早产高危风险因素分析及Nomogram预测模型构建

Analysis of high risk factors of premature birth and Nomogram predictionmodel construction

目的 探讨早产的高危风险因素并构建Nomogram预测模型。方法 选择2019年5月至2021年9月广东医科大学附属第二医院收治的296例早产产妇为早产组,同期108例足月产妇为足月组。收集并分析两组患者的临床资料,探索早产的相关因素;绘制受试者工作特征曲线,分析各影响因素预测早产的曲线下面积和最佳截断值;采用Logistic多元回归模型分析影响早产的独立危险因素;采用R语言软件4.0 “rms”包构建预测早产的Nomogram模型,校正曲线对Nomogram模型进行内部验证,决策曲线对Nomogram模型的预测效能进行评估。结果 与足月组相比,早产组孕妇年龄、孕周较小,体质量指数(body mass index, BMI)较大,流产次数较多(P<0.05),胎位异常、前置胎盘、妊娠期高血压疾病、妊娠期糖尿病、羊水过多、胎膜早破、多胎妊娠的比例较多(P<0.05)。年龄、孕周、BMI、流产次数AUC分别为0.883、0.855、0.809、0.780,最佳截断值为24岁、35周、28.58 kg/m2、4次。BMI(>28.58 kg/m2)、流产次数(>4次)、前置胎盘、妊娠期高血压疾病、妊娠期糖尿病、羊水过多、胎膜早破、多胎妊娠是影响早产的独立危险因素。Nomogram模型预测早产的C-index为0.674(95%CI:0.623-0.721),校准曲线显示观测值与预测值之间保持较好一致性;Nomogram模型预测早产的阈值>0.19,Nomogram模型提供临床净收益,此外,Nomogram模型临床净收益均高于BMI、流产次数、前置胎盘、妊娠期高血压疾病、妊娠期糖尿病、羊水过多、胎膜早破、多胎妊娠。结论 本研究基于BMI、流产次数、前置胎盘、妊娠期高血压疾病、妊娠期糖尿病、羊水过多、胎膜早破、多胎妊娠构建了预测早产的Nomogram模型,可用于早期识别早产的高危风险因素,有助于尽早干预,具有一定的临床应用价值。

Objective To explore the high risk factors of preterm labor and construct a Nomogram prediction model.Methods A total of 296 preterm women admitted to the Second Affiliated Hospital of Guangdong Medical University from May 2019 to September 2021 were selected as the preterm group and 108 full-term women during the same period were selected as the full-term group.Clinical data of the two groups were collected and analyzed to explore the related factors of preterm labor.The subjects working characteristic curve was drawn,and the area under the curve and the optimal cut-off value of each influencing factor were analyzed to predict the occurrence of preterm labor.Logistic multiple regression model was used to analyze the independent risk factors affecting the occurrence of preterm labor.This paper uses R language software 4.O"rms"package to construct a Nomogram model for predicting preterm labor,verifies the Nomogram model internally by correcting curves,and verifies the prediction efficacy of this model by decision curves.Results Compared with the full-term group,the preterm group had lower age,smaller gestational week,larger body mass index(BMI),more abortions(P<0.05),higher rates of abnormal fetal position,placenta previa,gestational hypertension,gestational diabetes,hyperamniotic fluid,premature rupture of membranes,and multiple pregnancies(P<0.05).Age,gestational week,BMI and AUC of abortion were 0.883,0.855,0.809 and 0.780,respectively,the optimal cut-off value was 24 years,35 weeks,28.58 kg/m^(2),4 times.BMI(>28.58 kg/m^(2)),number of abortions(>4),placenta previa,gestational hypertension,gestational diabetes,hyperamniotic fluid,premature rupture of membranes and multiple pregnancies were independent risk factors for preterm labor.The C-index of preterm labor predicted by Nomogram model was 0.674(95%CI:0.623-0.721),and the calibration curve showed a good agreement between the observed and predicted values.The threshold for preterm delivery predicted by Nomogram model>0.19,and this model provides a net clinical benefit that is higher than BMI,number of abortions,placenta previa,gestational hypertension,gestational diabetes,hyperamniotic fluid,premature rupture of membranes,and multiple pregnancies.Conclusion Based on BMI,number of abortions,placenta previa,gestational hypertension,gestational diabetes,hyperamniotic fluid,premature rupture of membranes,and multiple pregnancies,a Nomogram model for predicting preterm labor was established in this study,which can be used to identify the risk factors of preterm labor early,and facilitate early intervention.This model has certain clinical application value.