个体化预测胎儿和新生儿溶血的风险Nomgram模型建立

Establishment of Nomgram risk model for personalized prediction of fetal and neonatal hemolysis

目的 研究个体化预测胎儿和新生儿溶血(HFN)的风险Nomgram模型。方法 选择分娩的胎儿和新生儿16 450例,随机选择15 182例作为建模组,1 268例为验证组。根据溶血情况将建模组胎儿和新生儿分为HFN患儿3 882例(HFN组),其他为非HFN组共11 300例。Logistic分析HFN的危险因素,建立HFN风险Nomgram模型,并进行验证。结果 Logistic分析发现,较低出生体质量、胎龄、出生1 min Apgar评分、胎次、出生24 h血红蛋白,及较高产妇产前抗体IgG效价均是HFN的危险因素。建模组与验证组标准曲线、矫正曲线的拟合度较好。ROC曲线分析显示,建模组曲线下面积为0.852,验证组曲线下面积为0.811,提示该风险模型具有较好预测精准度。结论 成功建立HFN的风险Nomgram模型,对临床开展早期监测、及时干预以提高新生儿预后具有积极的意义。

Aim To study the Nomgram model for personalized prediction of fetal and neonatal hemolysis(HFN)risk.Methods A total of 16450 fetuses and newborns were randomly selected as the modeling group and 1268 as the validation group.According to the hemolysis status,the fetuses and newborns in the modeling group were divided into 3882 HFN patients(HFN group),and the other 11300 non HFN patients.Logistic analysis of risk factors for HFN,establishment of a Nomgram model for HFN risk,and validation.Results Logistic analysis showed that lower birth weight,gestational age,Apgar score at 1 min after birth,parity,hemoglobin at 24 h after birth,and higher prenatal antibody IgG titer were risk factors for HFN.The standard curve and corrected curve of the modeling group and the validation group fitted well.ROC curve analysis shows that the area under the curve of the modeling group is 0.852,and that of the verification group is 0.811,indicating that the risk model has good prediction accuracy.Conclusion In neonatal delivery,timely early monitoring and timely intervention for newborns with low birth weight,gestational age,Apgar score 1min after birth,parity,hemoglobin 24 h after birth and high IgG titer of prenatal antibodies are of positive significance for improving the prognosis of newborns.