Abstract

Objective We sought to measure the impact of specific peri-operative complications after primary cytoreductive surgery on relevant patient outcomes and use of resources. Methods A cohort of patients with advanced ovarian cancer who underwent primary cytoreductive surgery at two institutions (2006–2016) were studied. Specific known complications (‘exposures’) within 30 days of surgery were evaluated to determine the impact on outcomes. Exposures included bowel leak, superficial surgical site infection, deep surgical site infection, venous thromboembolic event, and cardiac event. Outcomes were prolonged lengths of stay, readmission or non-home discharge, reoperation, organ failure, delay to adjuvant chemotherapy, and 90-day mortality. Population attributable risk (PAR) was used to estimate the proportion of adverse outcomes that could be prevented by elimination of a causal exposure and considers both the strength of the association and the prevalence of the complication; adjusted PARs (aPAR) were calculated using adjusted relative risks (aRR) adjusted for stage (IIIC vs IV) and American Society of Anesthesiology score (<3 vs ≥3). Results A cohort of 892 patients was included. Each of the evaluated exposures had an impact on readmission/non-home discharge (aPAR range 5.3 to 13.5). A venous thromboembolic event was significantly associated with 90-day mortality (aRR=2.9 (95% CI 1.3 to 6.7); aPAR=8.6 (95% CI −1.8 to 19.1)) and organ failure (aRR=4.7 (95% CI 2.3 to 9.5); aPAR=13.9 (95% CI 2.8 to 25.1)). Similarly, a cardiac event was most strongly associated with organ failure and was very impactful (aPAR=19.0 (95% CI 6.8 to 31.1)). Bowel leak was a major contributor to poor outcome, including reoperation (aPAR=45.5 (95% CI 34.3 to 56.6)), organ failure (aPAR=13.6 (95% CI 2.6 to 24.6)), readmission/non-home discharge (aPAR=5.3 (95% CI 1.6 to 9.0)), delay to adjuvant chemotherapy (aPAR=5.9 (95% CI 2.3 to 9.4)), and prolonged lengths of stay (aPAR=13.0 (95% CI 9.1 to 16.9)). Conclusion Going beyond reporting complications using common scales to measure their genuine impact provides important information for providers, patients, and payers. We report that less frequent exposures, including a venous thromboembolic event, cardiac events, and bowel leaks, have a high impact on patients and use of resources.