HBV相关ACLF患者细菌感染的危险因素及预测模型建立  

Risk factors for bacterial infection in patients with HBV-associated ACLF and establishment of prediction model

作　　者：陈丹丹[1] 王文虎[1] 黄顺东[1] 尹明红[1] 李忠斌 CHEN Dan-dan;WANG Wen-hu;HUANG Shun-dong;YIN Ming-hong;LI Zhong-bin(The Second People's Hospital of Jingzhou City,Jingzhou,Hubei 434000,China;不详)

机构地区：[1]湖北省荆州市第二人民医院肝病二科,湖北荆州434000 [2]解放军总医院第五医学中心肝病医学部肝病科,北京100039

出　　处：《中华医院感染学杂志》2023年第8期1202-1206,共5页Chinese Journal of Nosocomiology

基　　金：湖北省自然科学基金资助项目(201969)。

摘　　要：目的分析乙型肝炎病毒(HBV)相关慢加急性肝衰竭(ACLF)患者住院期间细菌感染的危险因素,并建立、验证预测模型。方法选择荆州市第二医院肝病科2018年4月-2021年4月因HBV相关ACLF住院接受治疗患者120例为研究对象,将发生细菌感染患者列入感染组,未发生任何感染患者纳入对照组,另从120例患者中随机选取60例患者作为交叉验证组。通过床旁问卷或患者管理系统查询并记录临床资料。采用Kaplan-Meier生存曲线观察预测模型不同风险分层下全部研究对象和交叉验证组医院感染发生率差异。采用受试者工作特征(ROC)曲线分析预测模型对全部对象和交叉验证组对象发生医院感染的预测价值。结果感染组合并糖尿病、住院时间、合并腹水、降钙素原(PCT)、C-反应蛋白(CRP)高于对照组,白蛋白低于对照组(P<0.05);分析结果发现合并糖尿病、住院时间较长、合并腹水、PCT、CRP水平较高者和白蛋白水平较低发生细菌感染的风险也较大(P<0.05);所建立预测模型Hosmer-Lemeshow检验结果χ^(2)=7.645,P=0.225>0.05,表明模型一致性良好。根据各变量在模型中得分中位数将其分为高风险组和低风险组,Kaplan-Meier生存曲线分析结果显示预测模型分组下细菌感染发生率差异有统计学意义(P<0.05);ROC分析结果显示,预测模型预测研究对象总体、交叉验证组曲线下面积分别为0.842(95%CI 0.765~0.918)、0.839(95%CI 0.766~0.912)提示预测模型预测效能较好。结论以合并糖尿病、住院时间较长、合并腹水、PCT、CRP和白蛋白建立的预测模型对于HBV相关ACLF患者住院期间发生细菌感染有较好的预测价值。OBJECTIVE To analyze the risk factors for bacterial infection in patients with hepatitis B virus(HBV)-associated acute-on-chronic liver failure(ACLF)during hospitalization,and to establish a prediction model and validate it.METHODS A total of 120 patients who were treated in the Department of Hepatology at the Second Hospital of Jingzhou for HBV-associated ACLF from Apr 2018 to Apr 2021 were enrolled as the research subjects.Patients with bacterial infections were included in the infection group,and patients without any infection were included in the control group.Another 60 patients were randomly selected as the cross-validation group.The patients′clinical data were collected by bedside questionnaire or patient management system.The Kaplan-Meier survival curve was used to observe the difference in the nosocomial infection rates among all subjects with different risk grades and the cross-validation group.The receiver operating characteristic(ROC)curve was used to analyze the predictive value of the prediction model for nosocomial infections in all subjects and the cross-validation group.RESULTS The proportion of patients with diabetes,the length of hospital stay,the proportion of patients with ascites,procalcitonin(PCT)and C-reactive protein(CRP)in the infection group were higher than those in the control group,while albumin was lower than that in the control group(P<0.05).Analysis showed that the risk of bacterial infection in the patients with diabetes,longer hospital stay,ascites,higher PCT and CRP levels and lower albumin level were higher(P<0.05).Hosmer-Lemeshow test of the prediction model wasχ^(2)=7.645,P=0.2250.05,indicating that the mode had good consistency.According to the median score of each variable in the model,they were divided into the high-risk group and low-risk group.Kaplan-Meier survival curve analysis showed that there was a significant difference in the incidence of bacterial infection between the groups(P<0.05).ROC curve analysis showed that the area under the curve values of the predic

关 键 词：乙型肝炎病毒 慢加急性肝衰竭 细菌感染 预测模型 生存曲线 危险因素 

分 类 号：R512.62[医药卫生—内科学]