天津医药 ›› 2025, Vol. 53 ›› Issue (5): 503-508.doi: 10.11958/20242396

• 临床研究 • 上一篇    下一篇

妊娠期肝内胆汁淤积症合并GDM孕妇不良妊娠结局的影响因素及预测模型构建

王欣欣(), 许慧(), 吴晓   

  1. 南京医科大学附属苏州医院,苏州市立医院产科(邮编215000)
  • 收稿日期:2025-01-06 修回日期:2025-03-21 出版日期:2025-05-15 发布日期:2025-05-28
  • 通讯作者: △ E-mail:xu_monica@163.com
  • 作者简介:王欣欣(1990),女,主治医师,主要从事妊娠期糖尿病、妊娠期肝内胆汁淤积症研究。E-mail:wendy.wxx@163.com

Influencing factors and prediction model construction of adverse pregnancy outcomes in pregnant women with intrahepatic cholestasis of pregnancy complicated with GDM

WANG Xinxin(), XU Hui(), WU Xiao   

  1. Department of Obstetrics, the Affiliated Suzhou Hospital of Nanjing Medical University, Suzhou Municipal Hospital, Suzhou 215000, China
  • Received:2025-01-06 Revised:2025-03-21 Published:2025-05-15 Online:2025-05-28
  • Contact: △ E-mail:xu_monica@163.com

摘要:

目的 分析妊娠期肝内胆汁淤积症(ICP)合并妊娠期糖尿病(GDM)孕妇不良妊娠结局的影响因素并构建风险预测模型。方法 选择214例ICP合并GDM孕妇,按3∶1比例分为建模组(161例)和验证组(53例),建模组根据妊娠结局分为结局良好组(96例)和结局不良组(65例)。对比结局良好组与结局不良组的一般临床资料和实验室指标水平,采用多因素Logistic回归分析ICP合并GDM孕妇发生不良妊娠结局的危险因素,并建立风险预测模型,利用验证组对模型进行验证。结果 多因素Logistic回归分析显示,年龄≥35岁、妊娠期血糖管理未达标、妊娠晚期阴道微生态失调、高水平总胆汁酸(TBA)、高水平白细胞介素(IL)-12是ICP合并GDM孕妇发生不良妊娠结局的独立危险因素(P<0.05)。基于多因素分析结果构建Logistic回归模型,Logit(P)=1.338×年龄+2.196×妊娠期血糖管理+2.640×妊娠晚期阴道微生态+0.112×TBA+0.090×IL-12-14.898。受试者工作特征(ROC)曲线分析示,模型的曲线下面积(AUC)为0.930,敏感度为89.23%,特异度为83.33%;Hosmer-Lemeshow拟合优度检验χ2=4.625,P=0.797。预测模型外部验证的AUC为0.939,敏感度为80.95%,特异度为96.87%;Hosmer-Lemeshow拟合优度检验χ2=6.379,P=0.605。结论 ICP合并GDM孕妇发生不良妊娠结局的影响因素包括年龄、妊娠期血糖管理情况、妊娠晚期阴道微生态情况、TBA和IL-12水平,基于此建立的风险预测模型的预测能力良好。

关键词: 胆汁淤积, 肝内, 糖尿病, 妊娠, 妊娠结局, 危险因素, Logistic模型, ROC曲线, 列线图

Abstract:

Objective To analyze the influencing factors of adverse pregnancy outcomes in pregnant women with intrahepatic cholestasis of pregnancy (ICP) complicated with gestational diabetes mellitus (GDM) and to construct a risk prediction model. Methods A total of 214 ICP pregnant women with GDM were selected and divided into the modeling group (n=161) and the verification group (n=53) according to the ratio of 3∶1. Patients in the modeling group were classified into the good outcome group (96 cases) and the poor outcome group (65 cases) by means of pregnancy outcomes. The general clinical data and laboratory indicators were compared between the good outcome group and the poor outcome group. Multivariate Logistic regression analysis was used to analyze risk factors of adverse pregnancy outcomes in pregnant women with ICP complicated with GDM, and a risk prediction model was established. The verification group was used to verify the model. Results Multivariate Logistic regression analysis showed that age≥35 years old, substandard blood glucose management during pregnancy, vaginal microecological disorder in late pregnancy, high total bile acid (TBA) and high interleukin (IL)-12 were independent risk factors of adverse pregnancy outcomes in ICP pregnant women with GDM (P<0.05). Logistic regression model was constructed based on multivariate results, and Logit (P)=1.338 × age +2.196 × blood glucose management during pregnancy + 2.640 × late pregnancy vaginal microecology + 0.112 × TBA + 0.090× IL-12-14.898. Receiver operating characteristic (ROC) curve analysis indicated that the area under the curve (AUC) value, sensitivity and specificity of the prediction model were 0.930, 89.23% and 83.33%. Hosmer-Lemeshow goodness of fit test revealed χ2=4.625 and P=0.797. The AUC, sensitivity and specificity for external validation of the model were 0.939, 80.95% and 96.87%, and Hosmer-Lemeshow goodness of fit test indicated χ2=6.379 and P=0.605. Conclusion The factors of adverse pregnancy outcomes in pregnant women with ICP complicated with GDM include age, blood glucose management during pregnancy, vaginal microecology in late pregnancy, TBA and IL-12, and the established risk prediction model based on the above factors has good predictive efficiency.

Key words: cholestasis, intrahepatic, diabetes, gestational, pregnancy outcome, risk factors, Logistic models, ROC curve, nomograms

中图分类号: