Predicting United States Medical Licensure Examination Step 2 clinical knowledge scores from previous academic indicators

Abstract

The use of multiple academic indicators to identify students at risk of experiencing difficulty completing licensure requirements provides an opportunity to increase support services prior to high-stakes licensure examinations, including the United States Medical Licensure Examination (USMLE) Step 2 clinical knowledge (CK). Step 2 CK is becoming increasingly important in decision-making by residency directors because of increasing undergraduate medical enrollment and limited available residency vacancies. We created and validated a regression equation to predict students’ Step 2 CK scores from previous academic indicators to identify students at risk, with sufficient time to intervene with additional support services as necessary. Data from three cohorts of students (N=218) with preclinical mean course exam score, National Board of Medical Examination subject examinations, and USMLE Step 1 and Step 2 CK between 2011 and 2013 were used in analyses. The authors created models capable of predicting Step 2 CK scores from academic indicators to identify at-risk students. In model 1, preclinical mean course exam score and Step 1 score accounted for 56% of the variance in Step 2 CK score. The second series of models included mean preclinical course exam score, Step 1 score, and scores on three NBME subject exams, and accounted for 67%–69% of the variance in Step 2 CK score. The authors validated the findings on the most recent cohort of graduating students (N=89) and predicted Step 2 CK score within a mean of four points (SD=8). The authors suggest using the first model as a needs assessment to gauge the level of future support required after completion of preclinical course requirements, and rescreening after three of six clerkships to identify students who might benefit from additional support before taking USMLE Step 2 CK.

Predicting performance on the United States Medical Licensing Examination Step 1 and Step 2 Clinical Knowledge using results from previous examinations

“…There was also a significant correlation between the scores in the USMLE Step 1 and USMLE Step 2 CK. This correlation supports prior work done by Monteiro et al,18 who reported that the USMLE Step 1 is a significant predictor of the USMLE Step 2 CK score.…”

Prediction of success at UK Specialty Board Examinations using the mandatory postgraduate UK surgical examination

“…For example, score and number of attempts at Part B MRCS were found to be significant independent predictors of both selection into higher surgical training (HST, year 3 onwards of training) and performance during HST 3,6 . Similar relationships have been observed between the written papers and clinical exams of the equivalent physician examination in the UK, the Membership of the Royal College of Physicians (MRCP), and for the Medical Licensing Examination ® (USMLE ® ) in the USA 7,8 . Performance at the USMLE ® has also been found consistently to predict performance in the American Board of Surgery Qualifying and Certifying Examinations 9,10 , and may explain why US surgical training directors rank USMLE ® performance as one of the most important factors in selecting applicants for surgical training 11 .…”

The COVID-19 Pandemic as an Imperative to Advance Medical Student Assessment: Three Areas for Change

“…Information about students’ medical knowledge achievement will still be needed and could be captured more flexibly through confirmation of minimum competence on clerkship examinations or one but not necessarily both of the Step examinations before residency, given the high correlation among scores on these examinations. 17 …”