M. M. Symer1, L. Gade3, J. Abelson1, J. A. Sosa2, H. Yeo1 1Weill Cornell Medical College,Surgery,New York, NY, USA 2Duke University Medical Center,Surgery,Durham, NC, USA 3NewYork-Presbyterian / Queens,Surgery,New York, NY, USA
Introduction: The American Board of Surgery In-Training Exam (ABSITE) has been demonstrated to predict passage of the ABS certifying exam, and is intended to guide education rather than penalize residents. Attrition from general surgery training is common and costly but poorly understood. We hypothesized that ABSITE scores would not predict attrition, but changes in score may be correlated with a resident who is struggling and at risk for impending attrition.
Methods: In 2007, all categorical general surgery interns were administered a survey during their first months of residency. De-identified survey results containing resident demographics were linked to a database of ABSITE scores assembled separately by the American Board of Surgery. Attrition was determined based on completion of training during eight years of follow-up. Residents without ABSITE scores, a matching survey, and/or missing scores were excluded. Resident ABSITE scores were analyzed based on average rank, a normalized percentile derived from their raw score. Year-to-year change in ABSITE score was used to compare residents with a significant change in performance as a possible predictor of impending attrition.
Results:Of 837 residents, 739 (88.3%) completed surveys and had continuous ABSITE data until completion or attrition from training. 108 (14.6%) did not complete training. Residents who dropped out were more likely to be female (18.7% vs 12.3% male, p = 0.02) and from programs with ≥6 residents (19.2% vs 13.0% <6 residents, p=0.04). Average ABSITE rank (median normalized percentile) was higher for participants who completed training (51.8 vs 42.7 percentile dropouts, p<0.001). Scores were also higher for residents without family nearby (53.0 vs 48.5 percentile family nearby, p<0.01). There was no difference in attrition between residents with a single ABSITE rank below the 25th percentile (12.2% vs 17.1% without a low ranking, p=0.06). Those residents who experienced an ABSITE score drop of >16.5 percentile points from the previous year were more likely to leave training (13.0% vs 6.0% without such a drop, p=0.003). In adjusted analysis, a one percentile increase in ABSITE rank was associated with decreased odds of attrition (OR 0.98 95%CI 0.97-0.99, p<0.01).
Conclusion:Lower ABSITE scores are associated with attrition, but there is only a small absolute difference in scores between those residents who complete training and those who drop out. Program directors should focus their efforts on residents with an acute drop in scores, which may signify that a resident is at risk of impending drop out.