比较Ｌｏｇｉｓｔｉｃ回归分析与多元判别分析对冠心病判别的效果，采用这两种分析方法分析了冠心病组、合并对照组的十一项冠心病危险因素，建立了相应的判别模型。回代结果显示Ｌｏ－ｇｉｓｔｉｃ回归模型和多元判别模型对冠心病判别的灵敏度分别为８４．００％、７８．００％，特异性分别为８２．４９％、８７．７６％，准确度分别为８３．７３％、８２．８３％。冠心病的判别分析研究设立两组。以Ｌｏｇｉｓｔｉｃ回归分析方法为宜 Logistic regression analysis and multivariate discriminant analysis were used to evaluate the effect of coronary heart disease. Eleven coronary heart disease risk factors in coronary heart disease group and control group were analyzed by these two analysis methods, and corresponding discriminant models were established. The results of regression analysis showed that the sensitivity of the Lo-gistic regression model and the multiple discriminant model were 84.00% and 78.00% respectively, the specificity was 82.49% and 87.76% respectively, and the accuracy was 83 .73%, 82.83%. Discriminant analysis of coronary heart disease set up two groups. Logistic regression analysis method is appropriate