USMLE Step 2 Clinical Knowledge performance by disability, race, and ethnicity in U.S. medical students: A multi-site study

Abstract

Importance With the transition of USMLE Step 1 to pass/fail, USMLE Step 2 Clinical Knowledge (CK) has become a high-stakes component of residency selection. This shift raises concerns that increased reliance on Step 2 CK may exacerbate existing inequities, particularly for medical students with disabilities and those with intersecting marginalized identities. Objective To examine differences in USMLE Step 2 CK performance by disability status, race, ethnicity, and their intersections among U.S. medical students. Methods We conducted a cross-sectional analysis of the Pathways II dataset, comprising deidentified student-level data from nine U.S. MD-granting medical schools. Graduated students from the 2020–2023 cohorts were included. Medical students with disabilities were defined as those formally registered with institutional disability resource offices and were matched 1:2 with nondisabled peers by institution, gender, graduation cohort, and MCAT score. Race and ethnicity were categorized as White, Asian, underrepresented in medicine (URiM, including American Indian/Alaska Native, Black/African American, Hispanic/Latino, and Native Hawaiian/Pacific Islander), or Other. Disability categories included ADHD, learning, psychological, mobility/sensory, chronic health, or other. Multivariable linear regression models estimated adjusted Step 2 CK score differences by disability status, race, ethnicity, and their interactions, adjusting for MCAT score and clustering standard errors at the school level. Results Among 1,350 students, 448 (33.2%) had a registered disability. The mean Step 2 CK score was 244.3 (SD, 15.1). Students with disabilities scored lower than nondisabled peers (mean 239.6 vs 246.7; adjusted β, −7.02; 95% CI, −8.76 to −5.28). Students with ADHD, learning, or psychological disabilities scored 9–11 points lower than nondisabled peers (P < .001). URiM students scored lower than White students overall (adjusted β, −8.50; 95% CI, −10.74 to −6.25). Intersectional analysis demonstrated that disparities were largest among URiM and Asian students with disabilities. Conclusions Medical students with disabilities experience significant disparities in Step 2 CK performance, particularly those with cognitive-based disabilities and intersecting marginalized identities. As Step 2 CK plays a larger role in residency selection, addressing structural barriers in assessment and accommodation processes is critical to advancing equity.