论文部分内容阅读
目的建立受试者工作特征(ROC)曲线以评价血栓弹力图(TEG)各指标现有界值对老年人群发生出血或梗阻事件的预测效果,并为该老年人群寻找到各指标用于预测的最佳界值。方法选取2007年1月至2010年12月在中国人民解放军总医院接受TEG检测、年龄≥65岁、同时排除血液系统疾病的老年患者403例作为研究对象。收集研究对象的性别、年龄、TEG检查结果、卧床史、药物史以及疾病史;观测其出血或梗阻事件的发生情况。用研究对象ROC曲线和曲线下面积(AUC)比较TEG的凝血反应时间(R)、凝血形成时间(K)、最大振幅(MA)、凝固角(ANGLE)4项指标及其联合指标预测出血或梗阻事件发生的价值和能力。结果随访期内25例患者发生出血事件,78例患者发生梗阻事件。TEG现有界值不适用于在老年人群中进行梗阻的预测,依据约登指数最大,利用ROC曲线为各指标重新划分最佳界值,各指标的新界值分别为R=7.0 min、K=1.5 min、MA=63.5 mm、ANGLE=67.1°。将4指标并联组成的联合指标预测效果更佳,AUC达到0.602 8,灵敏度为85.9%,特异度为34.7%。利用ROC曲线为各指标预测出血事件重新划分的最佳界值分别为:R=7.8min、K=2.3 min、MA=50.5 mm、ANGLE=53.7°。将R和MA并联后的联合指标,AUC达到0.658 3,灵敏度为60.0%,特异度为71.7%。多因素Cox回归分析提示,TEG的R和MA指标具有较好的预测价值。结论建立有效的实验室检查方法及指标评价体系进行出血或梗阻的识别和预测,对老年人群预防出血和血栓栓塞性疾病的发生有积极意义。
Objective To establish a receiver operating characteristic (ROC) curve to evaluate the predictive value of the existing cut-off value of each index of thromboelastography (TEG) on bleeding or obstruction in the elderly population and to find out the indicators for the elderly population for prediction Best value. Methods From January 2007 to December 2010, 403 elderly patients aged> 65 years and at the same time excluded TBC from blood diseases were enrolled in the PLA General Hospital. The subjects’ gender, age, TEG examination results, bed rest history, drug history and disease history were collected. The occurrence of hemorrhage or obstruction was observed. The ROC curve and the area under the curve (AUC) were used to compare the TEG coagulation reaction time (R), coagulation time (K), maximum amplitude (MA), ANGLE angle and their combined indicators predict bleeding or The value and ability of an obstruction to occur. Results Twenty-five patients had bleeding events during the follow-up period and 78 patients had obstructive events. The existing cutoff value of TEG is not suitable for prediction of obstruction in the elderly population. According to the maximum YoY index, the best cutoff value of each index is recalculated by ROC curve. The new cutoff value of each index is R = 7.0 min, K = 1.5 min, MA = 63.5 mm, ANGLE = 67.1 °. The combination of four indicators in parallel to predict the joint effect of a better outcome, AUC reached 0.602 8, the sensitivity was 85.9%, the specificity was 34.7%. The best cutoff values of ROC curve for each index were R = 7.8min, K = 2.3min, MA = 50.5mm, ANGLE = 53.7 °. After combining R and MA in parallel, the AUC reached 0.658 3, the sensitivity was 60.0% and the specificity was 71.7%. Multivariate Cox regression analysis suggested that the T and R indicators of R and MA had good predictive value. Conclusion The establishment of an effective laboratory test method and index system for the identification and prediction of hemorrhage or obstruction is of positive significance for the prevention of bleeding and thromboembolic diseases in the elderly population.