中国实用妇科与产科杂志2024,Vol.40Issue(12) :1241-1244.DOI:10.19538/j.fk2024120117

妊娠合并心脏病不良事件发生危险因素分析及风险预测模型初探

Risk factors analysis and preliminary study on risk prediction model of adverse events in pregnancy complicated with heart disease

路侨 陆俊玲 王馨犹 李亚 那晶 王军 李玉岩
中国实用妇科与产科杂志2024,Vol.40Issue(12) :1241-1244.DOI:10.19538/j.fk2024120117

妊娠合并心脏病不良事件发生危险因素分析及风险预测模型初探

Risk factors analysis and preliminary study on risk prediction model of adverse events in pregnancy complicated with heart disease

路侨 1陆俊玲 1王馨犹 1李亚 1那晶 1王军 1李玉岩2
扫码查看

作者信息

  • 1. 大连医科大学附属第二医院,辽宁大连 116021
  • 2. 大连市妇女儿童医疗中心(集团),辽宁大连 116021
  • 折叠

摘要

目的 探讨妊娠合并心脏病患者发生不良事件的相关危险因素,并对相关风险预测模型进行初步探讨.方法 收集2016年1月至2022年12月146例来自大连医科大学附属第二医院的妊娠合并心脏病患者的临床资料,以母胎不良事件的发生为结局变量,采用logistic回归分析探讨结局变量的独立影响因素,并基于机器学习进行风险预测模型列线图初步探讨.结果 孕次、体重指数(BM1)、心功能分级、心脏病类型及是否行辅助生殖技术是母胎不良事件发生的独立危险因素.基于二元logistic回归结果绘制母胎不良事件风险预测模型列线图,计算不同危险因素妊娠女性发生不良事件的风险预测值,并绘制出风险预测模型列线图及多指标联合ROC曲线,ROC曲线下面积为0.90.结论 在心脏病孕妇管理中通过实时数据监控和分析以及风险预测模型的建立,可以为临床提供更加精确和实时的决策支持.

Abstract

Objective To investigate the risk factors associated with adverse events in pregnant women with heart disease,and to make a preliminary study on the related risk prediction model.Methods Clinical data of 146 pregnant patients with heart disease treated in the Second Affiliated Hospital of Dalian Medical University from January 2016 to December 2022 were collected.With the occurrence of maternal-fetal adverse events as outcome variables,logistic regression analysis was adopted to explore the independent influencing factors for outcome variables,and the risk prediction model was preliminarily explored based on machine learning.Results The number of pregnancies,BMI,cardiac function grade,type of heart disease and whether assisted reproductive technology was used were independent risk factors for adverse events.Based on the results of binary logistic regression,the nomogram of the risk prediction model for adverse maternal and fetal outcomes was drawn,and the risk prediction value of adverse events in pregnant women with different high-risk factors was calculated.The nomogram and the multi-indicator combined ROC curve were drawn,and the area under ROC curve was 0.90.Conclusion The management of pregnant women with heart disease through real-time data monitoring and analysis,as well as the establishment of predictive models,can provide more accurate and real-time decision support for clinical practice.

关键词

妊娠合并心脏病/母胎不良结局/风险预测模型

Key words

pregnancy complicated with heart disease/adverse maternal-fetal outcomes/risk prediction model

引用本文复制引用

出版年

2024
中国实用妇科与产科杂志
中国医师协会 中国实用医学杂志社

中国实用妇科与产科杂志

CSTPCD北大核心
影响因子:1.97
ISSN:1005-2216
段落导航相关论文