Risk-adjusted Outcomes of Clinically Relevant Pancreatic Fistula Following Pancreatoduodenectomy: A Model for Performance Evaluation

Objective: To evaluate surgical performance in pancreatoduodenectomy using clinically relevant postoperative pancreatic fistula (CR-POPF) occurrence as a quality indicator.Background: Accurate assessment of surgeon and institutional performance requires (1) standardized definitions for the outcome of interest and (2) a comprehensive risk-adjustment process to control for differences in patient risk.Methods: This multinational, retrospective study of 4301 pancreatoduodenectomies involved 55 surgeons at 15 institutions. Risk for CR-POPF was assessed using the previously validated Fistula Risk Score, and pancreatic fistulas were stratified by International Study Group criteria. CR-POPF variability was evaluated and hierarchical regression analysis assessed individual surgeon and institutional performance.Results: There was considerable variability in both CR-POPF risk and occurrence. Factors increasing the risk for CR-POPF development included increasing Fistula Risk Score (odds ratio 1.49 per point, P < 0.00001) and octreotide (odds ratio 3.30, P < 0.00001). When adjusting for risk, performance outliers were identified at the surgeon and institutional levels. Of the top 10 surgeons (>= 15 cases) for nonrisk-adjusted performance, only 6 remained in this high-performing category following risk adjustment.Conclusions: This analysis of pancreatic fistulas following pancreatoduodenectomy demonstrates considerable variability in both the risk and occurrence of CR-POPF among surgeons and institutions. Disparities in patient risk between providers reinforce the need for comprehensive, risk-adjusted modeling when assessing performance based on procedure-specific complications. Furthermore, beyond inherent patient risk factors, surgical decision-making influences fistula outcomes.