我国冠状动脉旁路移植术住院死亡预测模型研究及欧洲心血管手术危险评分系统效能检验

Prediction model for mortality in coronary artery bypass grafting and EuroSCORE system performance test

目的:分析并初步建立我国冠状动脉旁路移植术(coronary artery bypass grafting,CABG)风险预测模型,并与欧洲心血管手术危险因素评分系统即EuroSCORE及EuroSCORE对数回归模型对比。方法:回顾性分析2006年1月1日至2007年6月30日,北京安贞医院心脏外科行CABG手术或CABG合并手术的1637例患者资料。收集影响手术死亡的42个术前危险因素,经过单因素与多因素Logistic回归分析确立独立危险因素,并据此初步建立CABG手术死亡危险评估模型,再对模型进行分辨度、校准度检验,并与EuroSCORE及EuroSCORE对数回归模型进行对比研究。结果:全组患者年龄(61.9±9.7)岁,实际病死率4.03%(66/1637),CABG合并手术3.85%(63/1637),多因素Logistic回归分析结果:慢性肺部疾病、外周血管疾病、急性心肌梗死、既往介入治疗(PTCA、溶栓或支架)、心源性休克、主动脉瓣反流及二尖瓣反流为CABG手术死亡的独立危险因素。据此建立CABG手术死亡危险评估模型。并与EuroSCORE及EuroSCORE对数回归模型进行对比研究。受试者工作特征(Receiver-OperatingCharacteristic,ROC)曲线下面积:新建立风险模型(0.83)>EuroSCORE对数回归模型(0.82)>EuroSCORE模型(0.81),Hosmer-Lemeshowχ2检验新模型P=0.225,P>0.05,即预计病死率与实际观测病死率差异无统计学意义;而其他2种模型P<0.05。结论:慢性肺部疾病、外周血管疾病、急性心肌梗死、既往介入治疗(PTCA、溶栓或支架)、心源性休克、主动脉瓣反流及二尖瓣反流等7个因素为CABG手术死亡的独立危险因素。据此建立的CABG手术死亡危险评估模型具有良好的分辨度和校准度。

Objective：To analyze and the initial establishment of our coronary artery bypass graft surgery （ coronary artery bypass grafting, CABG ） risk prediction model, and with the EuroSCORE, EuroSCORE logistic regression model for comparison. Methods：Analyse 1 637 cases of CABG or combined CABG carried out in Beijing Anzhen Hospital,Cardiac Surgery,from January 1,2006 to June 30,2007. Collection affect the surgical mortality of 42 preoperative risk factors,through univariate and multivariate logistic regression analysis to establish an independent risk factor and a basis for the initial establishment of CABG operative mortality risk assessment model,then the discrimination and calibration of the result were tested,and risk factors for cardiovascular surgery with the European scoring system that the EuroSCORE and EuroSCORE logistic regression model for comparative study. Results：All patients aged （61. 9 ± 9. 7） years,the actual mortality rate 4. 03% （ 66 /1637 ）,CABG surgery combined 3. 85% （ 63 /1637 ）,logistic multivariate regression analysis： chronic lung disease,peripheral vascular disease,acute myocardial infarction,past intervention （ PTCA, thrombolytic therapy, or stent ）, cardiogenic shock, aortic regurgitation, mitral regurgitation operative mortality for CABG independent risk factors. Accordingly,the establishment of a risk assessment model of CABG operative mortality. And with the EuroSCORE and EuroSCORE logistic regression model for comparative study. Receiver operating characteristic （Receiver-Operating Characteristic,ROC） area under the curve： the new establishment of a risk model （0. 83） EuroSCORE logarithmic regression model （0. 82） EuroSCORE model （0. 81）,Hosmer-Lemeshow χ2 test the new model P =0. 225,P 0. 05,which is expected to the observed mortality and mortality rates were no statistically significant difference; while the other two models P 0. 05. Conclusion：Chronic lung disease,peripheral vascular disease,acute myocardial infarction,past intervention （PTCA,thrombolytic therapy,or stent）,cardiogenic shock,aortic regurgitation, mitral regurgitation seven factors such as CABG surgery an independent risk factor for death. Based on established risk assessment model of CABG operative mortality has a good degree of resolution and calibration degrees.