Association Between Racial and Ethnic Diversity in Medical Specialties and Residency Application Rates.

Importance: The lack of racial and ethnic diversity in the US medical profession is a well-recognized problem, and racial and ethnic representation is highly variable across the medical specialties. Residency selection is a crucial juncture at which diversity and representation in specialties can be increased. Objective: To identify factors associated with residency application rates for medical specialties by race and ethnicity. Design, Setting, and Participants: This national cross-sectional study of medical student residency applications used American Association of Medical Colleges data on 2019-2020 applicants and information about the racial and ethnic characteristics of practicing physicians (including medical school faculty) and department chairs. A total of 26 320 applicants to medical residency programs, 592 296 practicing physicians, and 2121 department chairs across the US were included. Residency application rates for 18 medical specialties were evaluated. Main Outcomes and Measures: The main outcome was the specialty representation quotient (SRQ), which estimated the extent to which students from a racial or ethnic group were overrepresented (an SRQ >1) or underrepresented (an SRQ <1) in a given specialty compared with the racial and ethnic demographic characteristics of the corresponding graduating class. Covariates included the racial and ethnic demographic characteristics of practicing physicians and department chairs by specialty based on American Association of Medical Colleges data and student academic factors (mean United States Medical Licensing Examination step 1 score, number of research experiences, and AΩA honor society membership among matched students from the previous application cycle). Multivariable logistic regression analysis was used to examine associations between these covariates and application rates by race and ethnicity. Results: Among 26 320 specialty-specific applications to medical residency programs in 18 specialties, 90 (0.3%) were from American Indian or Alaska Native students, 6718 (25.5%) were from Asian students, 2575 (9.8%) were from Black students, 1896 (7.2%) were from Hispanic students, and 15 041 (57.1%) were from White students. Among 592 296 practicing physicians, 2777 (0.5%) were American Indian or Alaska Native, 117 358 (19.8%) were Asian, 36 639 (6.2%) were Black, 41 071 (6.9%) were Hispanic, and 394 451 (66.6%) were White. Among 2121 department chairs, 5 (0.2%) were American Indian or Alaska Native, 212 (10.0%) were Asian, 86 (4.1%) were Black, 88 (4.1%) were Hispanic, and 1730 (81.6%) were White. The specialties with the greatest representation among applicants from racial and ethnic groups underrepresented in medicine (URM) were family medicine (SRQ, 1.70), physical medicine and rehabilitation (SRQ, 1.60), and obstetrics and gynecology (SRQ, 1.47). The specialties with the lowest URM representation among applicants were plastic surgery (SRQ, 0.47), otolaryngology (SRQ, 0.53), and orthopedic surgery (SRQ, 0.86). Membership in AΩA was negatively associated with SRQ among American Indian or Alaska Native students only (β = -0.11; 95% CI, -0.17 to -0.05; P = .002). Racial and ethnic representation among practicing physicians was positively associated with SRQ for American Indian or Alaska Native students (β = 6.05; 95% CI, 4.26-7.85; P < .001), Asian students (β = 0.07; 95% CI, 0.06-0.09; P < .001), Black students (β = 0.10; 95% CI, 0.06-0.15; P < .001), and URM students overall (β = 0.05; 95% CI, 0.01-0.08; P = .02). Conclusions and Relevance: This study's findings suggest that the propensity of medical students, particularly those from racial and ethnic minority groups, to apply to a given specialty for residency was associated with the representation of their racial or ethnic group among the specialty's practicing physicians. Future work to characterize the mechanisms of occupational sorting may guide interventions to improve equity within the physician workforce.