广西医学

目的分析孕妇围产期发生深静脉血栓（DVT）的影响因素，基于血清学标志物构建孕妇围产期发生DVT风险预测模型。方法回顾性分析194例孕妇的临床资料。将孕妇随机分为训练集（143例）和测试集（51例）。根据围产期是否发生DVT将训练集孕妇分为发生组（n=45）和未发生组（n=98），比较两组孕妇的临床资料及实验室指标，应用Logistic回归模型分析围产期发生DVT的影响因素。基于影响因素构建列线图预测模型，利用测试集进行外部验证，计算一致性指数并绘制校正曲线评价列线图预测模型的预测效能，并绘制决策曲线评估列线图预测模型的临床应用效果。结果发生组孕妇的年龄、妊娠期伴随疾病发生率、D-二聚体水平、同型半胱氨酸（Hcy）水平高于未发生组，25-羟基维生素D［25-（OH）D］水平低于未发生组（P＜0.05）。多因素Logistic回归模型分析结果显示，年龄、D-二聚体水平、Hcy水平及妊娠期伴随疾病是孕妇围产期发生DVT的危险因素，25-（OH）D水平是其保护因素（P＜0.05）。基于上述影响因素构建的列线图预测模型具有良好的一致性和区分度。决策曲线分析结果显示，当预测概率为0～0.59时，使用该模型预测孕妇围产期发生DVT可获得更多的临床收益。结论年龄、D-二聚体水平、Hcy水平及妊娠期伴随疾病是孕妇围产期发生DVT的危险因素，25-（OH）D水平是其保护因素，基于上述影响因素构建的列线图预测模型具有良好的预测效能和一定的临床应用价值。

ObjectiveTo analyze the influencing factors for occurrence of perinatal deep vein thrombosis (DVT) in pregnant women, and to establish a risk prediction model of occurrence of perinatal DVT in pregnant women based on serological markers. MethodsThe clinical data of 194 pregnant women were retrospectively analyzed, and they were randomly divided into training set (143 cases) or test set (51 cases). According to the presence of DVT occurred during perinatal period, pregnant women in the training set were divided into occurrence group (n=45) or non-occurrence group (n=98). The clinical data and laboratory indicators were compared between pregnant women of the two groups. The influencing factors for occurrence of perinatal DVT were analyzed by employing the Logistic regression model. Based on the influencing factors, a nomogram prediction model was established. The test set was used for external validation, the consistency index was calculated, the calibration curve was drawn to evaluate predictive efficiency of the nomogram prediction model, and the decision curve was drawn to evaluate the clinical application effect of the nomogram prediction model. ResultsThe occurrence group exhibited an older age, a higher incidence rate of pregnancy-accompanied diseases, and higher levels of D-dimer and homocysteine (Hcy), while a lower level of 25-hydroxyvitamin D (25-［OH］D) as compared with the non-occurrence group (P＜0.05). The results of multivariate Logistic regression model revealed that age, D-dimer level, Hcy level, and pregnancy-accompanied diseases were the risk factors for occurrence of perinatal DVT in pregnant women, and 25-(OH)D level was its protective factor (P＜0.05). The nomogram prediction model established based on the aforementioned influencing factors exerted favorable consistency and discrimination. The results of decision curve analysis indicated that when the prediction probability was 0-0.59, more clinical benefits could be obtained by using this model for predicting occurence of perinatal DVT in pregnant women. ConclusionAge, D-dimer level, Hcy level, and pregnancy-accompanied diseases are the risk factors for occurrence of perinatal DVT in pregnant women, and 25-(OH)D level is its protective factor. The nomogram prediction model established based on the aforesaid influencing factors exerts favorable predictive efficiency and clinical application value to a certain extent.