Investigating the Relationship Between a Clinical Science Composite Score and USMLE Step 2 Clinical Knowledge and...

Area of Science:

• Medical Education
• Assessment and Evaluation
• Physician Licensure

Background:

• Previous studies indicate a moderate link between individual clinical science subject exams and USMLE performance.
• Clinical science subject exams are widely used and needed to predict future performance in medical licensing exams and residency.

Purpose of the Study:

• To investigate the predictive validity of composite scores derived from clinical science subject examinations for clinical knowledge outcomes.
• To determine if composite scores enhance prediction of performance on USMLE Steps 2 CK and 3 compared to Step 1 scores alone.

Main Methods:

• Utilized data from 65,516 students who completed five major clinical science subject examinations between 2013-2017.
• Calculated composite scores using average equated percent correct scores across exam combinations.
• Employed stepwise linear and logistic regression analyses with composite scores and Step 1 scores as predictors for Step 2 CK and Step 3 outcomes.

Main Results:

• Composite scores explained a higher proportion of variance in Step 2 CK (0.62-0.65) and Step 3 (0.45-0.48) scores compared to Step 1 alone (0.52 for Step 2 CK, 0.37 for Step 3).
• Higher composite scores correlated with an increased likelihood of passing USMLE Steps 2 CK and 3.
• Composite scores demonstrated significant predictive power for both exam success and failure risk.

Conclusions:

• Composite scores derived from clinical science subject examinations are effective predictors of USMLE Steps 2 CK and 3 performance.
• These composite scores, alone or with Step 1, can identify students at risk of failing Steps 2 CK and 3, enabling timely remediation.
• The findings support the use of composite scores as valuable tools in medical student assessment and support systems.