Dugosh JW, Ostrem WE. — American Board of Internal Medicine
Song H. — National Board of Osteopathic Medical Examiners
Presented: Association of American Medical Colleges Conference, November 2013
Background: The face of graduate medical education in the U.S. has changed dramatically over the past two decades. International Medical Graduates (IMGs) now account for over one fourth (27%) of physicians in graduate medical training in the U.S. This change in the composition of graduate medical trainees has resulted in a need to reexamine the role that international medical education may play in influencing the outcomes of ACGME-accredited programs in graduate medical education.
The American Board of Internal Medicine (ABIM) is the primary body for certifying physicians in geriatric medicine, after successful completion of a one-year ACGME-approved fellowship. The initial certification exam in geriatric medicine is unique in that the vast majority (67%) of physicians who sit for the exam did their undergraduate medical training in a medical school outside of the U.S. Therefore, an investigation of possible differences in performance on the initial certification exam in geriatric medicine between U.S. medical graduates (USMGs) and IMGs deserves attention.
Methodology: Differential item functioning (DIF) analyses were performed on the 2009-2011 certification exams in geriatric medicine to identify item-level differences in performance between IMGs and USMGs who have comparable overall exam scores. Instead of its common usage for defending fairness of exam programs, DIF was used here as a diagnostic tool. The Mantel-Haenszel (MH) procedure was used for detecting DIF.
Results: Across the three exam administrations, 8%-13% of items (16, 26 and 18, respectively, of 200 items) were flagged for intermediate to large DIF. About half of the items favored USMGs while the other half favored IMGs, suggesting that the overall exam was fair for both groups. The content of nine items on which USMGs outscored IMGs was examined and categorized into one of four distinct categories: Cultural/Demographic, Ethics, Psychiatry and Medicare. All four of these topics are required to be taught in ACGME-accredited fellowship training programs in geriatric medicine.
Conclusions: IMGs may be differentially motivated to learn certain concepts–especially if they plan to return to their countries of origin. IMG fellows may simply not attend to information that they perceive as less relevant to a practice outside of the U.S. (e.g., Medicare rules; U.S. cultural/demographic information; ethics based on U.S. mores). DIF on psychiatry items may be explained by a relative lack of emphasis on psychiatric treatment in primary care outside of the U.S. The geriatric medicine certification exam is a fair assessment of the prescribed curricula of training programs. IMGs, however, may be selectively inattentive to certain information that they perceive to be of low relevance. Training programs should proactively emphasize these concepts–particular for their IMG trainees.
For more information about this presentation, please contact Research@abim.org.