Vandergrift J, Gray B, Weng W. — American Board of Internal Medicine
Presented: AcademyHealth Annual Research Meeting, June 2015
Research Objective: There exists wide variation among U.S. states in continuing medical education (CME) requirements, ranging from none to 50 credit-hours per year, for physicians reregistering their medical license. The motivation for these standards is largely to ensure physicians keep abreast of current medical knowledge. However, no research has looked at the relationship between CME requirements and a physician’s medical knowledge despite the considerable differences in effort required among physicians. This study addressed this gap by evaluating how changes in these policies across both states and years affect medical knowledge as measured by performance on the ABIM’s 10-year maintenance of certification (MOC) examination.
Study Design: Linear regression was used to evaluate general internists’ MOC exam performance before versus after state CME changes utilizing states with stable CME regulations to control for secular changes in performance. Models included state, county and year fixed effects as well as trend-state interactions. The model controlled for a physician’s initial certification (i.e., baseline) exam score, as well as their sex and whether they were a U.S. allopathic, U.S. osteopathic or international medical graduate. State level cluster adjustments were applied to account for correlated errors.
To address the research question, between 2006 and 2013, seven states increased the credit-hours per year (n=5, mostly instituting CME requirements) and/or decreased the term (n=3, years to complete) for CME (one state made both changes in different years). Overall, 5% of physicians practiced in states that increased credit-hours per year and 14% practiced in states that decreased their term. The study’s policy variables were indicators for a state: (1) instituting CME or increasing credit-hours per year mandates or (2) decreasing their CME term (interval for completing CME) holding credit-hours per year constant. These indicators, combined with the state and year fixed effects and baseline exam performance, provide a measure of the change in exam performance associated with increased CME requirement stringency.
Population Studied: Board certified general internists (n=21,314), initially certified 1996-2003, that took the ABIM Internal Medicine MOC exam 2006-2013 (76% of required physicians).
Principal Findings: An increase in credit-hours per year, or instituting CME, was associated with a 12.4 point increase in MOC exam score (p<.001). This corresponds with a shift from the 50th to 55th percentile of exam performance (standard difference = 0.13). A decrease in CME term was not associated with exam performance (4.2 point increase, p=0.17). Sensitivities found results were not driven by any one state. Further, baseline exam performance was not different in states with, and without, credit-hours per year changes (p=0.28).
Conclusions: Supporting the changes in CME requirements observed, there was an increase in the credit-hours per year or institution of CME requirements was associated with a statistically significant increase in a general internist's medical knowledge. No association between CME term and performance was found.
Implications for Policy or Practice: There is evidence supporting the use of general state CME mandates in that they have a measurable effect on a physician’s medical knowledge. Additional research is required to identify the optimal level and frequency with which CME should be required.
For more information about this presentation, please contact Research@abim.org.