输卵管妊娠破裂的预测因素分析

Analysis of predictors of tubal pregnancy rupture

目的分析各指标的受试者工作特征曲线(ROC)及β-人绒毛膜促性腺激素(β-HCG)的临界值,探究预测输卵管妊娠破裂的理想指标。方法纳入517例输卵管妊娠患者,其中输卵管破裂103例,输卵管未破裂414例,对可能影响输卵管破裂的10项因素进行筛选,采用单因素及多因素Logistic回归分析,以ROC曲线分析各因素诊断效能,分析血清β-HCG的预测临界值。结果采用单因素分析筛选出具有的统计学意义的5个变量,它们作为自变量,输卵管妊娠破裂作为因变量,进行多因素Logistic回归分析,筛选出停经时间、包块最大直径、妊娠部位、血清β-HCG水平为4个独立危险因素。受试者工作特征曲线分析结果显示,妊娠部位、血清β-HCG水平2个指标AUC>0.7,表示有较高诊断效能。4项指标联合判断输卵管妊娠破裂风险AUC值为0.861,β-HCG的预测临界值为3756.50 IU/L。结论建议联合应用停经时间、包块最大直径、妊娠部位、血清β-HCG水平作为预测输卵管妊娠患者破裂的危险因素。

Objective To explore the ideal predictors of tubal pregnancy rupture by analyzing ROC curves of the indexes and critical values ofβ-HCG.Methods A total of 517 cases of tubal pregnancy patients were included,comprising 103 cases of tubal rupture and 414 cases of tubal non-rupture.Ten factors that may affect tubal rupture were screened.Single-factor and multiple-factor Logistic regression analyses were employed to assess the diagnostic power of each factor using receiver operating characteristic curves(ROC).The predictive critical value of serumβ-human chorionic gonadotropin(β-HCG)was analyzed.Results Univariate analysis identified five variables with statistical significance,which were subsequently used as independent variables,with tubal pregnancy rupture as the dependent variable.Multivariate Logistic regression analysis revealed four independent risk factors:menopausal time,maximum mass diameter,pregnancy site and serumβ-HCG level.Analysis of the subject performance characteristic curve indicated that pregnancy site and serumβ-HCG level both had an AUC(Area Under the Curve)>0.7,indicating high diagnostic efficacy.The AUC value of the four indexes combined to determine the risk of tubal rupture was 0.861,with the predictive critical value forβ-HCG set at 3756.50 IU/L.Conclusion It is suggested combining the use of menopausal time,maximum mass diameter,pregnancy site and serumβ-HCG level as a risk factor for predicting tubal pregnancy rupture.