USMLE Step 2 CK Scores and Radiology Residency Applications: Does it Affect Diversity in Selection of Trainees?

Abstract

As residency programs in diagnostic radiology aspire to broaden trainee diversity, reliance on certain criteria may affect the selection of candidates from underrepresented groups. With the conversion of reporting of the United States Medical Licensing Examination (USMLE) Step 1 scores to pass/fail, programs may rely more on numerical USMLE Step 2 Clinical Knowledge (CK) scores. The purpose of our investigation is to assess the effects of Step 2 CK scores on the selection of underrepresented minority (URM) and female candidates. Applications from United States senior allopathic medical students to a radiology residency program from the 2021-2023 National Residency Matching Program cycles were analyzed. Subjects were classified as male or female and URM or non-URM by self-identification. Step 2 CK scores were compared and the use of cutoff scores was examined for disparate effects. 1017 subjects fulfilled the entry criteria. There were 721 males and 296 females, with 164 URM and 853 non-URM candidates. Comparing males to females, there was no significant difference in the mean score (p=0.21) and no disparate effects of cutoff scores. There was a significant difference between the mean score of URM versus non-URM candidates of eight points (p<0.00011). The use of cutoffs showed a disparate effect on URM candidates with a cutoff score of 250 (average score for 2022 matched applicants) excluding 71% of URM candidates while excluding only 46% of non-URM candidates. Reliance on USMLE Step 2 CK scores to screen applications for radiology residency can disadvantage URM candidates. Females are not adversely affected.