合并系统性红斑狼疮的妊娠期女性并发子痫前期的影响因素分析与预测模型建立  

Influence factors of preeclampsia in pregnant women with systemic lupus erythematosus and establishment of prediction model

作　　者：蒋萌[1] 吴珈悦 王酉[1] 常艳玲 林思涵 狄文[1] JIANG Meng;WU Jiayue;WANG You;CHANG Yanling;LIN Sihan;DI Wen(Department of Obstetrics and Gynecology,Renji Hospital,Shanghai Jiao Tong University School of Medicine,Shanghai 200127,China)

机构地区：[1]上海交通大学医学院附属仁济医院妇产科,上海200127

出　　处：《上海医学》2023年第9期605-610,共6页Shanghai Medical Journal

基　　金：国家自然科学基金(82101768、81901494);上海市公共卫生体系建设三年行动计划(2020-2022年)(GWV-10.2-YQ12)。

摘　　要：目的 探究合并系统性红斑狼疮(SLE)的妊娠期女性发生子痫前期(PE)的危险因素并建立预测模型。方法选取2010年11月—2019年6月间上海交通大学医学院附属仁济医院收治的541例妊娠合并SLE的患者资料,并最终纳入568例次(即每次妊娠视为1次独立数据)病历资料。收集患者初次建卡时(即孕16~20周)的基本信息、临床表现和实验室检查结果,以及SLE疾病活动、PE和子痫的发生情况。应用多因素logistic回归分析合并SLE的妊娠期女性发生PE的危险因素,并建立预测模型。应用Hosmer-Lemeshow和Omnibus检验评估预测模型拟合度,ROC的AUC评估模型区分能力。结果 568例次中,96例次(16.9%)妊娠期间发生PE,其中2例次并发子痫(0.4%)。单因素分析结果显示,在分类变量中,PE史、高血压、孕前SLE病情稳定<6个月,妊娠前20周疾病活动,以肾损伤、血液系统异常或浆膜炎为主要临床表现,补体C3/C4降低,核小体抗体和抗心磷脂抗体(aCL)-IgM呈阳性,以及孕期使用阿司匹林或孕后使用免疫抑制剂的占比在PE组和非PE组间的差异均有统计学意义(P值均<0.05);而在连续变量中,平均动脉压(MAP)值、血清尿酸水平、24 h尿蛋白定量在PE组和非PE组间的差异亦均有统计学意义(P值均<0.01)。将差异有统计学意义的连续变量转换为分类变量,确定MAP≥96.6 mmHg(1 mmHg=0.133 kPa)、血清尿酸≥300μmol/L、24 h尿蛋白定量≥0.272 g为合并SLE的妊娠期女性发生PE的危险因素。三者的AUC均>0.700,诊断效能可。将分类变量纳入多因素分析显示,高血压、血液系统异常、MAP≥96.6 mmHg、血清尿酸≥300μmol/L、24 h尿蛋白定量≥0.272 g是合并SLE的妊娠期女性并发PE的独立危险因素,建立包含该5项因素的预测模型。经Hosmer-Lemeshow拟合优度检验P=0.94(χ^(2)=2.96,df=8),提示预测值与观察值间无显著差异;经Omnibus检验P<0.001(χ^(2)=284.31,df=16),表明预测模型具有较好的拟�Objective To identify the risk factors of preeclampsia(PE)in pregnant women with systemic lupus erythematosus(SLE)and establish a prediction model.Methods The clinical data of 541 SLE pregnant women(568 pregnancies)admitted to Renji Hospital affiliated to Shanghai Jiao Tong University School of Medicine from November 2010 to June 2019 were retrospectively studied.The basic information at 16~20 gestational weeks of pregnancy,clinical manifestations and laboratory examination results of patients,as well as the disease activity of SLE and the incidence of PE and eclampsia were collected.The risk factors of PE in pregnant women with SLE were determined by multivariate logistic regression analysis,and a prediction model for PE was established.Hosmer-Lemeshow and Omnibus tests were used to evaluate the model fitting and area under the curve(AUC)was used to evaluate the model discrimination performance.Results Of the 568 pregnancies,96(16.9%)had PE during pregnancy,and 2(0.4%)had concurrent eclampsia.There were significant differences in the categorical variables,such as PE history,hypertension,pre-pregnancy SLE stable period<6 months,disease activity in the first 20 weeks of pregnancy,renal injury,hematological abnormalities or serositis as the main clinical manifestations,decreased complement C3/C4,positive nucleosome antibody and aCL-IgM,and the proportion of aspirin or immunosuppressive agents used during pregnancy between PE group and non-PE group(all P<0.05).There were significant differences in the continuous variables,including mean arterial pressure(MAP),serum uric acid,and 24-hour urinary protein between PE group and non-PE group(all P<0.05).It was shown that MAP≥96.6 mmHg(1 mmHg=0.133 kPa),serum uric acid≥300μmol/L,and 24-hour urinary protein≥0.272 g were risk factors of PE in women with SLE during pregnancy(after the continuous variables with statistically significant differences were converted into categorical variables).The AUC of the three risk factors of PE was greater than 0.700,and the diagnosti

关 键 词：系统性红斑狼疮 子痫前期 预测模型 高血压 尿酸 

分 类 号：R71[医药卫生—妇产科学]