Previous Abstract | Next Abstract
Printable Version
Does USMLE Step I Examination Predict Performance on In-Training Examination in Anesthesiology Residents?
Amr E. Abouleish, M.D., M.B.A.; S. Lynn Knox, M.D.; Nhung H. Nguyen, M.D.; J. Sean Funston, M.D.; Ann W. Frye, Ph.D.
Department of Anesthesiology, The University of Texas Medical Branch, Galveston, Texas
Introduction: In reviewing applications for residency, the only standardized test available to compare applicants from different medical schools is the United States Medical Licensing Examination (USMLE) Step 1 score. Generally, USMLE Step 2 is not available at the time the resident applicant is evaluated. Unfortunately, it is not clear whether the USMLE Step 1 score is predictive of future anesthesiology specialty examination (in-training examination and written board) scores for the applicant. Therefore, we undertook this study to determine whether in-training scores on written examinations can be predicted by the USMLE Step 1 score.

Methods: The USMLE Step 1 scores were collected for all residents entering anesthesiology residency at one medical school from 1994 to 2001. Anesthesiology in-training examinations and written board examinations are given in July of each year. The categorization is based on the years of training completed at the time of the examination. In-training scores for the first 4 years of training (PGY-0, CB, CA-1, and CA-2) and written board examination (CA-3) scores were collected, if available, for the same residents. Pearson's correlation was performed on USMLE Step 1 scores and each level of the in-training examination and written board examination. Because levels of residency completion differed among residents (i.e., a resident beginning in 2001 would have only have taken the PGY-0 examination, a resident beginning in 2000 would have only taken the PGY-0, and CB examinations, and a resident beginning in 1999 would have only taken the PGY-0, CB, and CA-1 examinations), there are more PGY-0 scores than CA-3 scores.

Results: From 1994-2001, there were 91 residents for whom USMLE Step 1 scores were available. In-training scores were available for 60 PGY-0, 74 CB, 59 CA-1, 49 CA-2, and 33 CA-3 residents. The USMLE Step 1 score correlated least with PGY-0 scores (r= 0.30) and most with CA-3 scores (r = 0.44; see Figure). If USMLE step 1 scores less than 180 or greater than 220 (greater than 1 standard deviation from national mean of 200) are excluded, then scores between 180 and 220 did not correlate at all with CA-3 scores (r = 0.14).

Discussion: USMLE Step 1 scores did not correlate with any of the in-training written examinations for anesthesiology. Because of the low numbers for those residents who scored high (>220) or low (<180) on USMLE Step 1, it is not clear whether these very high or low scores correlate with written board (CA-3) examination scores.


Am J Surgery 182:143, 2001

Teach Learn Med 13:9, 2001

Acad Med 68:753,1993

Anesthesiology 2002; 96: A1115
Figure 1