Academic Metrics and Predictive Models for First-Attempt Residency Placement Success in a U.S. Accredited Medical School.

Resumo

Objective: This study examined the contribution of academic performance indicators across the medical education continuum to predicting first-attempt residency placement success. Methods: A predictive, non-experimental, quantitative design with logistic regression analysis was used to analyze academic records of medical students from a U.S.-accredited medical school in Puerto Rico. Predictor variables included Medical College Admission Test (MCAT) scores, grade point average (GPA), class rank, Comprehensive Basic Science Examination (CBSE), USMLE Step 1, USMLE Step 2 Clinical Knowledge (CK), and number of Electronic Residency Application Service (ERAS) attempts. Logistic regression models were applied to evaluate predictive validity. Results: USMLE Step 2 CK emerged as the strongest predictor of first-attempt residency placement. A Step 2 CK score of approximately 240 corresponded to a predicted probability of ≥90% for successful first-attempt matching. Conclusions: Academic metrics, particularly Step 2 CK, remain critical predictors of residency placement success. These findings support integrating predictive dashboards and targeted curricular interventions to enhance student readiness and institutional outcomes.