Impact of blinding interviewers to written application on ranking of applicants from underrepresented minorities (URM) in the gynecologic oncology fellowship match

Abstract

<h3>Objectives:</h3> Preconceptions from application review in preparation for a professional interview may bias the professional interview. Biases may limit the access of underrepresented minorities (URM) to advanced medical training opportunities, such as subspecialty training. We evaluated whether blinding interviewers to applicants' written application was associated with difference in proportion of URM applicants ranked highly for our gynecologic oncology fellowship match compared to interviewers having reviewed written applications prior to the interviews. <h3>Methods:</h3> During our 2020 video Gynecologic Oncology interviews, we blinded one group of our interviewers to the written applications of the interviewees. Interviewers were able to visually interact with the candidates, but were blinded to the written aspects of the application, such as test scores and letters of reference. There were no opportunities to interact with the applicants before or after the interview, and interviewers submitted independent rank lists. Interviewers also did not know that they were being assessed as to ranking of URM candidates. We compared the pooled rankings of URM applicants of our blinded and non-blinded interviewers using a Medians Test and Mann-Whitney U test. <h3>Results:</h3> For 2 positions through the National Residency Program Match Process, we received 94 applications of which 18/94 (19%) were self-identified URM. Based on review of written application and minimum application criteria for our fellowship, we invited 36 applicants to interview and interviewed 30 over 6 half-day video sessions. A total of 11 of the 30 interviewees (37%) self-identified as URM. Of the eight faculty interviewers during each interview session, 3 or 4 were blinded to the written application. There was no statistically significant difference in rank order when comparing blinded to non-blinded reviewers overall. However, blinded interviewers were more likely to rank URM applicants towards the top of the rank list (p=0.05 Median Test, 0.044 Mann-Whitney U test). <h3>Conclusions:</h3> Interviewers blinded to written applications ranked URM applicants higher than interviewers who reviewed written applications prior to video interviews. Additional research is necessary to understand the etiology of this observed difference.