摘要
目的探讨耐碳青霉烯类肠杆菌科细菌(CRE)医院感染风险预测模型构建及其直接经济负担评价.方法以常州市金坛第一人民医院2018年1月-2022年12月住院患者中432例肠杆菌科细菌医院感染患者为研究对象,分为碳青霉烯类敏感肠杆菌科细菌(CSE)组372例和CRE组60例,建立Logistic回归风险预测模型,以Hosmer-Lemeshow检验、受试者工作特征(ROC)曲线及曲线下面积(AUC)检测该模型的校准度和区分度,利用灵敏度、特异度来判断模型判别能力;采用倾向得分匹配法(PSM)进行1:1病例对照研究,计算CRE医院感染的直接经济负担.结果CSE组和CRE组均以感染大肠埃希菌和肺炎克雷伯菌为主;趋势性卡方检验显示2018-2022年CRE医院感染患者占比逐年增高;合并其他多重耐药菌感染、合并其他敏感菌感染、15<住院天数≤30 d为CRE医院感染的危险因素(P<0.05);Logistic回归风险预测模型预测能力较好,ROC曲线AUC为0.789、灵敏度为73.33%、特异度为75.27%;除手术费和材料费外,CRE组患者的诊疗费、药费、床位费、护理费、化验费、检查费、治疗费、其他费均高于CSE组(P<0.05),CRE医院感染的直接经济负担为26219.81元.结论CRE医院感染会增加患者直接经济负担,本研究建立的风险预测模型判别能力较好,可为临床预防和控制CRE医院感染提供可靠的理论依据.
OBJECTIVE To explore the establishment of risk prediction model for carbapenem-resistant Enterobacteriaceae(CRE)nosocomial infection and evaluate the direct economic burden.METHODS A total of 432 patients with Enterobacteriaceae nosocomial infection who were hospitalized in Changzhou Jintan Fist People's Hospital from Jan 2018 to Dec 2022 were recruited as the research subjects and were divided into the carbapenem-sensitive Enterobacteriaceae(CSE)group with 372 cases and the CRE group with 60 cases.Logistic regression risk prediction model was established.The calibration and discrimination of the model were detected by Hosmer-Lemeshow test,receiver operating characteristic(ROC)curve and area under curve(AUC);the discriminative ability of the model was assessed by sensitivity and specificity.A 1:1 case-control study was conducted by propensity score matching(PSM)method,and the direct economic burden due to CRE nosocomial infection was calculated.RESULTS Escherichia coli and Klebsiella pneumoniae infections were dominant in both the CSE group and the CRE group.The tendency chi-square test showed that the proportion of the patients with CRE nosocomial infection was increased year by year in 2018-2022.Complication with other multidrug-resistant bacteria infection,complication with other drug-sensitive bacteria infections and 15<length of hospital stay<30 days were the risk factors for the CRE nosocomial infection(P<o.05).Logistic regression risk prediction model had favorable prediction performance,the AUC of ROC curve was 0.789,the sensitivity was 73.33%,the specificity was 75.27%.Except for the surgical and material costs,the diagnosis and treatment fees,medication fees,bed fees,nursing fees,laboratory test fees,examination fees,treatment fees and other expenses of the CRE group were significantly higher than those of the CSE group(P<0.05);the direct economic burden to CRE nosocomial infection was 26219.81 yuan.CONCLUSION The CRE nosocomial infection can increase the direct economic burden of the patients.The risk prediction model has favorable discrimination ability,and it can provide reliable theoretical basis for clinical prevention and control of CRE nosocomial infection.
作者
王玉沐
尹伟琴
杨乐
WANG Yu-mu;YIN Wei-qin;YANG Le(Changzhou Jintan First People's Hospital,Changzhou,Jiangsu 213000,China;不详)
出处
《中华医院感染学杂志》
CAS
CSCD
北大核心
2023年第21期3219-3224,共6页
Chinese Journal of Nosocomiology
基金
江苏省医院急诊管理专项基金资助项目(JSYGY-3-2021-JZ33)。
关键词
耐碳青霉烯类肠杆科细菌
医院感染
病原学
危险因素
风险预测模型
倾向得分匹配法
经济负担评价
Carbapenem-resistant Enterobacteriaceae
Nosocomial infection
Etiology
Risk factor
Risk prediction model
Propensity score matching
Economic burden evaluation
