摘要
目的建立由超声及血清指标组成的综合诊断模型，以获得早期诊断异位妊娠（EP）患者的较佳指标组合，提高早期不明位置妊娠患者（PUL）的综合诊断水平。方法早期PUL患者随机分成建模组（184例）和测试组（90例），按最终妊娠结果分为EP组与非EP组。以拟合累积方式得出建模组每个病例EP的诊断概率，并利用受试者工作特性曲线（ROC）评价诊断指标，最后在独立的测试组中检验模型的诊断效率、正确性与稳定性。结果采用logistic逐步回归分析显示，血清孕酮、内膜厚度和对称性的意义较大，对PUL的妊娠结局有影响。受试者工作特征曲线（ROC）证实，logistic模型较单一变量的诊断性能高（P〈0．05）。当模型预测概率临界值取0．25时，对EP鉴别的敏感度和特异度分别为98．4％、92，6％，正确指数为0．91，曲线下面积为0．992。以同样标准应用于测试组，2组的诊断效率相仿。结论综合考虑子宫内膜形态学变化与相应的生化指标，将有助于提高阴道超声诊断EP的正确率，能较为准确地判断早期PUL的妊娠状态。
Objective To acquire better combination of variables of ectopic pregnancy （EP） by mean of setting up the diagnostic models which was composed of ultrasound and serum indexes and elevate the general diagnostic ability of early pregnancy of unknown location. Methods Women with PUL were rolled in and random divided into modeling group （ 184 cases） and test group （90 cases）. These cases were followed until the diagnosis was established as： EP or nonP. The diagnostic probability of EP in modeling group was calculated. Finally, model performance, such as efficiency, accuracy and stability, was evaluated with test model. Results High influence on pregnancy outcome were found when the variables （progesterone,endometrial thickness and symmetry）were calculated with the logistic stepwise regression. The diagnostic value of logistic model was higher than univariate variable according to receiver operating characteristic （ROC） （ P 〈0. 05 ）. With the cut-offs value （0. 25） of logistic-regression model to predict EP, sensitivity was 98.4% and specificity was 92. 6% and wrrect index was 91.0%. The area under the receiver operating characteristic curve （AUC） for the logistic-regression model was 0. 992. The diagnostic value of test group was similar to modeling group. Conclusions The logistic regression model which was developed by combining morphology parameters with biochemical indicators could be used to rationalize the management of PULs as it can successfully diagnose EP.
出处
《中国医师杂志》
CAS
2008年第2期188-190,共3页
Journal of Chinese Physician
