足月新生儿发生胎粪吸入综合征的影响因素分析及预测模型构建

李娟; 石云; 徐霞铭; 陈轩

doi:10.13898/j.cnki.issn.1000-2200.2023.12.018

足月新生儿发生胎粪吸入综合征的影响因素分析及预测模型构建

Analysis of influencing factors and prediction model construction of meconium aspiration syndrome in full-term neonates

摘要

摘要:
目的探讨足月新生儿发生胎粪吸入综合征（MAS）的影响因素并构建预测模型。
方法回顾性分析2018年3月至2021年5月产检发现有胎粪污染羊水的507例足月儿及其母亲资料，按照新生儿出生3 d内是否发生MAS，分为MAS组和非MAS组，比较2组相关指标，并进行多因素logistic回归分析，建立预测模型，采用ROC曲线分析模型区分度，采用拟合优度检验模型校准度，并选取2021年7月至2022年3月产检发现有胎粪污染羊水的184例足月儿作为模型的临床验证。
结果507例胎粪污染羊水的足月儿发生MAS共62例（12.23%）。胎儿窘迫、分娩孕周、分娩方式、羊水污染、新生儿脐血pH值、新生儿1 min Apgar评分均为足月新生儿发生MAS的独立影响因素（P < 0.05~P < 0.01）。构建得出预测模型方程为：Logit（P）=0.704×胎儿窘迫（0=否；1=是）+0.625×分娩孕周（实测值）+0.443×分娩方式（0=顺产；1=剖宫产）+1.267×羊水污染分度（0=Ⅰ度；1=Ⅱ度；2=Ⅲ度）+0.694×新生儿脐血pH值（实测值）+0.783×新生儿1 min Apgar评分（实测值）-27.894。该模型的ROC曲线下面积为0.877（95%CI：0.828~0.927），最大约登指数（0.736）对应的灵敏度和特异度分别为88.70%和84.60%。拟合优度检验显示χ²=1.13，P>0.05，表明模型不存在过拟合现象。经临床验证，模型灵敏度为87.50%，特异度为85.63%，准确率为85.87%。
结论足月妊娠产妇所娩新生儿MAS的影响因素有胎儿窘迫、分娩孕周、剖宫产、羊水污染、新生儿脐血pH值、新生儿1 min Apgar评分，以此构建预测模型，区分良好，预测效能高。

Abstract:
ObjectiveTo explore the influencing factors of meconium aspiration syndrome (MAS) in full-term neonates and construct a prediction model.
MethodsThe data of 507 full-term neonates and mothers with meconium-contaminated amniotic fluid detected by prenatal examination from March 2018 to May 2021 were retrospectively analyzed.According to whether MAS occurred within 3 days after birth, they were divided into MAS group and non-MAS group.The related indexes of the two groups were compared.Multivariate logistic regression analysis was performed.A prediction model was established.ROC curve was used to analyze the model discrimination.The goodness of fit was used to test the model calibration.A total of 184 full-term neonates with meconium-contaminated amniotic fluid from July 2021 to March 2022 were selected as the clinical verification of the model.
ResultsThere were 62 cases (12.23%) of MAS in 507 full-term neonates with meconium-contaminated amniotic fluid.The fetal distress, gestational age of delivery, mode of delivery, amniotic fluid pollution, neonatal umbilical cord blood pH value and neonatal Apgar 1 min score were independent influencing factors of MAS in full-term neonates (P < 0.05 to P < 0.01).The prediction model equation was constructed as follows: Logit (P)=0.704×fetal distress (0=no; 1=yes)+0.625×gestational age at delivery (measured value)+0.443×mode of delivery (0=natural birth; 1=cesarean section) +1.267×amniotic fluid pollution index (0=degree Ⅰ; 1=Ⅱ degree; 2=degree Ⅲ)+0.694×neonatal umbilical cord blood pH value (measured value) +0.783×neonatal Apgar 1 min score (measured value)-27.894.The area under the ROC curve was 0.877 (95%CI: 0.828-0.927).The sensitivity and specificity of the maximum Youden index (0.736) were 88.70% and 84.60%, respectively.Goodness of fit test showed χ²=1.13, P>0.05, indicating that there was no over-fitting phenomenon in the model.After clinical verification, the sensitivity of the model was 87.50%, specificity was 85.63%, and accuracy rate was 85.87%.
ConclusionsThe influencing factors of neonatal MAS in full-term pregnant women are fetal distress, gestational age, cesarean section, amniotic fluid pollution, neonatal umbilical cord blood pH value, and neonatal Apgar 1 min score.A prediction model is constructed based on above, which is well distinguished and has high prediction efficacy.

HTML全文

参考文献(19)

施引文献

资源附件(0)