摘要
目的:构建妊娠女性重度子痫前期的辅助诊断模型,并建立个体诊断重度子痫前期的列线图。方法:回顾性分析2017年1月至2019年12月在安徽省妇幼保健院妇产科住院分娩的16387例孕妇病历资料,按是否发生重度子痫前期,分为重度子痫前期组(n=192)和非重度子痫前期组(n=16195)。对两组临床特征和生化指标等各项因素进行单因素比较,采用多因素Logistic回归分析重度子痫前期的危险因素并建立诊断模型,应用R语言软件中回归建模策略程序包建立诊断的列线图模型,采用受试者工作特征(ROC)曲线分析列线图模型对重度子痫前期的诊断效率。结果:Logistic回归分析发现,胎儿生长受限以及生化指标中尿酸、磷、同型半胱氨酸、天门冬氨酸氨基转移酶、甘油三酯、乳酸脱氢酶是重度子痫前期的危险因素(OR>1,P<0.001);白蛋白、钙是重度子痫前期的保护因素(OR<1,P<0.001)。根据多因素Logistic回归分析筛选出来的9个预测变量,建立辅助诊断重度子痫前期的列线图。ROC曲线显示列线图模型诊断重度子痫前期的AUC为0.891,最佳阀值为0.016,预测敏感度与特异度分别为89.0%、78.6%。自助抽样1000次进行内部验证,模型的AUC为0.888。结论:根据是否发生胎儿生长受限和血生化指标可以构建针对妊娠女性发生重度子痫前期的辅助诊断模型,为临床制定诊疗措施及合理转诊提供参考依据。
Objective:The aim of this study was to construct anassistant diagnostic model of severe preeclampsia in pregnant women,and to establish a nomogram model for individual diagnosis of severe preeclampsia.Methods:16387 pregnant women who deliveried in Anhui Maternal and Child Health Care Hospital from January 2017 to December 2019 were selected.They were divided into severe preeclampsia group and non-severe preeclampsia group(n=16195).Single factor comparison was used to analyze the clinical characteristics and biochemical indicators between the two groups,multiple Logistic regression was used to analyze the risk factors of severe preeclampsia in pregnant women,and the regression modeling strategies package in R software was used to establish a nomogram model to diagnose the occurrence of severe preeclampsia.ROC curve was used to analyze the efficiency of the nomogram model to diagnose severe preeclampsia in pregnancy women.Results:Univariate analysis showed that Fetal growth restriction(FGR)and biochemical indicators includinguric acid,phosphorus,homocysteine,aspartate aminotransferase,triglycerides and lactate dehydrogenase were risk factors(OR>1,P<0.001),and albumin and calcium were protect factors(OR<1,P<0.001)for severe preeclampsia in pregnancy.According to the 9 predictive variables screened by multifactorial Logistic regression analysis,a nomogram was established to assist in the diagnosis of severe preeclampsia.ROC curve showed that the area under the curve(AUC)predicted by nomogram model was 0.891,the cutoff point was 0.016,and the prediction sensitivity and specificity were 89.0%and 78.6%,respectively.In validation by bootstrap,the prediction model owned an AUC of 0.888.Conclusions:Whether FGR and blood biochemical indicators factors could provide a favorable reference for severe preeclampsia diagnosis,and formulating a reference for clinical diagnosis and reasonable referral.
作者
王霞
王群华
吴明珠
陈红波
程吉
杨湘玲
WANG Xia;WANG Qunhua;WU Mingzhu(Department of Obstetrics and Gynecology,Anhui Maternal and Child Health Care Hospital,Hefei Anhui 230001,China;Department of Obstetrics and Gynecology,The First Affiliated Hospital of USTC,Hefei Anhui 230031,China)
出处
《实用妇产科杂志》
CAS
CSCD
北大核心
2023年第8期603-608,共6页
Journal of Practical Obstetrics and Gynecology
基金
安徽省自然科学基金面上项目(编号:2108085MH260)
安徽省重点研究与开发计划项目(编号:2022e07020001)
合肥市围生医学名医工作室建设项目(编号:2018-164)。
关键词
妊娠
重度子痫前期
危险因素
诊断模型
Pregnancy
Severe preeclampsia
Risk factors
Diagnostic model
