摘要
目的探讨支气管哮喘患儿并发肺部感染的影响因素,并构建列线图模型进行分析与验证。方法回顾性分析2020年1月-2024年1月在江苏省人民医院宿迁医院儿科接受治疗的500例支气管哮喘患儿的临床资料,按7∶3的比例分为训练集(n=350)和验证集(n=150)。根据肺部感染情况将训练集患儿分为合并感染组(n=25)和非感染组(n=325),收集患儿基本信息(年龄、性别),临床症状(咳嗽、喘息、发热),实验室检查[血红蛋白(Hemoglobin,Hb)、白蛋白(Albumin,Alb)、C反应蛋白(C-reactive protein,CRP)、降钙素原(Procalcitonin,PCT)、白细胞计数(White blood cells,WBC)、血小板计数(Platelet count,PLT)、中性粒细胞与淋巴细胞比值(Neutrophil-to-lymphocyte ratio,NLR)],影像学检查(是否合并胸腔积液)及抗哮喘药物使用频率。采用SPSS 26.0及R42.2软件进行数据分析,通过多因素Logistic回归分析筛选支气管哮喘并发肺部感染的危险因素,构建预测模型。采用受试者工作特征(Receiver operating characteristic,ROC)曲线、校准曲线及决策曲线评估模型的区分能力、校准度及临床实用性等内部验证。结果训练集与验证集临床资料比较,差异无统计学意义(P>0.05)。合并感染组患儿发热、CRP、PCT、NLR、合并胸腔积液比例以及抗哮喘药物使用频率高于非感染组(P<0.05)。多因素Logistic回归分析显示,CRP、PCT、NLR、合并胸腔积液以及抗哮喘药物使用频率为小儿支气管哮喘并发肺部感染的影响因素(P<0.05)。基于训练集构建小儿支气管哮喘并发肺部感染的列线图风险预测模型,内部验证的ROC曲线下面积(Area under curve,AUC)分别为0.880、0.920,校准曲线均趋于实际曲线(Hosmer-Lemeshowχ^(2)=0.615、0.598,P均>0.05)。决策曲线分析结果显示,列线图模型预测小儿支气管哮喘并发肺部感染风险的阈值>0.15时,列线图模型提供临床净收益。结论基于小儿支气管哮喘并发肺部感染的独立影响因素构建的列线图模型具有良好预测效能和临床应用价值。
Objective To investigate the risk factors for pulmonary infection in children with bronchial asthma and to develop and validate a nomogram prediction model.Methods The clinical data of 500 children with bronchial asthma treated in the hospital from January 2020 to January 2024 were retrospectively analyzed,and it randomly divided into a training set(n=350)and a validation set(n=150)in a 7∶3 ratio.The children in the training set were divided into the co-infected group(n=25)and the non-infected group(n=325)according to the pulmonary infection situation.Data collected included basic information(age,sex),clinical symptoms(cough,wheezing,fever),laboratory tests[hemoglobin(Hb),albumin(Alb),C-reactive protein(CRP),procalcitonin(PCT),white blood cell count(WBC),platelet count(PLT)and neutrophil-to-lymphocyte ratio(NLR)],imaging findings(presence of pleural effusion),and frequency of anti-asthmatic medication use.Data analysis was performed using SPSS 26.0 and R42.2 software.Multivariate Logistic regression analysis was used to identify risk factors for pulmonary infection in children with bronchial asthma,and a prediction model was constructed.The model's discriminative ability,calibration and clinical utility was evaluated using receiver operating characteristic(ROC)curves,calibration curves and decision curve analysis(DCA)for internal validation.Results There was no significant difference in clinical data between training set and verification set(P>0.05).The proportion of fever,CRP,PCT,NLR,pleural effusion and the frequency of use of anti-asthma drugs in the co-infected group were higher than those in the non-infected group(P<0.05).Multivariate Logistic regression analysis identified CRP,PCT,NLR,pleural effusion and frequency of anti-asthmatic drug use as the independent risk factors for pulmonary infection in children with bronchial asthma(P<0.05).Based on training set,a nomogram risk prediction model for pediatric bronchial asthma complicated with pulmonary infection was constructed.The area under ROC curve(AUC)verified internally were 0.880 and 0.920,respectively,and the calibration curve was close to the actual curve(Hosmer-Lemeshowχ^(2)=0.615,0.598,all P>0.05).The decision curve showed that when the threshold for predicting the risk of pulmonary infection in children with bronchial asthma was greater than 0.15,the nomogram model provided a clinical net benefit.Conclusion The nomogram model based on independent risk factors for pulmonary infection in children with bronchial asthma demonstrates a good predictive performance and clinical utility.
作者
沈勤
李秋侠
李欢欢
王千
张军
孙炳霞
SHEN Qin;LI Qiuxia;LI Huanhuan;WANG Qian;ZHANG Jun;SUN Bingxia(Department of Pediatrics,Jiangsu Province(Suqian)Hospital,Suqian Jiangsu 223800,China)
出处
《新疆医科大学学报》
2025年第3期322-327,共6页
Journal of Xinjiang Medical University
基金
江苏省妇幼健康科研项目(F201643)
宿迁市科技计划资助市第一人民医院科研专项项目(SY202304)。
关键词
小儿支气管哮喘
肺部感染
列线图
预测模型
pediatric bronchial asthma
pulmonary infection
nomogram
prediction model
