摘要
Abstract
Objective:To systematically analyze the independent risk factors for postpartum hemorrhage(PPH)in women with gestational diabetes mellitus(GDM)and to develop and validate an individualized predictive nomogram model,aiming to provide scientific evidence for the early identification of high-risk patients and the formulation of preventive strategies.Methods:Clinical data was collected on 357 pregnant women with GDM who delivered at our institution between January 2016 and May 2023.Patients were divided into PPH group and control group based on postpartum hemorrhage occurrence,and they were divided into a training set(n=268)and a validation set(n=89)at a 7∶3 ratio using random number table method.Twenty-one clinical indicators were collected for univariate analysis,and multivariable logistic regression analysis was performed to identify the independent risk factors for PPH.A predictive nomogram was subsequently constructed using the"rms"package in R software.The model's performance was evaluated by its discrimination,measured by the area under the receiver operating characteristic(ROC)curve(AUC),and its accuracy,assessed by calibration curves.External validation was performed using the validation set.Results:The incidence of PPH in the training set was 19.03%.Univariate analysis showed statistically significant differences between the PPH group and control group in pre-pregnancy body mass index(BMI),concurrent gestational hypertensive disorders,history of uterine surgery,abnormal coagulation function,prenatal negative emotions,placental abnormalities,and macrosomia(P＜0.05).Multivariable logistic regression analysis identified five main influencing factors for PPH in GDM pregnant women:concurrent gestational hypertensive disorders,history of uterine surgery,abnormal coagulation function,prenatal negative emotions,and placental abnormalities.Based on the factors mentioned above,a predictive model was developed.Internal validation results showed an AUC of 0.822(95%CI=0.743-0.901),and the Hosmer-Lemeshow goodness-of-fit test indicated good model fit(χ²=9.171,P=0.246).External validation results showed an AUC of 0.828(95%CI=0.765-0.891),and the Hosmer-Lemeshow goodness-of-fit test also indicated good model fit(χ²=7.282,P=0.407).Conclusion:The nomogram model constructed based on five factors which are concurrent gestational hypertensive disorders,history of uterine surgery,abnormal coagulation function,prenatal negative emotions,and placental abnormalities can effectively predict the risk of PPH occurrence in GDM pregnant women,providing scientific evidence for individualized clinical prevention and intervention strategies.关键词
糖尿病,妊娠/产后出血/影响因素/列线图模型Key words
Diabetes,gestational/Postpartum hemorrhage/Influencing factors/Column chart model分类
医药卫生
