中文摘要：

运用Logistic回归方法分析加拿大安大略省实施的中风保健系统的效用水平.与KAPRAL等的分段线性回归方法相比,Logistic回归考虑到了不同时间点上的死亡率降低速度的不同,并且假设了在实施中风保健系统之后,死亡率最终将稳定在某个值的附近,从而对加拿大安大略省实施的中风保健系统的效用有了新的认识.在文章的实证研究中,与KAPRAL等的分段线性回归方法相比,Logistic回归得到了较好的拟合效果,拟合的Logistic增长曲线通过了统计检验.此外,还得出了与KAPRAL等不同的拐点位置.

英文摘要：

We conducted a Logistic regression analysis to study the effectiveness of a system of stroke care deliv -ery promoted in Ontario , Canada .Compared with piecewise regression model used in KAPRAL etc , Logistic re-gression model considers that decreased rate of death may vary with time and assumes that death rate will even -tually remain stable at the range of a certain value after promoting Ontario Stroke System .It leads to some new understanding of the effectiveness of Ontario Stroke System .In our empirical study Logistic regression obtains a better fitting cure which is significant for the statistical test .In addition , we find alternative inflection point compared with KAPRAL etc , which may have interesting interpretation .