females are at a disadvantage during the residency application process, one of the first steps in a neurosurgical career. To date, there are no published studies on specific applicant characteristics, including gender, that are associated with match outcome among neurosurgery resident applicants. The purpose of this study is to determine which characteristics of neurosurgery residency applicants, including gender, are associated with a successful match outcome. METHODS De-identified neurosurgical resident applicant data obtained from the San Francisco Fellowship and Residency Matching Service for the years 1990-2007 were analyzed. Applicant characteristics including gender, medical school attended, year of application, United States Medical Licensing Exam (USMLE) Step 1 score, Alpha Omega Alpha (AOA) status, and match outcome were available for study. RESULTS Of the total 3426 applicants studied, 473 (13.8%) applicants were female and 2953 (86.2%) were male. Two thousand four hundred forty-eight (71.5%) applicants successfully matched. USMLE Step 1 score was the strongest predictor of match outcome with scores > 245 having an OR of 20.84 (95% CI 10.31-42.12) compared with those scoring < 215. The mean USMLE Step 1 score for applicants who successfully matched was 233.2 and was 210.8 for those applicants who did not match (p < 0.001). Medical school rank was also associated with match outcome (p < 0.001). AOA status was not significantly associated with match outcome. Female gender was associated with significantly lower odds of matching in both simple (OR 0.59, 95% CI 0.48-0.72) and multivariate analyses (OR 0.57, 95% CI 0.34-0.94 CI). USMLE Step 1 scores were significantly lower for females compared to males with a mean score of 230.1 for males and 221.5 for females (p < 0.001). There was no significant difference in medical school ranking or AOA status when stratified by applicant gender. CONCLUSIONS The limited historical applicant data from 1990-2007 suggests that USMLE Step 1 score is the best predictor of match outcome, although applicant gender may also play a