摘要
目的:构建个体化预测成都市儿童反复呼吸道感染发生风险的列线图模型,并评价该列线图模型的准确性以及临床适用性。方法:选取近期我院儿科收治的呼吸道感染患儿作为研究对象,根据患儿是否发生反复呼吸道感染将患儿分为反复呼吸道感染患儿和普通呼吸道感染患儿;从整体研究对象中随机将选取70%患儿作为训练集,30%作为测试集,在训练集中利用Logistic回归分析成都市儿童反复呼吸道感染发生风险的独立危险因素,基于上述危险因素构建列线图模型,在训练集中利用Bootstrap方法进行1000次抽样检测并绘制校准曲线,利用ROC曲线和临床决策曲线对列线图模型进行内部验证,并在测试集中验证模型的临床效能。结果:在训练集中Logistic分析显示成都市儿童反复呼吸道感染发生风险的独立危险因素为:过敏史、哮喘病史、维生素A以及维生素D偏低、低出生体重、被动吸烟、低血红蛋白;训练集中Bootstrap验证显示C-index为0.879(95%CI:0.812~0.932),校准曲线显示列线图模型预测儿童反复呼吸道感染发生风险效果与实际观察曲线接近,临床决策曲线显示列线图模型预测预测成都市儿童反复呼吸道感染发生风险发生阈值为0.04~0.95之间时该模型图的适用性最佳,ROC曲线显示,AUC面积为0.921(95%CI:0.851~0.968),模型的灵敏度为0.848,特异度为0.872;在验证集中ROC曲线显示,AUC面积为0.912(95%CI:0.871~0.941),模型的灵敏度为0.845,特异度为0.836。结论:构建的个体化预测成都市儿童反复呼吸道感染发生风险列线图模型具有良好的准确性和临床适用性。
Objective To construct an individualized nomogram model for predicting the risk of recurrent respiratory tract infections in urban children in Chengdu,and to evaluate the accuracy and clinical applicability of this nomogram model.Methods Children with respiratory tract infections admitted to our hospital recently were selected as the study subjects.The children were divided into two groups:those with recurrent respiratory tract infections and those with ordinary respiratory tract infections.70%of the subjects were randomly selected as the training set,and the remaining 30%were used as the testing set.Logistic regression analysis was conducted on the training set to identify independent risk factors for recurrent respiratory tract infections in Chengdu children.Based on the identified risk factors,a nomogram model was constructed.The calibration curve was drawn after performing 1000 rounds of bootstrap sampling on the training set.Internal validation of the nomogram model was performed using ROC curves and clinical decision curves,and the clinical efficacy of the model was validated on the testing set.Results Logistic regression analysis on the training set revealed the following independent risk factors for recurrent respiratory tract infections in urban children in Chengdu:history of allergies,history of asthma,low levels of vitamin A and vitamin D,low birth weight,passive smoking,and low hemoglobin levels.The C-index of the nomogram model in the training set was 0.879(95%CI 0.812-0.932).The calibration curve showed that the predicted risk of recurrent respiratory tract infections in children using the nomogram model was close to the actual observed curve.The clinical decision curve indicated that the nomogram model was most applicable when the predicted risk of recurrent respiratory tract infections in urban children in Chengdu was between 0.04 and 0.95.The ROC curve showed an AUC of 0.921(95%CI 0.851-0.968),sensitivity of 0.848,and specificity of 0.872 for the model.In the testing set,the ROC curve showed an AUC of 0.912(95%CI 0.871-0.941),sensitivity of 0.845,and specificity of 0.836.Conclusion The constructed individualized nomogram model for predicting the risk of recurrent respiratory tract infections in urban children in Chengdu demonstrates good accuracy and clinical applicability.
作者
李娟
蔡和花
王志强
陈桂华
吴晓艳
刘茜洋
LI Juan;CAI Hehua;WANG Zhiqiang;CHEN Guihua;WU Xiaoyan;LIU Qianyang(Department of Pediatrics,Chengdu Third People's Hospital,Chengdu 610014,China;Department of Immunology,Institute of Basic Research,China Medical University,Shenyang 110012,China)
出处
《湖南师范大学学报（医学版）》
2024年第1期52-57,共6页
Journal of Hunan Normal University(Medical Sciences)
基金
辽宁省教育厅科学研究项目“基于儿童自然人群队列的能量代谢效率分型标志物研究及与肠道菌群的相关性分析”(QN2019040)
