儿童难治性支原体肺炎预后预测模型构建及应用
Construction and application of a prognostic prediction model for refractory mycoplasma pneumonia in children
廖丽娟 1余俊浩1
作者信息
- 1. 深圳市龙华区妇幼保健院,儿科,518000
- 折叠
摘要
目的 研究儿童难治性支原体肺炎的预后预测模型的构建和效果.方法 选取2022年5月到2023年11月于深圳市龙华区妇幼保健院治疗的难治性支原体肺炎患儿100例.采集资料并分析患儿预后的单因素及预后的影响因素,以及BP神经网络模型结果.结果 预后不良组的年龄明显低于预后良好组,中性粒细胞数目、淋巴细胞数目、血小板数目、NLR、PLR、CRP、DD、PCT及LDH水平明显高于预后良好组(P＜0.05).Logistic分析显示,患儿预后影响因素包含NLR、PLR、CRP、DD、PCT及LDH.经过模型优化后,产生的最大AUC值为0.877(95％CI:0.704～0.961),准确度是94.27％,存在良好的一致性.结论 儿童难治性支原体肺炎预后预测模型所产生的应用效果较好,值得在临床上推广应用.
Abstract
Objective To study the construction and application of a prognostic prediction model for refractory mycoplasma pneumonia in children.Methods 100 children with refractory myco-plasma pneumonia treated from May 2022 to November 2023 were selected.Data were collected and analyzed for prognosis factors and influencing factors,as well as the results of BP neural net-work model.Results The age of the poor prognosis group was significantly lower than the good prognosis group,and the number of neutrophils,lymphocyte number,platelet number,NLR,PLR,CRP,DD,PCT and LDH levels were significantly higher than good prognosis group(P＜0.05)o Logistic analysis showed that the prognosis factors included NLR,PLR,CRP,DD,PCT and LDH.After model optimization,the maximum AUC value was 0.877(95％CI:0.704 0.961),and the accuracy was 94.27％,showing good agreement.Conclusion The prognosis prediction model of refractory mycoplasma pneumonia in children is effective and worthy of clini-cal application.
关键词
儿童/难治性支原体肺炎/预后/预测模型/构建/应用Key words
Children/Refractory mycoplasma pneumonia/Prognosis/Predictive model/Con-struction/Application引用本文复制引用
基金项目
深圳市龙华区医疗卫生机构区级科研项目(2021036)
出版年
2024
NETL
NSTL
万方数据