Defining Benchmarks for Major Liver Surgery A multicenter Analysis of 5202 Living Liver Donors

Objective: To measure and define the best achievable outcome after major hepatectomy. Background: No reference values are available on outcomes after major hepatectomies. Analysis in living liver donors, with safety as the highest priority, offers the opportunity to define outcome benchmarks as the best possible results. Methods: Outcome analyses of 5202 hemi-hepatectomies from living donors (LDs) from 12 high-volume centers worldwide were performed for a 10-year period. Endpoints, calculated at discharge, 3 and 6 months postoperatively, included postoperative morbidity measured by the Clavien-Dindo classification, the Comprehensive Complication Index (CCI), and liver failure according to different definitions. Benchmark values were defined as the 75th percentile of median morbidity values to represent the best achievable results at 3 month postoperatively. Results: Patients were young (34 +/- [9] years), predominantly male (65%) and healthy. Surgery lasted 7 +/- [2] hours; 2% needed blood transfusions. Mean hospital stay was 11.7 +/- [5] days. 12% of patients developed at least 1 complication, of which 3.8% were major events (>= grade III, including 1 death), mostly related to biliary/bleeding events, and were twice higher after right hepatectomy. The incidence of postoperative liver failure was low. Within 3-month follow-up, benchmark values for overall complication were <= 31 %, for minor/major complications <= 23% and <= 9%, respectively, and a CCI <= 33 in LDs with complications. Centers having performed >= 100 hepatectomies had significantly lower rates for overall (10.2% vs 35.9%, P < 0.001) and major (3% vs 12.1%, P < 0.001) complications and overall CCI (2.1 vs 8.5, P < 0.001). Conclusions: The thorough outcome analysis of healthy LDs may serve as a reference for evaluating surgical performance in patients undergoing major liver resection across centers and different patient populations. Further benchmark studies are needed to develop risk-adjusted comparisons of surgical outcomes.