胰十二指肠切除胰胃吻合术后胰瘘术前风险因素分析

A preoperative clinical risk factor analysis for postoperative pancreatic fistula of pancreatogastrostomy after pancreatoduodenectomy

目的 分析胰十二指肠切除胰胃吻合术后胰瘘的术前风险因素.方法 回顾性分析2009年6月至2013年12月在我科接受胰十二指肠切除胰胃吻合患者临床病理资料及胰瘘发生情况,应用Logistic回归分析胰十二指肠切除胰胃吻合术后胰瘘的危险因素,并建立术前预测胰瘘的评分系统.应用ROC曲线分析评分系统预测患者术后胰瘘发生的敏感度和特异度.结果 共有139例患者纳入本研究,39例患者发生A级胰瘘（28.5％）,15例为B级胰瘘（10.8％）,6例为C级胰瘘（4.3％）.Logistic回归多因素分析提示术前减黄（wald χ^2=5.515,P =0.019）、超重（wald χ^2=6.233,P=0.013）、胰管直径＜3 mm（wald χ^2=10.392,P=0.001）与术后胰瘘发生密切相关.根据回归系数建立术后胰瘘风险评分系统,评分越高患者发生胰瘘风险越高,0分患者胰瘘发生率为9.9％,而4分患者胰瘘发生率为92.9％,评分系统的ROC曲线下面积为0.771 （95％ CI：0.694～0.849）.结论 术前减黄、超重、胰管直径＜3 mm与术后胰瘘发生密切相关,根据回归分析建立的术前风险评分系统能够比较有效的预测胰十二指肠切除术胰胃吻合术后胰瘘的发生.

Objective To explore the preoperative risk factors for postoperative pancreatic fistula （PF） of pancreatogastrostomy after pancreatoduodenectomy （PD）.Methods We retrospectively summarize the patients＆#39; clinical data and postoperative complications who underwent PD with pancreatogastrostomy form June 2009 to Dec 2013.Risk factors associated with PF were assessed and preoperative clinical risk score system was constructed by Logistic regression analysis.The validity of the scoring system was assessed by ROC curve.Results 139 patients were recruited.Grade A PF was recognized in 39 patients （28.5%）,grade B in 15 （10.8%）,grade C in 6（4.3%）.In a multivariate analysis,only overweight（wald χ^2 =6.233,P =0.013）,preoperative biliary drainage （wald χ^2 =5.515,P =0.019）,pancreatic duct size 〈 3 mm （wald χ^2 =10.392,P =0.001） were significant predictors of PF.A score based on the number of risk factors present divided patients into 5 subgroups carrying a risk of PF ranging from 9.9% （score =0） to 92.9% （score =4）.The ROC curve showed that AUC of this score was 0.771（95% CI：0.694-0.849）.Conclusions Overweight,preoperative biliary drainage,pancreatic duct size 〈 3 mm are significant predictors of PF after pancreatogastrostomy.The preoperative scoring system effectively predicts the subsequent PF.