Impact of a USMLE Step 2 Prediction Model on Medical Student Motivations

Area of Science:

• Medical Education
• Assessment and Evaluation

Background:

• USMLE Step 1 transitioning to Pass/Fail increases the importance of Step 2 CK for residency applications.
• Early career mentoring is crucial for medical students navigating residency selection.

Purpose of the Study:

• Develop a predictive model for USMLE Step 2 CK scores.
• Assess the impact of predicted scores on medical students' career and study plans.

Main Methods:

• Utilized traditional statistical models and machine learning to identify predictors of Step 2 CK performance.
• Provided predicted scores to medical students.
• Conducted a cross-sectional survey to evaluate the influence of predicted scores on student plans.

Main Results:

• A predictive model incorporating CBSE, Organ System exam, and NBME Phase 2 scores was developed (Step2CK = 191.984 + 0.42(CBSE) + 0.294(Organ Systems) + 0.409(NBME)).
• The model had a standard error of 7.6, with better accuracy for scores above 230.
• 19% of students altered study plans; feedback included reassurance and increased anxiety.

Conclusions:

• A Step 2 Predicted Score, derived from existing metrics, accurately estimates Step 2 CK performance.
• This predictive tool can aid in specialty and residency selection counseling for medical students.