摘要
目的:评价中国冠状动脉旁路移植术(CABG)风险评估系统(SinoSCORE)预测单中心接受CABG患者术后早期病死率及并发症发生率的临床价值。方法:收集1103例行CABG患者的围手术期资料,应用SinoSCORE预测患者的病死率,通过比较实际病死率和预期病死率以评价模型的预测能力。运用Hosmer-Lemeshow拟合优度检验评价其校准度,运用受试者工作特征(ROC)曲线下面积(AUC)评价其分辨力。再应用Hosmer-Lemeshow拟合优度和AUC评价SinoSCORE对并发症发生率的预测能力。计算Youden指数得出对应最佳诊断界值(cut-off值),根据cut-off值分组,χ2检验比较两组实际病死率和并发症发生率。结果:全组术后30日内22例死亡,实际病死率1.99%。SinoSCORE预计全组病死率为3.01%,较实际病死率略高。Hosmer-Lemeshow拟合优度检验评价示SinoSCORE对于术后早期病死率校准度较好(χ2=3.473,P>0.05)。ROC检验证明SinoSCORE对于本组患者的死亡分辨力良好(AUC=0.751)。拥有良好预测价值的并发症为:低心排综合征、脑血管事件、多脏器衰竭、气管切开及术中(术后)植入主动脉球囊反搏仪(IABP)。5种并发症Hosmer-Lemeshow拟合优度检验ROC的AUC分别为0.785、0.745、0.829、0.821和0.757。cut-off值分别为6.50、5.50、3.50、3.50和5.50。结论:SinoSCORE对本中心接受CABG患者的术后早期病死率及术后低心排综合征、脑血管事件、气管切开及术中(术后)植入IABP的预测效能良好。
Objective： To predict the early mortality and postoperative complication by SinoSCORE in patients undergoing coronary artery bypass graft （CABG）. Methods： The clinical information of 1 103 patients undergoing CABG in our department was collected. The SinoSCORE was used to predict the early mortality, which was compared with the actual mortality in CABG patients. The Hosmer-Lemeshow goodness of fit test was used to access the calibration. Discrimination was tested by the area under the receiver operating characteristic （ROC） curve （AUC）. The performance of SinoSCORE on postoperative complications was detected by AUC and Hosmer-Lemeshow test. The optimal cut-off point for $inoSCORE predicting major complications was obtained by the Yonden index.The actual mortality and incidence rate of postoperative complications were compared by X~ test between two groups divided by cut-off point. Results： The observed mortality in postoperative 30 days was 1.99%. The predictive mortality calculated by the SinoSCORE was 3.01%, which was slightly higher than that of the actual mortality. The text proved SinoSCORE had a good calibration for postoperative early mortality （χ^2=3.473,P 〉 0.05）. SinoSCORE showed a higher discriminatory ability in predicting mortality by ROC （AUC =0.751）. SinoSCORE also showed the good calibration power and vary high discriminatory ability for the main complication risk evaluation such as low cardiac output syndrome, cerebmvascular events, multiple organ failure, tracheal incision and implanted aortic balloon counterpulsation （IABP）. Areas under the ROC （AUC） curves measured by hosmer-Lemeshow goodness of fit test were 0.785,0.745,0.829,0.821 and 0.757. The optimal cut-off points were 6.50,5.50,3.50,3.50 and 5.50. Conclusion： SinoSCOBE has a good forecast performance for the postoperative early mortality and some postoperative complication, including low cardiac output syndrome, cerebrovascular events, tracheal incision and intraoperative/postoperative IABP implantation, in patients undergoing CABG.
出处
《天津医药》
CAS
北大核心
2012年第11期1114-1117,共4页
Tianjin Medical Journal
