External validation of fistula risk scores for postoperative pancreatic fistula after distal pancreatectomy

Abstract

BackgroundFistula risk scores such as distal fistula risk scores and DISPAIR have been recently developed to assess the risk of clinically relevant postoperative pancreatic fistula after distal pancreatectomy. This study aimed to validate these models externally using a large-scale Chinese cohort. MethodsThe study enrolled adult patients who underwent distal pancreatectomy at a high-volume single center between January 2011 and December 2021. The clinically relevant postoperative pancreatic fistula was defined as grade B/C, according to the 2016 International Study Group of Pancreatic Surgery. Model performance was evaluated using the area under the curve. ResultsAmong 653 eligible patients, 126 (19.3%) suffered from clinically relevant postoperative pancreatic fistulas. Independent predictors for clinically relevant postoperative pancreatic fistulas included body mass index, diabetes mellitus, pancreatic thickness at both neck and transection sites, main pancreatic duct diameter, and soft pancreas. Clinically relevant postoperative pancreatic fistula risk increased with increasing score severity. All 3 prediction models showed acceptable discrimination, with area under the curve values of preoperative distal fistula risk score at 0.723 (95% confidence interval 0.687–0.757), intraoperative distal fistula risk score at 0.737 (95% confidence interval 0.701–0.770), and DISPAIR at 0.721 (95% confidence interval 0.685–0.755). No significant differences were found among them. ConclusionDistal fistula risk scores and DISPAIR are useful tools for predicting clinically relevant postoperative pancreatic fistula after distal pancreatectomy, highlighting their importance in guiding surgical approach decisions and mitigating strategies against this complication in clinical practice.