RETENTION OF BASIC SCIENCE INFORMATION BY SENIOR MEDICAL STUDENTS

David Swanson*¹, Yu Ling¹, Kathy Holtzman¹ and S. Deniz Bucak¹

¹National Board of Medical Examiners (NBME), Philadelphia, PA 19104 U.S.A.

PURPOSE: Over the past 30 years, NBME studies of the retention of basic science information have consistently shown performance declines as students progress through medical education. This study investigated whether the changes occurring in basic science education and assessment over the past decade have altered patterns of basic science retention.

METHODS: 502 content and statistically representative basic science items from 2004-05 forms of USMLE Step 1 were included in unscored sections of 2004-05 USMLE Step 2 Clinical Knowledge (CK) test forms, and the performance of 15,000+ first-time examinees from US and Canadian schools was analyzed to identify item characteristics affecting retention.

RESULTS: Across all 502 items, the mean item difficulty on Step 1 was 76.1%; on Step 2 CK, this value declined 6.4% to 69.7%. Performance improvement was only observed for Behavioral Sciences items (8.7%); performance declined in all other disciplines, with the largest declines in Biochemistry (17.5%), Microbiology (12.6%) and Pharmacology (10.1%). Performance on items presented in the context of patient vignettes declined relatively little (3.1%), with larger declines observed for experimental vignettes (12.0%) and non-vignettes (14.2%).

CONCLUSION: Shifts in examinee performance were similar to those observed in four previous NBME studies, although the magnitude of the overall decline was somewhat larger. These results were somewhat disappointing: one might anticipate that use of integrated basic science curricula would improve retention of basic science information, particularly when assessed in a clinical context. Clearly, additional research on teaching, learning and testing students’ understanding of basic science is desirable.