Abstract
Objective: The American College of Surgeons (ACS) NSQIP® Risk Calculator is designed to estimate postoperative risk. Using ACS-predicted complication rates, we identified significant discrepancies in outcomes for patients undergoing major hepatectomy (≥3 Couinaud segments) at our high-volume HPB center. The goal of this study was to develop and validate an institution-specific risk calculator for patients undergoing major hepatectomy at our institution. Methods: Outcomes generated by the ACS calculator were recorded for 136 major hepatectomies performed at our institution (2008–2015). In parallel, novel predictive models for seven postoperative outcomes were constructed and probabilities calculated. Brier score and area under the curve (AUC) were employed to assess predictive accuracy. Internal validation was performed using bootstrap logistic regression. Logistic regression models were constructed using bivariate (p<0.25) and multivariate (p<0.05) analyses. Model accuracy was validated using retrospective and prospective data. Results: Brier scores showed no significant difference in predictive ability of the ACS and institution-specific models. However, significant differences in the discriminative ability of the ACS and institution-specific models were identified at the individual level. The ACS model weakly predicted individual postoperative risk for 6/7 outcomes (AUC: 0.5200.619, p>0.05) compared to the institution-specific model (AUC: 0.7550.969, p<0.05). Predictive capacities were similar for 30-day mortality (ACS AUC: 0.858, p<0.05; institution-specific AUC: 0.815, p<0.05). Conclusion: Institution-specific models provide superior outcome predictions for individual perioperative risk following major hepatectomy and, if properly developed/validated, can be used to generate more accurate, patient-specific delivery of care.Table 1Comparison of predictive capacity of ACS-NSQIP® and Institution-specific risk-prediction models for postoperative complications of major hepatectomy performed at a high-volume center.ACS modelInstitution-specific modelAUCAUC p-valueBrier scoreAUCAUC p-valueBrier scoreSurgical site infection0.6190.2340.0300.9690.0030.015Renal failure0.5230.4110.0290.958<0.0010.021Urinary tract infection0.5970.2180.0280.9120.0020.035Discharge to nursing/rehabilitation facility0.6010.0920.0900.8920.0000.060Serious complication0.5680.1010.1890.7640.0000.15130-day mortality0.858<0.0010.0670.8150.0010.04730-day readmission0.5200.3450.1560.7550.0000.128Abbreviations: ACS-NSQIP, American College of Surgeons National Quality Improvement Program; AUC, Area under the curve. Open table in a new tab Abbreviations: ACS-NSQIP, American College of Surgeons National Quality Improvement Program; AUC, Area under the curve.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have