基于随机森林模型的耐碳青霉烯类鲍曼不动杆菌性呼吸机相关性肺炎风险预测模型的构建  

Construction of a risk prediction model for ventilator⁃associated pneumonia caused by carbapenem resistant Acinetobacter baumannii based on random forest model

作　　者：冯清 贺培凤[2] FENG Qing;HE Peifeng(First Hospital of Shanxi Medical University,Shanxi 030001 China)

机构地区：[1]山西医科大学第一医院,山西030001 [2]山西医科大学医学数据科学研究院

出　　处：《护理研究》2024年第19期3410-3416,共7页Chinese Nursing Research

摘　　要：目的:分析耐碳青霉烯类鲍曼不动杆菌(CRAB)性呼吸机相关性肺炎发生的风险因素,并采用随机森林模型和Logistic回归两种方法构建预测模型,为重症监护室降低CRAB性呼吸机相关性肺炎的发生风险提供理论依据。方法:选取2018年1月—2022年12月我院重症监护室收治的291例呼吸机相关性肺炎病人为研究对象,分析CRAB性呼吸机相关性肺炎的影响因素,基于随机森林模型和Logistic回归构建预测模型,计算受试者工作特征曲线(ROC)和曲线下面积(AUC),比较两种模型的差异。结果:多因素分析结果显示,氧合指数、气管切开、昏迷是CRAB性呼吸机相关性肺炎的独立影响因素。随机森林模型的AUC为0.78,Logistic回归模型AUC为0.61,随机森林模型的准确率(77.97%)、灵敏度(85.37%)、特异度(61.11%)均高于Logistic回归模型(66.10%、73.17%、50.00%)。结论:氧合指数、抗菌药物使用时间、气管切开、昏迷是CRAB性呼吸机相关性肺炎的高危风险因素,随机森林模型对CRAB性呼吸机相关性肺炎的预测性能优于Logistic回归模型。Objective:To analyze the risk factors for ventilator-associated pneumonia caused by carbapenem-resistant Acinetobacter baumannii(CRAB).Random forest model and Logistic regression two methods were used to construct prediction models,to provide theoretical basis for ICU to reduce the risk of CRAB ventilators associated pneumonia.Methods:A total of 291 patients with ventilators associated pneumonia admitted to ICU from January 2018 to December 2022 were selected as the study subjects.The influencing factors of CRAB ventilators associated pneumonia were analyzed.The predictive model was constructed based on random forest model and Logistic regression,and the operating characteristic curve(ROC)and area under the curve(AUC)were calculated to compare the differences between the two models.Results:Multivariate analysis results showed that oxygenation index,tracheotomy and coma were independent influencing factors of CRAB ventilator-associated pneumonia.The AUC of the random forest model was 0.78,and the AUC of the Logistic regression model was 0.61.The accuracy(77.97%),sensitivity(85.37%),and specificity(61.11%)of the random forest model were higher than those of the Logistic regression model(66.10%,73.17%,50.00%).Conclusion:Oxygenation index,duration of antibiotic use,tracheotomy and coma are risk factors for CRAB ventilator-associated pneumonia.The random forest prediction model outperforms the Logistic regression model in predicting CRAB ventilator-associated pneumonia.

关 键 词：耐碳青霉烯类鲍曼不动杆菌 呼吸机相关性肺炎 影响因素 随机森林 LOGISTIC回归 预测模型 

分 类 号：R563.1[医药卫生—呼吸系统]