Predictive Value Of FRS System For Pancreatic Fistula After Pancreaticoduodenectomy

Objective:To investigate the predictive value of pancreatic fistula risk scoring system(FRS)for pancreatic fistula and clinically relevant pancreatic fistula after pancreaticoduodenectomy(PD).Methods:Retrospective analysis of clinical data of patients with PD in the First Affiliated Hospital of Xinjiang Medical University from June 2016 to June 2018,collecting pancreatic texture,main pancreatic duct diameter,intraoperative blood loss and pathological diagnosis according to pancreas The Fistula Risk Score(FRS)system accumulates points and divides patients into four risk groups.The sensitivity and specificity of the FRS model and the area under the AUC curve were analyzed using the ROC curve.Results:Of 104 patients with PD,the incidence of postoperative pancreatic fistula was 48.0%(50/104),including 30 cases of biochemical sputum,18 cases of B grade,and 2 cases of c grade,the incidence rate was 28.8%,17.3%,19.2% respectively.There was no CR-POPF in patients with negligible risk group.The proportion of CR-POPF in patients with moderate risk group and high risk group was significantly higher than that of negligible and low risk group.ROC curve analysis showed: the sensitivity of the FRS system for predicting POPF was 71.1%,the specificity was 59.6%,the AUC was 0.655(95% CI: 0.547-0.764),and the sensitivity of FRS ?3 for predicting CR-POPF was 51.4%,and the specificity of that is 80.3% and the AUC is 0.654(95% CI: 0.485 to 0.822).Conclusion:FRS system can more accurately predict the occurrence of pancreatic fistula after PD,meanwhile,FRS?3 points have a good predictive value for CR-POPF.