产妇产后出血的危险因素分析及基于循证医学的预警信息系统构建

Risk factors analysis of postpartum hemorrhage and construction of early warning information system based on evidence-based medicine

目的探究产妇产后出血的危险因素并基于循证医学构建预警信息系统。方法选取2018年6月至2020年6月广州市妇女儿童医疗中心收取的2 872名进行系统产前检查并住院分娩的产妇作为研究对象。分为产后未出血组(n=2 467)和产后出血组(n=405)。比较两组的基线资料,采用多因素Logisitic回归分析影响产后出血的因素,构建Logisitic回归分析模型、列线图预测模型和产后出血预警模型,采用ROC曲线分析3种模型的预测效能。结果年龄≥35岁、子宫收缩乏力、胎盘因素、孕产史、孕期出血史、剖宫产、产前体重、产前宫高≥36cm、器械助产、产道裂伤、多胎妊娠、巨大儿、凝血功能障碍、子宫肌瘤、不良心理情绪均是影响产妇产后出血的独立危险因素;构建的预警模型中将产妇年龄重新划分为"35~39岁"和"≥40岁";产后出血预警模型的ROC曲线下面积、灵敏度、特异度、约登指数均明显高于Logisitic回归分析模型和列线图预测模型。结论基于循证医学构建预测产妇产后出血的预警模型可用于早期识别产后出血高危人群,有助于制订有效的预防措施,对降低产妇产后出血的发生率有重要意义。

Objective To explore the risk factors of postpartum hemorrhage and build an early warning information system based on evidence-based medicine. Methods 2,872 women who received systematic prenatal examination and were hospitalized for delivery in Guangzhou Women and Children′s Medical Center from June 2018 to June 2020 were selected and divided into two groups: group without postpartum hemorrhage(n=2,467) and group with postpartum hemorrhage(n=405). The baseline data of the two groups were compared, and multivariate Logisitic regression analysis was used to analyze the factors affecting postpartum hemorrhage, and the Logisitic regression analysis model, the histogram prediction model and the postpartum hemorrhage warning model were constructed. The prediction efficacy of the three models was analyzed by ROC curve. Results Age ≥35 years old, weak uterine contraction, placental factors, history of pregnancy, history of bleeding during pregnancy, cesarean section, prenatal weight, prenatal uterine height ≥36 cm, device-assisted delivery, birth canal laceration, multiple pregnancy, macrosomia, coagulation dysfunction, uterine fibroids, and bad psychological emotions are all independent risk factors that affect postpartum hemorrhage. In the early warning model constructed, the maternal age is reclassified as "35~39 years old" and "≥40 years old". ROC of the early warning model the area under the curve, sensitivity, specificity, and Youden index are all significantly higher than the other two prediction models. Conclusions Based on evidence-based medicine, an early warning model for predicting postpartum hemorrhage can be used to identify people at high risk of postpartum hemorrhage early, and to formulate effective preventive measures. It is of great significance to reduce the incidence of postpartum hemorrhage.