摘要
目的:探讨阴道分娩初产妇产后尿潴留(PUR)发生影响因素并构建风险预警模型进行验证。方法:应用目的抽样法回顾性选取2022年1月-2024年8月本院阴道分娩初产妇临床资料,出现PUR的52例为PUR组,1:2比例未出现PUR的阴道分娩初产妇104例为非PUR组。收集产妇一般人口学资料、妊娠与产时资料,应用单因素分析筛选出阴道分娩初产妇出现PUR的可能影响因素,有意义指标采用LASSO回归行降维处理,将优化后获得的因素纳入多因素logistic回归模型确定独立影响因素,通过R软件建立阴道分娩初产妇出现PUR的Nomogram风险预警模型。受试者操作特征(ROC)曲线、校准曲线与H-L检验、决策曲线与临床影响曲线判断模型预警效能、校准能力、临床效益。Bootstrap法行内部验证。结果:PUR组存在产前焦虑症状、有分娩镇痛、使用缩宫素、产钳助产、会阴侧切占比以及第一、第二产程时间、新生儿体质量均高于非PUR组,有妊娠期锻炼习惯、产后2 h内排尿占比低于非PUR组(均P<0.05)。LASSO回归获得6个非零系数变量(产前焦虑症状、分娩镇痛、产钳助产、第二产程时间、新生儿体质量、产后2 h内排尿),多因素logistic回归分析显示,产前焦虑症状、应用分娩镇痛、产钳助产、第二产程时间长为阴道分娩初产妇出现PUR危险因素,产后2 h内排尿为保护因素(P<0.05)。ROC曲线分析显示,Nomogram风险预警模型预测阴道分娩初产妇出现PUR的曲线下面积为0.925,预测敏感性0.904、特异性0.788、准确性0.827。校准曲线显示,平均绝对误差为0.014,H-L检验,χ^(2)=9.101、P=0.334。决策曲线显示,当阈值概率在0.04~0.97时Nomogram风险预警模型有较高临床受益,临床影响曲线发现阈值概率>0.6时预测高风险患者例数和实际高风险患者例数具有高度一致性。Bootstrap法内部验证显示,C指数为0.924,内部验证一致性良好。结论:依据产前焦虑症状、分娩镇痛、产钳助产、第二产程时间、产后2 h内排尿构建的Nomogram风险预警模型具有良好PUR预测性能,可为PUR风险预测提供有效工具。
Objective:To investigate the risk factors of postpartum urinary retention(PUR)of primiparas after vaginal delivery,and to construct and validate a risk warning model for PUR.Methods:The purposive sampling was used to select the clinical data of the primiparas after vaginal delivery in the hospital from January 2022 to August 2024 retrospectively.Among these primiparas,52 primiparas with PUR were included in study group and 104 primiparas without PUR were randomly selected in control group at a 1:2 ratio.The general demographic data,the information about pregnancy and delivery of the primiparas were collected.Univariate analysis was used to screen the potential risk factors of PUR of the primiparas.The significant factors were undergone dimensionality reduction through LASSO regression,and the optimized factors were incorporated into the multivariate logistic regression model to identify the independent risk factors.A Nomogram risk warning model for predicting the PUR of the primiparas after vaginal delivery was established by R software.The predictive performance,the calibration ability and the clinical utility of this model were evaluated by receiver operating characteristic(ROC)curve,calibration curve,Hosmer-Lemeshow test,decision curve and clinical impact curve.The internal validation was performed by Bootstrap method.Results:The proportions of the prenatal anxiety,the labor analgesia,the oxytocin used,the forceps delivery,the episiotomy,the durations of the first and the second stages of labor and the neonatal weight of the primiparas in the study group were significantly higher than those of the primiparas in the control group.The proportions of the exercise habits during pregnancy and the urination within 2 hours after delivery of the primiparas in the study group were significantly lower than those of the primiparas in the control group(all P<0.05).LASSO regression identified 6 non-zero coefficient variables,which included the prenatal anxiety symptoms,the labor analgesia,the forceps delivery,the second stage duration,the newborn weight and the urination within postpartum 2 hours.Multivariate logistic regression showed that the prenatal anxiety symptoms,the labor analgesia used,the forceps delivery,and the long duration of the second stage of labor of the primiparas were the risk factors of their PUR,while the urination within postpartum 2 hours of the primiparas was a protective factor of their PUR(P<0.05).ROC curve analysis showed the area under the curve(AUC),the sensitivity,the specificity and the accuracy of the Nomogram risk warning model for predicting the PUR of the primiparas after vaginal delivery were 0.925,0.904,0.788 and 0.827,respectively.The calibration curve showed the mean absolute error was 0.014,with Hosmer-Lemeshow test of χ^(2)=9.101 and P=0.334.The decision curve analysis indicated that the Nomogram risk warning model had high clinical benefit within the threshold probabilities of 0.04-0.97.The clinical impact curves showed that the high consistency between the predicted cases and the actual high-risk cases was found at the threshold probability>0.6.Bootstrap internal validation showed that the C-index was 0.924,and the internal verification consistency was good.Conclusion:The Nomogram risk warning model constructed based on the prenatal anxiety symptoms,the labor analgesia,the forceps delivery,the duration of the second stage of labor and the urination within postpartum 2 hours of the primiparas demonstrates the good predictive performance,and which can provide an effective tool for predicting the postpartum PUR risk of the primiparas.
作者
吴晓远
杜青心
马华姝
马雨鸿
王君颖
王松梅
WU Xiaoyuan;DU Qingxin;MA Huashu;MA Yuhong;WANG Junying;WANG Songmei(Xingtai People's Hospital,Xingtai,Hebei Province,054001;Xingtai Medical College)
出处
《中国计划生育学杂志》
2025年第6期1445-1450,共6页
Chinese Journal of Family Planning
基金
河北省邢台市科技局计划项目
河北省邢台市重点研发项目(2022zz084)
河北省卫生健康委医学科学研究课题计划(20252691)。
关键词
初产妇
阴道分娩
产后尿潴留
风险预警模型
Primipara
Vaginal delivery
Postpartum urinary retention
Risk warning model
