USMLE step 1 scores as a significant predictor of future board passage in pediatrics.

PURPOSE

To explore the relationship(s) between USMLE, In-Training Exam, and American Board of Pediatrics (ABP) board-certifying exam scores within a Pediatric residency-training program.

METHODS

Data were abstracted from records of graduating residents from the Pediatric residency program at the University of Florida College of Medicine Jacksonville from 1999 to 2005. Seventy (70) residents were identified and their files reviewed for the following information: USMLE Step 1 and 2 scores, in-training exam results and eventual board scores as reported by the ABP. Correlation and regression analyses were performed and compared across all tests.

RESULTS

The correlation coefficients between the three types of tests were all statistically significant. Using logistic regression, however, only USMLE Step 1 scores (compared to Step 2) had a statistically significant association with board performance. Interestingly, none of the three in-training exam scores had any additional impact on predicting board performance given one's USMLE Step 1 score. USMLE Step 1 scores greater than 220 were associated with nearly a 95 per cent passage rate on the board-certifying exam.

CONCLUSIONS

The data suggests that performance on USMLE Step 1 is an important predictor of a resident's chances of passing the pediatric boards. This information, which is available when a resident initiates training, can be used to identify those at risk of not passing the boards. While Step 1 scores should not be used as a sole determinant in the recruiting process, individual learning plans can be developed and implemented early in training to maximize one's ability to pass the certifying exam.