国际妇产科学杂志 ›› 2023, Vol. 50 ›› Issue (1): 88-93.doi: 10.12280/gjfckx.20220591

• 产科生理及产科疾病:论著 • 上一篇    下一篇

建立预测妊娠期高血糖患者分娩巨大儿风险的列线图模型

李慧, 赵欣, 张眉花()   

  1. 030000 山西省太原市妇幼保健院产科
  • 收稿日期:2022-07-21 出版日期:2023-02-15 发布日期:2023-03-02
  • 通讯作者: 张眉花 E-mail:zhangmeihua121@sina.com

Establishment of A Nomogram Model to Predict the Risk of Macrosomia in Patients with Hyperglycemia in Pregnancy

LI Hui, ZHAO Xin, ZHANG Mei-hua()   

  1. Department of Obstetrics, Taiyuan Maternal and Child Health Care Hospital, Taiyuan 030000, China
  • Received:2022-07-21 Published:2023-02-15 Online:2023-03-02
  • Contact: ZHANG Mei-hua E-mail:zhangmeihua121@sina.com

摘要:

目的: 构建并验证妊娠期高血糖(hyperglycemia in pregnancy,HIP)患者分娩巨大儿风险的列线图模型。 方法: 回顾性分析2020年11月—2022年2月在太原市妇幼保健院分娩的HIP患者资料。 采用多因素Logistic回归分析筛选发生巨大儿的独立影响因素, R软件构建列线图模型, 采用受试者工作特征曲线下面积对该模型的效能进行评估, 决策曲线分析(decision curve analysis,DCA)评估模型的临床使用价值。 结果: ①纳入1 098例HIP患者进行建模, 其中92例 (8.38%) 孕妇分娩巨大儿。 按7∶3比例将所有患者随机分为训练集(761例)和测试集(337例)。 ②多因素Logistic回归分析发现, 经产妇 (OR=3.19, 95%CI:1.58~6.54, P=0.001)、 高血压家族史 (OR=2.28, 95%CI:1.06~4.90, P=0.034)、 妊娠前体质量指数(OR=1.18, 95%CI:1.08~1.30, P<0.001)、 双顶径(OR=13.52, 95%CI:4.04~48.38,P<0.001)、 腹围(OR=2.83, 95%CI:2.17~3.81, P<0.001) 是孕妇分娩巨大儿的独立危险因素, 并据此建立列线图模型。 ③该模型在训练集和测试集的受试者工作特征曲线下面积分别为0.93(95%CI:0.90~0.97)和0.92(95%CI:0.88~0.97),差异无统计学意义(P=0.69), 说明模型在训练集和测试集中效果均良好。④DCA结果显示,当阈值概率≥7%时,采用该列线图预测模型可以使孕妇的净获益提高,该模型有一定的临床使用价值。结论: 初步建立了预测HIP患者分娩巨大儿的列线图模型。该模型有一定准确度,有望成为指导临床制定终止妊娠时机、进行个体化产程监护、决定分娩方式的量化工具。

关键词: 妊娠期高血糖, 巨大胎儿, 列线图, 预测, 模型, 统计学, 危险因素

Abstract:

Objective: To establish and validate a nomogram model which can predict the risk of macrosomia in patients with hyperglycemia in pregnancy (HIP). Methods: A retrospective analysis was performed on the data of pregnant women with HIP who deliveried at Taiyuan Maternal and Child Health Care Hospital from November 2020 to February 2022. Multivariate logistic regression analysis was used to screen independent influencing factors for the occurrence of macrosomia, R software was used to construct the column line graph model, the area under the receiver operator characteristic curve (AUC) was used to assess the efficacy of the model, and decision curve analysis (DCA) was used to evaluate the clinical value of the model. Results: ①A total of 1 098 HIP medical records were included in the model, and 92 (8.38%) pregnant women gave birth macrosomia. All records were randomly divided into training set (761 cases) and test set (337 cases) according to 7∶3 ratio. ② The multivariate logistic regression analysis revealed that multiparous history (OR=3.19, 95%CI: 1.58-6.54, P=0.001), family history of hypertension (OR=2.28, 95%CI: 1.06-4.90, P=0.034), pre-pregnancy body mass index (OR=1.18, 95%CI: 1.08-1.30, P<0.001), biparietal diameter (OR=13.52, 95%CI: 4.04-48.38, P<0.001) and abdominal circumference (OR=2.83, 95%CI: 2.17-3.81, P<0.001) were independent risk factors for macrosomia and the column line graph model was developed accordingly. ③ The AUC on the training and test sets were 0.93 (95%CI: 0.90-0.97) and 0.92 (95%CI: 0.88-0.97), respectively. No significant difference was observed on the area of AUC (P=0.69). The results indicated that the model worked well in both the training and test sets. ④ DCA results showed that when the threshold probability (Pt) ≥7%, the use of this nomogram prediction model can improve the net benefits of pregnant women. That is, the model has certain clinical value. Conclusions: A nomogram model which could assess the risk of macrosomia in patients with HIP was preliminarily established. The model has certain accuracy and is expected to be a quantitative tool to guide clinical timing of delivery, individual labor process monitoring, and decision of delivery mode.

Key words: Hyperglycemia in pregnancy, Fetal macrosomia, Nomograms, Forecasting, Models, statistical, Risk factors