Race/Ethnicity Trends Among US Internal Medicine Residency Applicants and Matriculants: A Cross-Sectional Study

Background: Previous research has detailed the impor-tance of a racially and ethnically diverse physician workforce in improving access to care, communication, patient satisfaction, and health outcomes, particularly for underserved and systemi-cally marginalized patients (1). Despite this need, disparities in representation within medicine, including internal medicine (IM), persist. Many identities fall under the umbrella of diversity; here, references to diversity pertain to race/ethnicity. Objective: To elucidate trends in representation for IM resi-dency applicants and matriculants who identify as underrepre-sented in medicine (UIM). Methods and Findings: Data about applicants and matriculants to Accreditation Council for Graduate Medical Education IM resi-dency programs between July 2010 and July 2018 were acquired from the Association of American Medical Colleges. Institutional review board review was not required for the deidentified, publicly available data. Analysis was performed using Prism (version 9.2.0; GraphPad Software) and Excel (version 16.43; Microsoft). Between 2010 and 2018, a total of 214656 unique persons applied to IM residency programs and 87489 matriculated; 28222 of the applicants (13.2%) and 9269 of the matriculants (10.6%) identified as a UIM race/ethnicity (Figure, A). The proportion of aggregate UIM applicants grew minimally but significantly (slope, 0.34 [95% CI, 0.29 to 0.39]; Figure, B). On stratified analysis, a statistically significant increase was seen only for applicants who were Black or African American (slope, 0.11 [CI, 0.05 to 0.17]) and those who were Hispanic, Latino, or of Spanish origin (slope, 0.22 [CI, 0.15 to 0.29]). Trends for the proportion of aggregate UIM matriculants similarly had minimal but significant growth (slope, 0.11 [CI, 0.02 to 0.20]; Figure, C). In examining disaggregated UIM matriculant data, only the proportion of matriculants who were Hispanic, Latino, or of Spanish origin significantly changed (slope, 0.13 [CI, 0.05 to 0.21]).