Comparison between different predicting systems for pancreatic fistula after pancreaticoduodenectomy:which is better?

Abstract

Introduction: Though the safety of PD has been significantly improved, the post-operative pancreatic fistula (POPF) always is a formidable complication. It is important to choose different peri-operative strategy to manage POPF according the risk predicting scores. There were three scoring systems to predict the risk of clinically relevant postoperative pancreatic fistula (CR-POPF), National Cancer Center Hospital (NCCH) system ; University Hospitals Birmingham (UHB) system and Fistula risk score (FRS). Though these systems were established and validated by external data, there was less comparison between each other. Method: The study was a retrospective analysis of 202 patients who underwent PD from January 2015 to October 2017 in the general surgery department of Peking University Third Hospital by the same staffs. All data were collected, including preoperative medical status, operative procedure, final pathology, gland texture, pancreatic duct diameter and predicting system scores. We defined POPF according to International Study Group of Pancreatic Fistula (ISGPF) definition. Result: The incidence of POPF was 43.1% and the incidence of CR-POPF was 29.2% in 202 patients. The ROC area under the curve (AUC) for UHB(0.802) was higher than that for FRS (0.785) and NCCH (0.739) (p<0.01). Conclusion: University Hospitals Birmingham (UHB) scoring system is better than National Cancer Center Hospital (NCCH) system and Fistula risk score (FRS) in predicting CR-POPF.