Predicting and Modifying Success on the NBME Surgery Subject Exam

The National Board of Medical Examiners (NBME) surgery subject exam (SSE) is a high-stakes exam with few predictors of success. We hypothesized that performance on a mid-clerkship NBME surgery self-assessment (SSA) practice test would predict performance on the SSE and could be used to selectively intervene with struggling students. Participating students completed the SSA at the mid-point and the SSE at the end of their clerkship. All students were offered a test-taking skills intervention using the framework of self-regulated learning theory with microanalytic assessment and training (SRL-MAT). The relationship between SSA and SSE scores and effect of the intervention were investigated using linear correlation, logistic regression, and difference in differences analysis. Clerkship students (n = 314) completed both the SSA and the SSE. SSA and SSE scores positively correlated (r = 0.48, p < 0.001). Students who failed the SSA were ten times more likely to fail the SSE compared with students who passed the SSA (OR 9.94, 95