广西医科大学学报2025，Vol.42Issue(6)：852-860,9.DOI:10.16190/j.cnki.45-1211/r.2025.06.009
基于meta分析构建成人ECMO患者医院感染风险预测模型及验证
Construction and validation of a meta-analysis-based risk prediction model for nosocomial in-fections in adult patients undergoing extracorporeal membrane oxygenation
摘要
Abstract
Objective:To construct a meta-analysis-based risk prediction model for nosocomial infections in adult patients undergoing extracorporeal membrane oxygenation(ECMO),thereby providing an assessment tool to evaluate and reduce the risk of nosocomial infections in this population.Methods:Databases were systemati-cally searched from inception to June 20,2025,for relevant literature on risk factors associated with nosocomial infections in adult ECMO patients.The quality of included studies was evaluated using the Newcastle-Ottawa Scale(NOS).Meta-analysis was performed with Review Manager 5.4,and integrated risk values of identified fac-tors were used to construct a logistic regression prediction model.Patients who underwent ECMO treatment in a tertiary care hospital from January 2023 to August 2024 were enrolled as the model validation cohort.The model performance was assessed using receiver operating characteristic(ROC)curves,Hosmer-Lemeshow test,calibra-tion curves,and decision curve analysis(DCA).Results:Twenty-six literatures,involving 3,872 patients were in-cluded,and the overall incidence of nosocomial infections in adult ECMO patients was 34.19%.The logistic re-gression model was constructed as follows:Logit(P)=ɑ-0.02×age+0.09×BMI+0.08×duration of ECMO support+0.27×duration of mechanical ventilation+0.02×duration of central venous catheterization+2.06×SOFA score+1.07×CRRT use+1.78×IABP use.The sensitivity and specificity of the model were 80.0%and 68.9%,respec-tively.The area under the ROC curve(AUC)was 0.777(95%CI:0.659-0.894),indicating good discrimination.The Hosmer-Lemeshow test showed satisfactory model calibration(χ²=8.325,P=0.402).Calibration curve analy-sis revealed a prediction error between the predictive model and the actual observations was 0.013,indicating high accuracy and consistency.DCA demonstrated a positive net benefit,suggesting favorable clinical utility.Conclusion:The meta-analysis-based risk prediction model for nosocomial infections in adult ECMO patients demonstrates strong predictive performance.It serves as a useful tool for early identification of patients at high risk for nosocomial infections.关键词
体外膜肺氧合/医院感染/meta分析/危险因素/预测模型Key words
extracorporeal membrane oxygenation/nosocomial infections/meta-analysis/risk factors/predic-tion model分类
医药卫生引用本文复制引用
李燕玲,马惠,黄素,戴霞,黎业钊..基于meta分析构建成人ECMO患者医院感染风险预测模型及验证[J].广西医科大学学报,2025,42(6):852-860,9.基金项目
广西护理学会2023年度科研项目青年项目(No.GXHL202317) （No.GXHL202317）
