儿童呼吸道感染所致高热惊厥风险预警模型的构建与验证

Construction and verification of early warning model for febrile convulsion risk caused by respiratory tract infection in children

林燕霞 ¹林建锋²

扫码查看

作者信息

1. 莆田市儿童医院急诊科,福建莆田 351100
2. 莆田市秀屿区埭头镇中心卫生院,福建莆田 351100
折叠

摘要

目的构建并验证儿童呼吸道感染所致高热惊厥风险预警模型.方法回顾性选取2020年4月至2022年8月在我院治疗的156例呼吸道感染患儿作为研究对象,以7:3的比例分为模型样本(109例)和验证样本(47例),两样本均根据是否发生高热惊厥分为高热惊厥组和无高热惊厥组.比较两组患儿性别、年龄、既往是否发生高热惊厥、家族癫痫史、合并呼吸节律紊乱、血压、体温、呼吸频率(BR)、心率(HR)及血清学指标的情况.使用多因素Logistic回归分析高热惊厥的影响因素并构建患儿高热惊厥的风险预警模型,以ROC曲线判断模型准确度,并通过Hosmer-Lemeshow检验评价该模型的拟合度.结果模型样本中高热惊厥组43例(39.45％),未高热惊厥组66例(60.55％);验证样本中高热惊厥组16例(34.04％),未高热惊厥组31例(65.96％).两组患儿的年龄、既往是否发生高热惊厥、合并呼吸节律紊乱、体温、BR、HR及血钙方面比较,差异均有统计学意义(均P＜0.05).多因素Logistic回归分析结果显示,两组患儿的年龄(OR=0.044,95％CI:0.004～0.447,P=0.008)、合并呼吸节律紊乱(OR=10.678,95％CI:1.506～75.709,P=0.018)、体温(OR=11.963,95％CI:1.725～82.951,P=0.012)、BR(OR=1.571,95％CI:1.138～2.168,P=0.006)及 HR(OR=1.252,95％CI:1.076～1.457,P=0.004)为患儿高热惊厥的独立影响因素.在此基础上建立风险预警模型,预警模型的ROC曲线下面积为0.981,95％CI:0.960～1.000,敏感度为 0.930,特异度为 0.970,Youden 指数为 0.900.Hosemer-Lemeshow 检验显示 x2=4.136,P=0.845.结论基于列线图模型构建的呼吸道感染患儿高热惊厥预警模型的预警效能良好,医务人员可通过患儿年龄、合并呼吸节律紊乱、体温、BR及HR等指标判断呼吸道感染患儿高热惊厥的风险,并采取针对性干预方案.

Abstract

Objective To construct and verify the risk early warning model of febrile convulsion caused by respiratory tract infec-tion in children.Methods A total of 156 children with respiratory tract infection who were treated in the hospital from April 2020 to August 2022 were retrospectively selected as the research objects.They were divided into model samples(109 cases)and validation samples(47 cases)at a ratio of 7:3.Both samples were divided into febrile convulsion group and non-febrile convulsion group accord-ing to whether febrile convulsion occurred.Multivariate Logistic regression was used to analyze the influencing factors of febrile convul-sion and to construct a risk warning model for febrile convulsion in children.The accuracy of the model was judged by ROC curve,and the fitting degree of the model was evaluated by Hosmer-Lemeshow test.Results There were 43 cases(39.45％)in the febrile con-vulsion group and 66 cases(60.55％)in the non-febrile convulsion group.There were 16 cases(34.04％)in the febrile convulsion group and 31 cases(65.96％)in the non-febrile convulsion group.The results of multivariate analysis showed.Age(OR=0.044,95％CI:0.004-0.447,P=0.008),respiratory rhythm disorder(OR=10.678,95％CI:1.506-75.709,P=0.018),body temperature(OR=11.963,95％CI=1.725-82.951,P=0.012),respiratory rate(BR)(OR=1.571,95％CI:1.138-2.168,P=0.006)and heart rate(HR)(OR=1.252,95％CI:1.076-1.457,P=0.004)were independent influencing factors of febrile convulsion in the two groups(all P＜0.05).The area under the ROC curve was 0.981,95％CI:0.960-1.000,the sensitivity was 0.930,the specificity was 0.970,and the Youden index was 0.900.The Hosemer-Lemeshow test showed x2=4.136,P=0.845.Conclusion The early warn-ing model of febrile convulsion in children with respiratory tract infection based on the nomogram model has good early warning perform-ance.Medical staff can judge the risk of febrile convulsion in children with respiratory tract infection by age,combined respiratory rhythm disorder,body temperature,BR and HR,and take targeted intervention programs.

关键词

儿童/呼吸道感染/高热惊厥/预警模型

Key words

Children/Respiratory tract infection/High fever convulsion/Early warning model

引用本文复制引用

出版年

2024

蛇志

中国蛇协

蛇志

影响因子：0.439

ISSN：1001-5639

参考文献量4

段落导航