呼吸机相关性肺炎患者MDRO感染的诊断模型构建

Construction of diagnostic model for MDRO infection in patients with ventilator-associated pneumonia

目的通过筛选呼吸机相关性肺炎(VAP)患者多重耐药菌(MDRO)感染的影响因素构建诊断模型,为医院感染管理提供指导信息。方法选取2021年1月—2023年9月194例VAP患者,根据病原菌培养结果分为MDRO组(82例)、非MDRO组(112例),收集二组临床资料,采用Lasso-Logistic回归分析VAP患者MDRO感染的影响因素,构建VAP患者MDRO感染的诊断模型,采用校正曲线、ROC曲线、决策曲线(DCA)进行内部验证,另选取2023年10月—2024年3月80例VAP患者作为外部验证样本,采用ROC曲线进行外部验证。结果二组年龄、机械通气时间、抗菌药物使用时间、急性生理学与慢性健康状况系统(APACHE)评分、深静脉置管、气管切开、意识障碍、留置鼻胃管、抗菌药物联用、血清C-反应蛋白(CRP)、白细胞介素-6(IL-6)水平差异有统计学意义(P<0.05);Lasso-Logistic回归分析显示,年龄、机械通气时间、抗菌药物使用时间、APACHE评分、深静脉置管、抗菌药物联用、留置鼻胃管、血清CRP、IL-6是VAP合并MDRO的独立危险因素(P<0.05);根据Lasso-Logistic回归分析构建VAP患者MDRO感染的诊断模型,内部验证显示,该模型具有较高临床正向净获益和诊断效能,外部验证显示,该诊断模型诊断价值可靠。结论VAP合并MDRO的影响因素为年龄、机械通气时间、抗菌药物使用时间、APACHE评分、深静脉置管、留置鼻胃管、抗菌药物联用、血清CRP、IL-6,各独立因子构建的风险诊断模型具有良好应用价值,为医院感染防控和抗感染治疗提供参考。

Objective To construct a diagnostic model for screening the influencing factors of multiple drug resistant organism(MDRO)infection in patients with ventilator-associated pneumonia(VAP),and provide guidance information for hospital infection management.Methods One hundred and ninety-four patients with VAP from January 2021 to September 2023 were selected and divided into MDRO group(82 cases)and non-MDRO group(112 cases)according to the results of pathogen culture.Clinical data were collected from both groups,and Lasso-Logistic regression analysis was used to investigate the factors affecting MDRO infection in VAP patients.A diagnostic model for MDRO infection in VAP patients was constructed,and internal validation was performed using calibration curves,ROC curves,and decision curves(DCA).In addition,80 patients with VAP from October 2023 to March 2024 were selected as external validation samples,and ROC curves were used for external validation.Results There were statistically significant differences between the two groups in age,duration of mechanical ventilation,duration of antibiotic use,acute physiology and Chronic health status System(APACHE)score,deep vein catheterization,tracheotomy,disturbance of consciousness,indindent nasogastral tube,combination of antibiotics,serum C-reactive protein(CRP)and interleukin-6(IL-6)levels(P<0.05).Lasso-Logistic regression analysis showed that age,mechanical ventilation time,antibacterial drug use time,APACHE score,deep vein catheterization,antibacterial drug combination,indi-ndent nasogastric tube,serum CRP and IL-6 were independent risk factors for VAP and MDRO(P<0.05).Based on Lasso-Logistic regression analysis,a diagnostic model for MDRO infection in VAP patients was constructed.Internal validation showed that the model had high clinical positive net benefit and diagnostic efficacy.External validation showed that the diagnostic model had reliable diagnostic value.Conclusion The risk factors for VAP with MDRO are age,mechanical ventilation time,antibiotic use time,APACHE score,deep venous catheterization,indwelling nasogastric tube,antibiotic combination,serum CRP,and IL-6.The risk diagnosis model constructed from independent factors has good application value and provides a reference for hospital infection prevention and control and anti-infective treatment.