Is it appropriate to use core clerkship grades in the selection of residents?

This study challenges the appropriateness of using core clerkship grades for resident selection. The authors hypothesize that substantial variability occurred in the system of grading.

In this retrospective cross-sectional study, variability in the grading systems for third-year core clinical clerkships were examined. From the Medical Student Performance Evaluation of applicants from U.S. medical schools for residency training in the authors' department in 2004 and 2005, the authors gathered the following variables: medical school, third-year core clerkship grading systems, and percentage of students in each grade category. Descriptive analyses were conducted and within institution variability across clerkship scores was analyzed using repeated measure analysis of variance (ANOVA) and t-test.

University teaching hospital.

The survey covered 121 of 122 U.S. medical schools accredited by the AAMC/LCME.

Grading systems used included: variations of Honors/Pass/Fail (H,P,F) system in 76 schools, letter grade systems in 22 schools, and other variants (eg, Outstanding, Advanced, and Proficient in 6 schools and Pass/Fail in 4 schools). Thirteen schools (10%) provided either no grading system or no interpretable system. Grading systems included were further defined into 2 scores in 6 schools, 3 in 34 schools, 4 in 38 schools, 5 in 23 schools, and more than 6 in 6 schools. For schools using a grading system containing 3 or more scores, the percentage of students given the highest grade was significantly less in Surgery (28%) compared with Family Medicine (34%) and Psychiatry (35%) (p = 0.001).

Core clerkship grading systems and the percentage to which institutions grade students as having achieved the highest performance level vary greatly among U.S. medical schools. Within institutions, significant variability exists among clerkships in the percentage of the highest grade given, which makes interpersonal comparison based on core clerkship grades difficult and suggests that this method may not be a reliable indicator of performance.