An Institutional Review: Which Metrics Correlate With a Successful United States Medical Licensing Examination Step 1 Score?

J Oral Maxillofac Surg. 2020 Feb;78(2):179-183. doi: 10.1016/j.joms.2019.09.015. Epub 2019 Sep 21.

Abstract

Purpose: Since the introduction of the Comprehensive Basic Science Examination (CBSE) as an entrance examination in 2012, no studies have been performed correlating its relationship with passing rates on United States Medical Licensing Examination (USMLE) Step 1 in dental students. This study was designed to determine the impact of the CBSE score, dental grade point average (GPA), and undergraduate GPA on USMLE Step 1 performance and develop a CBSE cutoff score that correlates with a passing USMLE score.

Materials and methods: A single-blinded, retrospective, cross-sectional study was designed. Data were collected from University of Alabama at Birmingham oral-maxillofacial surgery residents who had matriculated from 2014 to 2018. The primary predictor variable was the CBSE score. The primary outcome variable was the USMLE Step 1 score. Additional predictor variables included undergraduate and dental school GPAs. Bivariate statistics were calculated using a 2-tailed Pearson correlation (P = .05). Confounders were investigated using multivariate linear regression (P = .05). A bivariate linear regression was created using the variables of CBSE and USMLE scores.

Results: Dental school GPA and CBSE score correlated with USMLE Step 1 score (P < .05). Bivariate linear regression between CBSE and USMLE scores yielded a predictive equation of USMLE score = 2.02 × CBSE score + 66.2 (R2 = 0.30). When second-attempt scores were included, this equation became USMLE score = 2.08 × CBSE score + 67 (R2 = 0.49). The positive predictive value for a CBSE cutoff score of 61 reached 100% on repeated USMLE attempts.

Conclusions: A CBSE score for dental students of 61 or greater correlates with a passing USMLE score and should be implemented to screen for dual-degree oral-maxillofacial surgery candidates.

MeSH terms

  • Benchmarking*
  • Cross-Sectional Studies
  • Educational Measurement*
  • Humans
  • Retrospective Studies
  • United States