气管切开长期机械通气老年患者肺部感染的nomogram模型构建OA
Construction of a nomogram model for pulmonary infection in elderly patients undergoing long-term mechanical ventilation via tracheotomy
目的:探讨气管切开长期机械通气老年患者肺部感染的影响因素,并构建nomogram模型.方法:选取2023年4月至2024年11月在南昌大学第二附属医院行气管切开长期机械通气的老年患者146例为训练集,依据有无发生肺部感染分为感染组(n=41)、未感染组(n=105),采用多因素Logistic回归分析对影响肺部感染发生的因素进行分析,构建nomogram模型.另选取2024年11月至2025年7月行气管切开长期机械通气的老年患者63例为验证集.通过受试者工作特征(receiver operating characteristic,ROC)曲线、校准曲线验证预测性能,用验证集数据进行验证.结果:多因素Logistic回归分析结果显示,急性生理和慢性健康状况Ⅱ(acute physiology and chronic health evaluationⅡ,APACHEⅡ)评分、营养不良、抗生素使用时间、超敏C反应蛋白水平、白蛋白水平、合并高脂血症、合并糖尿病均是气管切开长期机械通气老年患者肺部感染的独立影响因素(P＜0.05).训练集曲线下面积为0.905,最佳截断值0.346,灵敏度80.5%,特异度88.6%,区分度较佳;平均绝对误差为0.027,校准性能较优,可靠性较强.验证集模型的外部预测性能较好.结论:APACHEⅡ评分、营养不良、抗生素使用时间、超敏C反应蛋白水平、白蛋白水平、合并高脂血症、合并糖尿病为气管切开长期机械通气老年患者肺部感染的独立危险因素,构建的nomogram模型预测性能较好.
Objective:To investigate the influencing factors of pulmonary infection in elderly patients undergoing long-term mechanical ventilation via tracheotomy,and to construct a nomogram model.Methods:A total of 146 elderly patients who underwent long-term mechanical ventilation via tracheostomy at the Second Affiliated Hospital of Nanchang University from Apr 2023 to Nov 2024 were selected as the training set.These patients were divided into the infection group(n=41)and the non-infection group(n=105)based on the presence or absence of pulmonary infection.Multivariate logistic regression analysis was used to analyze the factors influencing the occurrence of pulmonary infection,and a nomogram model was constructed.Another 63 elderly patients who underwent long-term mechanical ventilation via tracheotomy from Nov 2024 to Jul 2025 were selected as the validation set.The predictive performance was verified through receiver operating characteristic(ROC)curves and calibration curves,and validation was conducted using the validation set data.Results:Multivariate logistic regression analysis showed that acute physiology and chronic health evaluation Ⅱ(APACHE Ⅱ)score,malnutrition,duration of antibiotic use,hypersensitive C-reactive protein(hs-CRP)level,albumin(ALB)level,combined hyperlipidemia,and combined diabetes were independent influencing factors of pulmonary infection in elderly patients undergoing long-term mechanical ventilation via tracheotomy(P＜0.05).The area under the ROC curve(AUC)of the training set was 0.905,with an optimal cutoff value of 0.346,sensitivity of 80.5%,and specificity of 88.6%,indicating good discrimination.The mean absolute error(MAE)was 0.027,demonstrating good calibration performance and strong reliability.The external predictive performance of the model in the validation set was also good.Conclusions:The APACHE Ⅱ score,malnutrition,duration of antibiotic use,hs-CRP level,ALB level,combined hyperlipidemia,and combined diabetes are independent risk factors for pulmonary infection in elderly patients undergoing long-term mechanical ventilation via tracheotomy.The constructed nomogram model exhibits good predictive performance.
杨芬芳;王海燕;余丽娜;习苏园;舒苹;刘颜颜
南昌大学第二附属医院重症医学科(综合ICU),江西 南昌 330000南昌大学第二附属医院重症医学科(综合ICU),江西 南昌 330000南昌大学第二附属医院重症医学科(综合ICU),江西 南昌 330000南昌大学第二附属医院重症医学科(综合ICU),江西 南昌 330000南昌大学第二附属医院重症医学科(综合ICU),江西 南昌 330000南昌大学第二附属医院重症医学科(综合ICU),江西 南昌 330000
医药卫生
老年患者气管切开长期机械通气肺部感染nomogram模型
elderly patientstracheotomylong-term mechanical ventilationpulmonary infectionnomogram model
《沈阳医学院学报》 2026 (3)
257-263,281,8
江西省卫生健康委科技计划(No.20203317)
评论