妊娠期高血压疾病发生的危险因素及早期预测模型构建研究

Research on risk factors and construction of early prediction models for the occurrence of hypertensive disorder complicating pregnancy

目的探讨妊娠期高血压疾病(Hypertensive Disorder Complicating Pregnancy,HDCP)发生的危险因素,构建相关风险预测模型。方法采用便利抽样法,将2022年6月—2023年6月杭州市妇产科医院收治的165例HDCP患者纳入HDCP组,按照1∶2的比例选取同期330例健康孕妇纳入非HDCP组,收集孕妇病历资料,采用二元logistic回归分析孕妇发生HDCP的危险因素。应用R4.21建立HDCP风险列线图模型,并绘制校准曲线图。结果年龄≥35岁(OR=2.503)、孕前BMI≥28kg/m^(2)(OR=2.115)、初产(OR=1.869)、合并妊娠期糖尿病(OR=3.027)、有高血压家族史(OR=4.101)、多胎妊娠(OR=3.240)是孕妇发生HDCP的危险因素(P<0.05)。以此建立孕妇发生HDCP的列线图模型,模型曲线下面积AUC值为0.920,灵敏度为83.5%,特异度为86.9%。结论孕妇HDCP的发生与高龄、孕前肥胖、初产、合并妊娠期糖尿病、高血压家族史、多胎妊娠等因素有关。预测模型的区分度良好,可以科学、直观、简便地识别存在发生HDCP风险的孕妇。

Objective To explore the risk factors for gestational hypertension(HDCP)and construct a related risk pre-diction model.Methods Using convenience sampling method,165 HDCPpatients admitted to Hangzhou Obstetrics and Gynecology Hospital from June 2022 to June 2023 were included in the HDCPgroup.330 healthy pregnant women in the same period were selected in a 1:2 ratio and included in the nonHDCPgroup.Medical records of pregnant women were collected,and binary logistic regression analysis was used to analyze the risk factors for HDCP in pregnant women.Establish an HDCP risk column chart model usingR 4.21 and draw a calibration curve.Results Logistic analysis showed that age≥35 years old(OR=2.503),pre pregnancy BMI≥28 kg/m^(2)(OR=2.115),primiparous delivery(OR=1.869),comorbid gestational diabetes mellitus(OR=3.027),family history of hypertension(OR=4.101),and multiple pregnancies(OR=3.240)were risk factors for the occurrence of HDCP(P<0.05).Based on this,a column chart model was established for the occurrence of HDCP in pregnant women.The AUC value under the model curve was 0.920,with a sensitivity of 83.5% and a specificity of 86.9%.Conclusion The occurrence of HDCP in preg-nant women is related to factors such as advanced age,pre pregnancy obesity,first delivery,comorbid gestational dia-betes mellitus,family history of hypertension,and multiple pregnancies.The predictive model has good discrimination and can scientifically,intuitively,and conveniently identify pregnant women at risk of developing HDCP.