National trends in centralization and perioperative outcomes of complex operations for cancer

Complex cancer operations performed at high-volume and teaching hospitals have been associated with better outcomes. The purpose of this study was to determine the national trends in the performance of these operations at large teaching hospitals. Patients who underwent elective esophagectomies, gastrectomies, pancreatectomies, and hepatectomies for cancer (2003–2015) were identified using the National Inpatient Sample. We determined average annual percent change (AAPC) in the proportion of operations at large teaching hospitals, inpatient complications, length of stay (LOS), and inpatient mortality. Between 2003 and 2015, 38,932 esophageal, 104,941 gastric, 96,098 hepatic, and 137,440 pancreatic cancer resections were performed. The proportion at large teaching hospitals increased with an AAPC of 2.5 for esophagectomies (P < .001), 3.6 for gastrectomies (P < .001), and 1.5 for pancreatectomies (P = .039), but did not change for hepatectomies (AAPC 0.48, P = .50). During the study period, mean LOS and inpatient mortality rates at large teaching hospitals decreased across hospital types. By 2013 to 2015, the operations at large hospitals were associated with decreased mortality only for pancreatectomies (odds ratio, 0.62, 95% confidence interval, 0.43–0.91, P = .015). Complex cancer operations are performed increasingly at large teaching hospitals, but perioperative outcomes have improved nationally across hospital types. Further studies should identify actionable areas for improvement to ensure accessible quality cancer care.