The effects of resident participation on patient morbidity and mortality in major gastrointestinal oncologic surgery.

Abstract

437 Background: While the effects of resident participation have been documented in various studies, there has yet to be a comprehensive study analyzing resident participation in overall gastrointestinal (GI) oncologic surgery. The aim of this study was to compare outcomes in major GI oncologic cases performed by an attending alone and those performed by an attending and resident. Methods: The ACS-NSQIP database from 2005-12 was utilized to study major (GI) operations (esophagectomy, gastrectomy, pancreatectomy, enterectomy, hepatectomy, and colectomy/proctectomy) in patients with an ICD-9 cancer diagnosis. Major complications and 30-day mortality were then compared in those patients who underwent surgery with an attending alone (AA) to those patients who underwent surgery with an attending and resident (AR). Results: A total of 64,637 patients met criteria for the study; AR n = 48,022 and AA n = 16,615. In 76.6% of AR cases, the resident assistance was classified as senior level PGY-4 or higher. On average, operative time was significantly increased in AR cases compared to AA cases (228 ± 130 vs 163 ± 104 min, p < 0.001). On multivariate analysis, AR cases were more likely to develop superficial incisional infection (OR 1.3, CI 95% 1.2-1.4, p < 0.001) and urinary tract infection (OR 1.2, CI 95%, 1.1-1.4, p < 0.001) compared to AA cases. However, on multivariate analysis, resident participation was associated with less likelihood of returning to the operating room (OR 0.9, 95% CI 0.8-0.9, p < 0.001) and lower mortality (OR 0.7, CI 95% 0.6-0.8, p < 0.001). Conclusions: The majority of major GI oncologic operations in the NSQIP database are performed by an attending with the assistance of a senior level resident. This may be due to the complex nature of GI oncologic operation. Potentially, operative time in cases with resident participation may be increased by teaching or by the complex nature of the operation due to the referral bias to teaching centers. However, even with potentially more complex operations, there was less mortality in cases performed by a resident and an attending.