To the Editor: Holistic Approach to Resident Selection May Eliminate Potential Biases Associated With USMLE Step 1 Pass/Fail Scoring System

We read with profound interest the article by Ganesh Kumar et al¹ highlighting trainee perspectives regarding the use of a United States Medical Licensing Examination (USMLE) pass/fail score in place of the traditional 3-digit numerical scoring system for resident selection in graduate medical education (GME) residency programs. The authors surveyed a diverse cohort of residents, fellows, and medical students, inclusive of US allopathic (MD) students, US osteopathic (DO) students, and international medical graduates (IMGs) in 2020. A total of 11 633 trainees (7254 residents/fellows and 4379 medical students), 14% of which were underrepresented in medicine (UiM), responded to the survey and provided insights on this controversial subject. The authors emphasized in their discussion that, although 35% of the respondents supported the proposed USMLE Step 1 pass/fail scoring system, 44% of the survey respondents disliked this system. US allopathic medical students and UiM trainees were the most supportive of a USMLE Step 1 pass/fail scoring system. In contrast, many non-UiM, DO, and IMG respondents felt that they would be at a disadvantage. Regarding the secondary outcome measure that a USMLE Step 1 pass/fail scoring system will reduce socioeconomic disparities in resident selection, it was fascinating to read that UiM medical students primarily supported this hypothesis. We applaud the Vanderbilt University investigators for performing this study, which provides GME leaders with vital data as we transition to a USMLE Step 1 pass/fail scoring system.

In the United States, GME is at a crucial crossroads regarding resident selection. Historically, USMLE Step 1 scores have served as a benchmark for screening resident candidates, especially in highly competitive programs. Although USMLE scores may reflect some degree of inherent knowledge, the preparation required to excel on the examination may also depend on socioeconomic factors. Medical students often take multiple preparation courses, such as the PASS Program and Kaplan, to improve their content knowledge and test-taking skills for the USMLE examinations. Many of these are high cost, and in an era of rising medical school tuition costs, access may be limited to students with adequate financial resources. Students from socioeconomically disadvantaged backgrounds may have lower USMLE Step 1 scores because of reduced resources. For this and other reasons, the decision to move to a USMLE Step 1 pass/fail scoring system was made. However, as found by Ganesh Kumar et al,¹ trainees who responded to this survey had varied perspectives—a significant subset of these respondents, particularly DO and IMG trainees, were skeptical of this transformation. As a result, we caution GME leaders to be thoughtful with this new binary USMLE Step 1 pass/fail scoring system, because it may introduce additional implicit biases against DO and IMG applicants. It may be helpful to remove credentials that could encourage implicit bias, such as the MD, DO, or IMG designation when reviewing an applicant.

The critical question now becomes: How will residency admission committees identify top-level applicants in the era of a USMLE Step 1 pass/fail scoring system? We believe that using a meritocratic, holistic approach to residency admission, inclusive of letters of recommendation, the binary USMLE Step 1 pass/fail score, medical school transcripts, trainee experiences, personal attributes, interview performance, and barriers encountered during their education and training, can identify the best and brightest medical students for residency programs.² Hopefully, this holistic approach to resident selection will eliminate ingrained socioeconomic disparities that may hinder UiM students from joining competitive residency programs.