经内镜逆行胰胆管造影术后胆道感染的危险因素分析及预警模型构建

Risk factors analysis of biliary tract infection after ERCP and construction of early warning model

目的探讨经内镜逆行胰胆管造影术(ERCP)术后胆道感染的危险因素并构建预警模型。方法回顾分析2018年7月~2021年6月钦州市中医医院153例行ERCP术治疗的胆总管结石患者临床资料,统计ERCP术后胆道感染发生率,分析ERCP术后发生胆道感染的危险因素。结果153例患者术后并发胆道感染18例,发生率为11.76%。18例胆道感染患者胆汁中共分离出20株病原菌,包括革兰氏阴性菌16株(占80.00%),革兰氏阳性菌4株(占20.00%)。糖尿病(x1)、结石直径(x2)、胆道梗阻位置(x3)和Oddi括约肌功能障碍(x4)是造成ERCP术后胆道感染的重要危险因素(P<0.05),胆汁引流充分(x5)则是保护因素(P<0.05)。最终logistic回归预警模型为Logit(P)=2.216x1+2.219x2+3.343x3+1.659x4-2.361x5-2.668。模型预测术后胆道感染的曲线下面积(AUC)为0.920。结论ERCP术后胆道感染风险较高,革兰氏阴性菌是主要病原菌。糖尿病、胆道梗阻位置、Oddi括约肌功能以及胆汁引流不充分均与ERCP术后胆道感染的发生有关,建立logistic预警模型有助于早期预测胆道感染风险。

Objective To investigate the risk factors of biliary tract infection after ERCP and to construct the early warning model.Methods The clinical data of 153 patients with choledocholithiasis treated by ERCP were retrospectively analyzed.The distribution of pathogenic bacteria in patients with biliary tract infection was analyzed,the risk factors of biliary tract infection after ERCP were analyzed.Results Postoperative biliary tract infection occurred in 18 of 153 patients(11.76%).A total of 20 pathogenic bacteria were isolated from the bile of 18 patients with biliary tract infection,including 16 gram-negative bacteria(80.00%)and 4gram-positive bacteria(20.00%).Diabetes mellitus(x1),stone diameter>1.5cm(x2),high obstruction(x3)and Oddi sphincter dysfunction(x4)were important risk factors for biliary tract infection after ERCP(P<0.05),while adequate biliary drainage(x5)was a protective factor(P<0.05).The final Logistic regression warning model was Logit(P)=2.216x1+2.219x2+3.343x3+1.659x4-2.361x5-2.668.The predictive area under curve(AUC)of biliary tract infection was 0.920.Conclusion The risk of biliary tract infection after ERCP is higher,and gram-negative bacteria are the main pathogenic bacteria.Diabetes,location of biliary tract obstruction,sphincter function of Oddi and biliary drainage are all related to biliary tract infection after ERCP.The establishment of Logistic early warning model is helpful to predict the risk of biliary tract infection in early stage.