RESUMO
Problem: Medical educators increasingly champion holistic review. However, in U.S. residency selection, holistic review has been difficult to implement, hindered by a reliance on standardized academic criteria such as board scores. Masking faculty interviewers to applicants' academic files is a potential means of promoting holistic residency selection by increasing the interview's ability to make a discrete contribution to evaluation. However, little research has directly analyzed the effects of masking on how residency selection committees evaluate applicants. This mixed-methods study examined how masking interviews altered residency selection in an anesthesiology program at a large U.S. academic medical center. Intervention: During the 2019-2020 residency selection season in the University of Pennsylvania's Department of Anesthesiology & Critical Care, we masked interviewers to the major academic components of candidates' application files (board scores, transcripts, letters) on approximately half of interview days. The intent of the masking intervention was to mitigate the tendency of interviewers to form predispositions about candidates based on standardized academic criteria and thereby allow the interview to make a more independent contribution to candidate evaluation. Context: Our examination of the masking intervention used a concurrent, partially mixed, equal-status mixed-methods design guided by a pragmatist approach. We audio-recorded selection committee meetings and qualitatively analyzed them to explore how masking affected the process of candidate evaluation. We also collected independent candidate ratings from interviewers and consensus committee ratings and statistically compared ratings of candidates interviewed on masked days to ratings from conventional days. Impact: In conventional committee meetings, interviewers focused on how to reconcile academic metrics and interviews, and their evaluations of interviews were framed according to predispositions about candidates formed through perusal of application files. In masked meetings, members instead spent considerable effort evaluating candidates' "fit" and whether they came off as tactful. Masked interviewers gave halting opinions of candidates and sometimes pushed for committee leaders to reveal academic information, leading to masking breaches. Higher USMLE Step 1 score and higher medical school ranking were statistically associated with more favorable consensus rating. We found no significant differences in rating outcomes between masked and conventional interview days. Lessons learned: Elimination of academic metrics during the residency interview phase does not straightforwardly promote holistic review. While critical reflection among medical educators about the fairness and utility of such metrics has been productive, research and intervention should focus on the more proximate topic of how programs apply academic and other criteria to evaluate applicants.