摘要
目的观察分析新生儿缺氧缺血性脑病(HIE)不良预后的预测模型,为协助临床简便、有效、快速识别HIE患儿不良预后的高危群体奠定基础。方法选取2020年5月至2022年4月四川省妇幼保健院新生儿科收治的128例HIE患儿作为建模组人群,2022年5月至2023年2月收治的103例HIE患儿作为验证组人群。收集患儿一般及临床资料,Logistic回归建模,限制性立方样条(RCS)优化,进一步绘制列线图,采用Bootstrap法、决策曲线(DCA)、受试者操作特征(ROC)曲线进行评价。结果128例建模组HIE患儿中,预后不良31例,预后良好97例,分别占24.219%、75.781%。5 min Apgar评分、病情程度、合并持续性肺动脉高压(PPHN)、昏迷、惊厥、脑电异常、早期干预时间窗<6 h、亚低温治疗是HIE患儿预后的影响因素(P<0.05);模型1、模型2的标准度良好,绝对误差依次为0.011、0.017,表明模型校准度良好;模型1、模型2列线图对HIE预后预测的AUC依次为0.870、0.922;DCA提示,任何阈值概率下,采用列线图(模型1、2)预测HIE患儿预后的净收益率均较高,且模型2净收益率大于模型1;验证组HIE患儿预后良好75例,预后不良28例,在验证组人群内,模型2列线图预测HIE患儿预后的AUC为0.914,模型2列线图在验证组、建模组人群中的AUC对比,无显著差异(P>0.05)。结论影响HIE患儿预后不良的风险因素众多,且列线图模型可对其进行有效预测,相关部门可加强预测模型构建和应用,给予高危人群有效处理对策,降低高危因素危害,从而减少不良预后事件。
Objective To observe and analyze the prediction model of poor prognosis of neonates with hypoxic-ischemic encephalopathy(HIE),and to lay a foundation for facilitating clinical simple,effective and rapid identification of high-risk groups with poor prognosis of HIE.Methods A total of 128 children with HIE admitted to the Department of neonatology of Sichuan Provincial Maternity and Child Health Care Hospital from May 2020 to April 2022 were selected as the modeling group,and 103 children with HIE admitted from May 2022 to February 2023 were selected as the validation group.General and clinical data of children were collected.Logistic regression modeling,restricted cubic spline(RCS)optimization,and nomogram were further drawn.Bootstrap method,decision curve(DCA)and receiver operating characteristic(ROC)curve were used for evaluation.Results Among 128 HIE children in the modeling group,31 cases had poor prognosis and 97 cases had good prognosis,accounting for 24.219%,75.781%,respectively.5 min Apgar score,disease degree,combined persistent pulmonary hypertension(PPHN),coma,convulsion,EEG abnormality,early intervention time window<6 h,mild hypothermia treatment,NBNA score were the prognostic factors of HIE children(P<0.05).The standard degree of model 1 and model 2 is good,and the absolute errors are 0.011 and 0.017,indicating that the calibration degree of the model is good.The AUC of model 1 and model 2 histogram for HIE prognosis was 0.870 and 0.922,respectively.DCA indicated that under any threshold probability,the net return rate of HIE patients predicted by nomogram(models 1 and 2)was higher,and the net return rate of model 2 was higher than that of model 1.In the verification group,75 cases had good prognosis with HIE,and 28 cases had poor prognosis.In the verification group,the AUC predicted by model 2 histogram was 0.914,and there was no significant difference in AUC between the verification group and the modeling group(P>0.05).Conclusion There are many risk factors affecting the poor prognosis of children with HIE,and the nomogram model can effectively predict them.Relevant departments can strengthen the construction and application of the prediction model,give effective treatment countermeasures to high-risk groups,reduce the harm of high-risk factors,and thus reduce adverse prognostic events.
作者
胡椿艳
张勇
石镜懿
HU Chunyan;ZHANG Yong;SHI Jingyi(Department of Neonatology,Sichuan Provincial Maternity and Child Health Care Hospital,Chengdu,Sichuan 610041,China)
出处
《中国优生与遗传杂志》
2024年第6期1162-1169,共8页
Chinese Journal of Birth Health & Heredity
基金
2021年四川省医学会(青年创新)科研课题项目:S21355
关键词
缺氧缺血性脑病
预后
影响因素
新生儿
预测模型
hypoxic-ischemic encephalopathy
prognosis
influencing factors
newborn
prediction model