心脏手术患者导管相关性血流感染病原菌及危险因素回归模型构建  

Distribution of pathogenic bacteria in patients with catheter-associated bloodstream infection after cardiac surgery and establishment of regression model for risk factor

作　　者：段海燕 赵俊娅[2] 王伟[3] 兰云霞[4] 张力[1] DUAN Hai-yan;ZHAO Jun-ya;WANG Wei;LAN Yun-xia;ZHANG Li(Henan Provincial Chest Hospital,Zhengzhou University,Zhengzhou,Henan 450o00,China)

机构地区：[1]河南省胸科医院,郑州大学附属胸科医院心血管外科,河南郑州450000 [2]河南省胸科医院,郑州大学附属胸科医院感染防控科,河南郑州450000 [3]河南省胸科医院,郑州大学附属胸科医院检验科,河南郑州450000 [4]河南省胸科医院,郑州大学附属胸科医院护理部,河南郑州450000

出　　处：《中华医院感染学杂志》2023年第24期3748-3752,共5页Chinese Journal of Nosocomiology

基　　金：河南省卫健委联合共建基金资助项目(LHGJ20200206)。

摘　　要：目的分析心脏手术患者导管相关性血流感染的病原菌分布和危险因素并构建Logistic回归模型。方法2018年4月-2022年12月河南省胸科医院收治的心脏手术患者4571例纳入研究,根据患者导管相关性血流感染发生情况分为感染组72例和无感染组4499例,回顾性收集研究对象临床资料。分析心脏手术患者导管相关性血流感染的病原学特点,通过单因素及多因素分析法分析其危险因素并构建Logistic回归模型,绘制受试者工作特征(ROC)曲线分析回归模型的预测价值。结果4571例心脏手术患者并发导管相关性血流感染72例,感染率为1.58%,共分离病原菌96株,其中革兰阳性菌49株,占比51.04%,金黄色葡萄球菌、屎肠球菌、肺炎克雷伯菌是最主要病原菌;多因素Logistic分析结果显示,置管时间≥7 d、异体输血、低蛋白血症、白细胞计数<4×10^(9)/L为心脏手术患者导管相关性血流感染的危险因素(P<0.05);将上述因素纳入Logistic回归模型:Logit(P)=-12.168+0.548×置管时间≥7 d+0.814×异体输血+0.769×低蛋白血症+0.618×白细胞计数<4×10^(9)/L;绘制Logistic回归模型预测心脏手术患者导管相关性血流感染发生的ROC曲线,结果显示,当Logit(P)>13.67时,曲线下面积(AUC)为0.817,诊断敏感度为83.33%、特异度为75.53%。结论心脏手术患者导管相关性血流感染病原菌以革兰阳性菌为主,多种因素与心脏手术患者导管相关性血流感染密切相关,构建Logistic回归模型对心脏手术患者导管相关性血流感染具有良好的预测价值。OBJECTIVE To analyze the distribution characteristics of pathogenic bacteria in patients with catheter-associated bloodstream infection after cardiac surgery, observe the risk factors and construct logistic regression model. METHODS A total of 4 571 patients undergoing cardiac surgery admitted to Henan Provincial Chest Hospital from Apr 2018 to Dec 2022 were included in the study. They were divided into the infected group(72 cases) and the non-infected group(4 499 cases) according to whether the catheter-related bloodstream infection occurred. Clinical data of the subjects were collected retrospectively. The etiological characteristics of catheter-associated bloodstream infection in patients undergoing cardiac surgery were analyzed, the risk factors for catheter-associated bloodstream infection in patients undergoing cardiac surgery were analyzed by univariate and multivariate analysis, and the logistic regression model was constructed. The receiver operating characteristic(ROC) curve was drawn to analyze the prediction value of the regression model. RESULTS Total of 72 cases in 4 571 patients with cardiac surgery were complicated by catheter-associated bloodstream infection, and the catheter-associated bloodstream infection rate was 1.58%. A total of 96 strains of bacteria were isolated, among which 49 strains were gram-positive, accounting for 51.04%. Staphylococcus aureus, Enterococcus faecium, Klebsiella pneumoniae were the main pathogens. Multivariate logistic analysis showed that catheterization time ≥7 d, allogeneic blood transfusion, hypoproteinemia, and white blood cell number <4×10^(9)/L were risk factors for catheter-related bloodstream infection in patients undergoing cardiac surgery(P<0.05). The above factors were included in the logistic regression model as follows: Logit(P) =-12.168+0.548×catheterization time ≥7 days+0.814×allogeneic transfusion+ 0.769×hypoproteinemia+0.618×leukocyte number <4×10^(9)/L. The ROC curve for the occurrence of catheter-associated bloodstream infection in patie

关 键 词：心脏手术 导管相关性血流感染 病原菌 危险因素 模型 预测价值 

分 类 号：R619.3[医药卫生—外科学]