Relationship between pathology curricular approaches and performance in the United States medical licensing examination (USMLE), step 1: a national cross-sectional study

Abstract

A collaborative multi-institutional study was conducted to study the relationship between characteristics of pathology instruction and student performance in the United States Medical Licensing Examination (USMLE), Step 1. Detailed descriptions of pathology instruction, such as curriculum format, style of instruction, and design of instruction of various areas of pathology for the 1996-1997 academic year were matched with Step 1 total scores and pathology subscores for 10,159 students from 88 schools who took the June 1997 examination. Hierarchical linear modeling was used for analysis while controlling for students' MCAT-bpv scores (Medical College Admission Test scores, average of biological and physical sciences and verbal reasoning). The Step 1 total scores correlate with Step 1 pathology subscores, and both correlate with students' MCAT-bpv scores. The MCAT-bpv scores account for most of the variation in Step 1 scores. Of all of the curricular variables analyzed, 2 possible factors that significantly influence the impact of MCAT-bpv scores on Step 1 scores are (1) multidisciplinary format of instruction and (2) discipline-based general pathology instruction in year 2. Overall, the multidisciplinary format reduced the impact of MCAT-bpv scores on both the Step 1 total score and the pathology subscore by 1 point per MCAT-bpv. Overall, the general pathology instruction in year-2 increased that impact on Step 1 total score by 1 point per MCAT-bpv. The slope became less steep with multidisciplinary format and more steep with year-2 general pathology instruction. As a result, students with higher MCAT-bpv scores tend to benefit from year-2 general pathology instruction, whereas those with lower MCAT-bpv scores appear to benefit from a multidisciplinary format. These differences become more apparent as the MCAT-bpv scores move away from the mean in either direction, indicating that scores of most students are not affected by curricular variables. Overall, there is no significant difference in the school means of Step 1 total scores and pathology subscores of schools with different curricular approaches. This is most likely due to the finding that the students' MCAT-bpv scores, and not curricular variables, are the major predictors of Step 1 scores, and all schools have a mix of students with various MCAT-bpv scores.