Can students' scores on preclerkship clinical performance examinations predict that they will fail a senior clinical performance examination?

Purpose: This study was designed to determine whether preclerkship performance examinations could accurately identify medical students at risk for failing a senior clinical performance examination (CPE).

Method: This study used a retrospective case-control, multiyear design, with contingency table analyses, to examine the performance of 412 students in the classes of 2005 to 2010 at a midwestern medical school. During their second year, these students took four CPEs that each used three standardized patient (SP) cases, for a total of 12 cases. The authors correlated each student's average year 2 case score with the student's average case score on a senior (year 4) CPE. Contingency table analysis was carried out using performance on the year 2 CPEs and passing/failing the senior CPE. Similar analyses using each student's United States Medical Licensing Examination (USMLE) Step 1 scores were also performed. Sensitivity, specificity, odds ratio, and relative risk were calculated for two year 2 performance standards.

Results: Students' low performances relative to their class on the year 2 CPEs were a strong predictor that they would fail the senior CPE. Their USMLE Step 1 scores also correlated with their performance on the senior CPE, although the predictive values for these scores were considerably weaker.

Conclusions: Under the conditions of this study, preclerkship (year 2) CPEs strongly predicted medical students at risk for failing a senior CPE. This finding opens the opportunity for remediation of deficits prior to or during clerkships.