妊娠伴甲状腺功能减退症患者自然流产风险预测模型构建

Construction of Risk Forecasting Model for Spontaneous Abortion in Patients with Pregnancy Complicated with Hypothyroidism

目的:探讨妊娠伴甲状腺功能减退症患者自然流产的危险因素并构建预测模型,旨在为自然流产高危人群准确识别及早期干预提供更多指导。方法:回顾性纳入2018年1月—2023年1月于宜春市妇幼保健院就诊的妊娠伴甲状腺功能减退症患者共278例,根据是否发生自然流产分为自然流产组(n=40)和妊娠组(n=238)。采用单因素和多因素分析评估妊娠伴甲状腺功能减退症患者自然流产的独立危险因素,并基于独立危险因素构建偏回归系数模型方程式,分析该模型用于妊娠伴甲状腺功能减退症患者自然流产风险预测的临床效能。结果:单因素分析结果显示,两组糖化血红蛋白水平、甘油三酯水平、同型半胱氨酸(Hcy)水平、促甲状腺激素(TSH)水平及甲状腺过氧化物酶抗体(TPOAb)阳性比较,差异均有统计学意义(P<0.05)。logistic回归模型多因素分析结果显示,高糖化血红蛋白水平、高甘油三酯水平、高Hcy水平、高TSH水平及TPOAb阳性均是妊娠伴甲状腺功能减退症患者自然流产的独立危险因素(P<0.05)。根据logistic回归模型证实的妊娠伴甲状腺功能减退症患者自然流产的独立危险因素,构建自变量偏回归系数模型方程式:妊娠伴甲状腺功能减退症患者自然流产风险logit(Y)=1.68+1.63×糖化血红蛋白(%)+1.85×甘油三酯(mmol/L)+1.48×Hcy(μmol/L)+0.60×TSH(mIU/L)+0.96×TPOAb阳性。受试者操作特征(ROC)曲线分析结果显示,上述临床模型用于妊娠伴甲状腺功能减退症患者自然流产风险预测的曲线下面积为0.90[95%CI(0.85,0.97)],最佳cut-off值为8.65,敏感度和特异度分别为86.15%、80.78%。结论:妊娠伴甲状腺功能减退症患者自然流产的发生与高糖化血红蛋白水平、高甘油三酯水平、高Hcy水平、高TSH水平及TPOAb阳性均有关,而基于上述指标构建的预测模型在指导临床医生早期准确识别自然流产高危人群方面显示出良好价值。

Objective:To investigate the risk factors of spontaneous abortion in patients with pregnancy complicated with hypothyroidism and construct a forecasting model,in order to provide more guidance for accurate identification and early intervention of spontaneous abortion high-risk groups.Method:A total of 278 patients with pregnancy complicated with hypothyroidism treated in Yichun Maternal and Child Health Hospital from January 2018 to January 2023 were retrospectively included,and they were divided into spontaneous abortion group(n=40)and pregnancy group(n=238)according to whether spontaneous abortion occurred.Independent risk factors for spontaneous abortion in patients with pregnancy complicated with hypothyroidism were evaluated by univariate and multivariate analysis,and a partial regression coefficient model equation was constructed based on independent risk factors,and the clinical efficacy of this model in predicting spontaneous abortion risk in patients with pregnancy complicated with hypothyroidism was analyzed.Result:The results of single factor analysis showed that there were statistically significant differences in the levels of glycosylated hemoglobin,triglyceride,homocysteine(Hcy),thyroid stimulating hormone(TSH)and thyroid peroxidase antibody(TPOAb)positive between the two groups(P<0.05).Multivariate analysis results of logistic regression model showed that high levels of glycosylated hemoglobin,triglyceride,Hcy,TSH and TPOAb positive were independent risk factors of spontaneous abortion in patients with pregnancy complicated with hypothyroidism(P<0.05).According to the independent risk factors of spontaneous abortion in patients with pregnancy complicated with hypothyroidism confirmed by logistic regression model,the independent variable partial regression coefficient model equation was constructed:the spontaneous abortion risk of patients with pregnancy complicated with hypothyroidism logit(Y)=1.68+1.63×glycosylated hemoglobin(%)+1.85×triglyceride(mmol/L)+1.48×Hcy(μmol/L)+0.60×TSH(mIU)+0.96×TPOAb positive.Receiver operating characteristic(ROC)curve analysis results showed that the area under the curve of the above clinical model for predicting the risk of spontaneous abortion in patients with pregnancy complicated with hypothyroidism was 0.90[95%CI(0.85,0.97)],the optimal cut-off value was 8.65,and the sensitivity and specificity respectively was 86.15%and 80.78%.Conclusion:The occurrence of spontaneous abortion in patients with pregnancy complicated with hypothyroidism is associated with high glycosylated hemoglobin level,high triglyceride level,high Hcy level,low TSH level and TPOAb positive,and the forecasting model based on the above indicators shows good value in guiding clinicians to accurately identify high-risk groups of spontaneous abortion at an early stage.