论文部分内容阅读
目的验证Caprini和Rogers风险评估模型在胸外科手术后患者中筛选静脉血栓栓塞症(VTE)的有效性。方法采用单中心回顾性研究,以2016年7—12月首都医科大学附属北京朝阳医院胸外科行手术治疗的194例患者为样本,以Caprini和Rogers风险评估模型对所有患者进行回顾性血栓风险评分,并通过Logistic二分类回归分析得到Caprini和Rogers联合使用后的预测概率值,分别绘制受试者工作特征(ROC)曲线,计算曲线下面积(AUC),进行两两比较,以Youden指数最大的分界点作为最佳诊断分界点。结果胸外科术后总的VTE发生率为13.4%。Caprini模型AUC为(0.713±0.043,P<0.001),Rogers模型AUC为(0.577±0.062,P=0.207),预测概率模型AUC为(0.730±0.041,P<0.001)。Caprini模型分别与Rogers和预测概率模型AUC比较,差异均无统计学意义(P>0.05);但预测概率模型较Rogers模型AUC显著增加,差异有统计学意义(P=0.015)。Caprini模型在Youden指数为0.393时,敏感度为0.923,特异度为0.47;Rogers模型在Youden指数为0.135时,敏感度为0.385,特异度为0.75;预测概率模型在Youden指数为0.444时,敏感度为0.962,特异度为0.48。结论 Caprini和Rogers风险评估模型联合使用可以提高胸外科手术后患者筛选VTE的准确性。
Objectives To validate the effectiveness of the Caprini and Rogers risk assessment models for screening venous thromboembolism (VTE) in patients after thoracic surgery. Methods A one-center retrospective study was performed on 194 patients who underwent surgery in the Department of Thoracic Surgery, Beijing Chaoyang Hospital, Capital Medical University from July to December in 2016. All patients underwent retrospective thrombosis risk assessment using the Caprini and Rogers risk assessment model , And the predictive probability values of Caprini and Rogers were obtained by logistic dichotomous regression analysis. The receiver operating characteristic (ROC) curves were plotted, the area under the curve (AUC) was calculated, and the pairwise comparison was performed. The maximum Youden index Cutoff point as the best diagnostic cutoff point. Results The overall incidence of VTE after thoracic surgery was 13.4%. The AUC of Caprini model was (0.713 ± 0.043, P <0.001). The AUC of Rogers model was (0.577 ± 0.062, P = 0.207). The AUC of predictive probability model was (0.730 ± 0.041, P <0.001) Caprini model compared with Rogers and AUC, respectively. There was no significant difference between them (P> 0.05). However, the predictive probability model was significantly higher than Rogers model AUC (P = 0.015). The Caprini model with a Youden index of 0.393 had a sensitivity of 0.923 and a specificity of 0.47. The Rogers model with a Youden index of 0.135 had a sensitivity of 0.385 and a specificity of 0.75. When the Youden index was 0.444, the sensitivity of the model Was 0.962 and the specificity was 0.48. Conclusion The combination of Caprini and Rogers risk assessment model can improve the accuracy of VTE screening in patients after thoracic surgery.