用逐步Logistic回归分析影响急性心肌梗塞的远期预后因素

作者对246例资料完整的急性心肌梗塞患者经过5年～10年随访,应用逐步Logistic回归方法,分析影响急性心肌梗塞的远期预后因素。结果预测生存或死亡总符合率为92.68%。证明这一方法可信而且具有一定的临床使用价值。

THE APPLICATION OF LOGISTIC REGRESSION TO ANALYSIS OF THE FACTORS INFLUENCING THE LONG-TERM PROGNOSIS OF A. M. I

The model of Logistic regression is used to analyse the factors that may influence the long-term prognosis of actue myocardial infarction(A. M. I). The analysis is based on the material taken from the complete records of 246 cases which have been selected out of the A. M. I patients who were in our hospital during 1977～1982. These cases have been followed up for more than 5 year since they recovered from their acute periods and were discharged from our hospital. During the following up interval, among these cases 72 died of the illness. The mortality is 29.26% Using the model of Logistic regression and stepwise algorithm, we have selected out 8 factors from 36 ones. The result of analysis has shown that the 8 factors are statistically significant in the affection of long-term prognosis of A.M.I and of great value in clinic practice. To confirm the result we have found, the data were substituted back for the variables in the regression equation and the mortality were calculated. It has been shown that conformity is up to 92.86% and the analysis techniques we use are believable.