热性惊厥患儿复发风险列线图预测模型构建的多中心研究  

Multicenter study on the construction of recurrence risk nomogram prediction model in children with febrile seizure

作　　者：李梅[1] 徐江燕[1] 黄树稳 黄丽梅 周春鸣 李世巧 邓海兰 梁修业 LI Mei;XU Jiangyan;HUANG Shuwen;HUANG Limei;ZHOU Chunming;LI Shiqiao;DENG Hailan;LIANG Xiuye(Pediatrics Department,Nanning Maternity and Child Health Hospital,Nanning 530011;Pediatrics Department,Long'an Maternal and Child Health Hospital,Nanning 532700;Pediatrics Department,Binyang Maternal and Child Health Hospital,Nanning 530400;Pediatrics Department,Wuming Maternal andChild Health Hospital of Nanning,Nanning 530100,China)

机构地区：[1]南宁市妇幼保健院儿科,广西南宁530011 [2]隆安县妇幼保健院儿科,广西南宁532700 [3]宾阳县妇幼保健院儿科,广西南宁530400 [4]南宁市武鸣区妇幼保健院儿科,广西南宁530100

出　　处：《临床医学研究与实践》2025年第7期33-36,共4页Clinical Research and Practice

基　　金：广西壮族自治区卫生健康委员会自筹经费科研课题(No.Z-A20221252)。

摘　　要：目的分析热性惊厥(FS)患儿复发的因素,并构建预测FS患儿复发情况的列线图模型。方法选择2021年1月至2022年6月在本课题多中心协同医院就医的FS患儿为研究对象,根据1年内FS复发情况将患儿分为复发组与未复发组,并按7:3的比例随机分为训练集与验证集。通过Logistic回归分析明确影响FS患儿复发的因素,并构建列线图模型,采用受试者工作特征(ROC)曲线与决策曲线检验模型的区分度与临床获益率。结果310例FS患儿中,1年内复发的有97例(31.29%)。多因素Logistic回归分析结果显示,首次发病年龄≤1岁、惊厥前体温≤39℃、惊厥持续时间、一次热程中惊厥次数≥2次及IL-6水平是FS患儿1年内复发的危险因素(P<0.05);同时,FS患儿1年内的复发风险随惊厥持续时间的延长与IL-6水平的升高而增加。模型在训练集与验证集的曲线下面积(AUC)分别为0.876、0.877。决策曲线分析显示,在31.29%阈值下,模型净获益率分别为60%和85%。结论基于多因素分析构建列线图预测模型对临床上识别FS复发高风险患儿与采取针对性措施进行干预具有一定指导价值。Objective To analyze the factors of recurrence in children with febrile seizure(FS),and to construct a nomogram model for predicting the recurrence of children with FS.Methods The children with FS who were treated in the multicenter collaborative hospital of this subject from January 2021 to June 2022 were selected as the research objects.According to the recurrence of FS within 1 year,the children were divided into recurrence group and non-recurrence group,and randomly divided into training set and verification set according to the ratio of 7∶3.Logistic regression analysis was used to identify the factors affecting recurrence in children with FS,a nomogram model was constructed,and receiver operating characteristic(ROC)curve and decision curve were used to test the discrimination and clinical benefit rate of the model.Results Of the 310 children with FS,97(31.29%)recurred within 1 year.Multivariate Logistic regression analysis results showed that the age of first onset≤1 year old,body temperature before convulsion≤39℃,duration of convulsion,number of convulsions in one fever≥2 times and IL-6 level were risk factors for recurrence within 1 year in children with FS(P<0.05);at the same time,the recurrence risk within 1 year of FS children increased with the prolongation of convulsion duration and the increase of IL-6 level.The area under curve(AUC)of the model in the training set and the validation set was 0.876 and 0.877,respectively.The decision curve analysis showed that the net benefit rate of the model was 60% and 85% respectively at the threshold of 31.29%.Conclusion The nomogram prediction model based on multivariate analysis has certain guiding value for clinical identification of children with high risk of FS recurrence and targeted intervention measures.

关 键 词：热性惊厥 复发 列线图模型 

分 类 号：R742.1[医药卫生—神经病学与精神病学]