基于logistic回归建立子宫弓状动脉超声参数对凶险性前置胎盘孕妇产后出血的风险预测模型  

Establishment of a risk prediction model for postpartum hemorrhage caused by pernicious placenta previa in pregnant women using uterine arcuate artery ultrasound parameters based on logistic regression

作　　者：黄巧如 李子琳 刘应锋 殷婉萍 陈静华 黄丽珊 周倩珺[1] HUANG Qiaoru;LI Zilin;LIU Yingfeng;YIN Wanping;CHEN Jinghua;HUANG Lishan;ZHOU Qianjun(Department of Obstetrics and Gynecology,the Tenth Affiliated Hospital of Southern Medical University(Dongguan People′s Hospital),Guangdong 523000,China)

机构地区：[1]南方医科大学第十附属医院(东莞市人民医院)妇产科,广东523000

出　　处：《中国临床新医学》2024年第11期1267-1271,共5页CHINESE JOURNAL OF NEW CLINICAL MEDICINE

基　　金：东莞市社会科技发展(一般)项目(编号:201950715001292);广东省医学科学技术研究基金项目(编号:C2019097)。

摘　　要：目的对凶险性前置胎盘(PPP)孕妇进行子宫弓状动脉超声检查,基于logistic回归建立超声参数对产后出血的风险预测模型。方法选取2019年6月至2022年6月在南方医科大学第十附属医院(东莞市人民医院)进行分娩的PPP孕妇103例,追踪其产后出血情况,根据产后出血量分为出血组(产后24 h内出血量≥1000 mL,定义为发生产后出血)及未出血组(产后24 h内出血量<1000 mL,定义为未发生产后出血)。对比两组一般资料、子宫弓状动脉超声参数,对产后出血进行多因素logistic回归分析,建立风险预测模型并进行评价,通过绘制受试者工作特征(ROC)曲线分析风险预测模型预测产后出血的价值。结果出血组收缩期峰值流速(PSV)高于未出血组,阻力指数(RI)和搏动指数(PI)显著低于未出血组(P<0.05)。子宫弓状动脉超声参数PSV、RI、PI为产后出血的影响因素(P<0.05)。建立logistic回归模型:logit(P)=-40.002+1.386×PSV-0.926×RI-0.783×PI,拟合度良好。根据拟合模型预测产后出血的曲线下面积(AUC)为0.860,大于PSV(χ^(2)=2.355,P=0.019)、RI(χ^(2)=2.332,P=0.020)、PI(χ^(2)=2.472,P=0.013)单一指标的预测。结论基于logistic回归建立子宫弓状动脉超声参数对PPP孕妇产后出血的风险预测模型可有效判断产后出血风险,有助于临床及早防治。Objective To examine the pregnant women with pernicious placenta previa(PPP)using uterine arcuate artery ultrasound,and to establish a risk prediction model for postpartum hemorrhage using ultrasound parameters based on logistic regression.Methods A total of 103 PPP pregnant women who gave birth in the Tenth Affiliated Hospital of Southern Medical University(Dongguan People′s Hospital)from June 2019 to June 2022 were selected and their postpartum hemorrhage was followed up,and the PPP pregnant women were divided into hemorrhage group(The amount of blood lost within the first 24 hours after delivery≥1000 mL was difined as the occurrence of postpartum hemorrhage)and non-hemorrhage group(The amount of blood lost within the first 24 hours after delivery<1000 mL was defined as the absence of postpartum hemorrhage)according to the amount of postpartum hemorrhage.The general data and the uterine arcuate artery ultrasound parameters were compared between the two groups,and multivariate logistic analysis was conducted on postpartum hemorrhage,and a risk prediction model was established and evaluated,and the value of the risk prediction model in predicting postpartum hemorrhage was analyzed by plotting receiver operating characteristic(ROC)curve.Results The peak systolic velocity(PSV)in the hemorrhage group was higher than that in the non-hemorrhage group,and the resistance index(RI)and pulsatility index(PI)in the hemorrhage group were lower than those in the non-hemorrhage group(P<0.05).The uterine arcuate artery ultrasound parameters of PSV,RI and PI were the influencing factors of postpartum hemorrhage(P<0.05).A logistic regression model was established:logit(P)=-40.002+1.386×PSV-0.926×RI-0.783×PI,with good fitting degree.The area under the curve(AUC)for predicting postpartum hemorrhage was 0.860 based on the fitted model,which was larger than that for predicting postpartum hemorrhage by using single indicator of PSV(χ^(2)=2.355,P=0.019),RI(χ^(2)=2.332,P=0.020)and PI(χ^(2)=2.472,P=0.013).Conclusion The risk pr

关 键 词：LOGISTIC回归 子宫弓状动脉超声参数 凶险性前置胎盘 产后出血 风险预测模型 

分 类 号：R714[医药卫生—妇产科学]