王笑权
本人认为：指示变量具有分组的意义，代表观测的不同状态，而这些不同状态对生存时间的影响无疑问就是要分析的内容。
例1
data cox;/*SAS统计分析 沈其君主编。216页 已改动*/
label t='生存天数' x1='指示0为截尾，1为生存' x2='分组' x3='肾功能损害1有0无';
infile 'd:\shqj.txt';/*调用固定形式的存盘数据*/
input t 1-3 (x1-x3)(3*1.);
proc phreg;/*比例风险回归模型*/
model t*x1(0)=x2;
proc print label noobs;
run;
结果
The SAS System 23:49 Friday, December 15, 2000 6
The PHREG Procedure
Model Information
Data Set WORK.COX
Dependent Variable t Éú´æÌìÊý
Censoring Variable x1 ָʾ0Ϊ½Ø⣬1ΪÉú´æ
Censoring Value(s) 0
Ties Handling BRESLOW
Summary of the Number of Event and Censored Values
Percent
Total Event Censored Censored
25 17 8 32.00
Convergence Status
Convergence criterion (GCONV=1E-8) satisfied.
Model Fit Statistics
Without With
Criterion Covariates Covariates
-2 LOG L 94.671 93.043
AIC 94.671 95.043
SBC 94.671 95.876
Testing Global Null Hypothesis: BETA=0
Test Chi-Square DF Pr > ChiSq
Likelihood Ratio 1.6279 1 0.2020
Score 1.6058 1 0.2051
Wald 1.5544 1 0.2125
Analysis of Maximum Likelihood Estimates
Parameter Standard Hazard Variable
Variable DF Estimate Error Chi-Square Pr > ChiSq Ratio Label
x2 1 0.63654 0.51056 1.5544 0.2125 1.890 •Ö×é
The SAS System 23:49 Friday, December 15, 2000 7
Éú´æÌìÊý ָʾ0Ϊ½Ø⣬1ΪÉú´æ •Ö×é Éö¹¦ÄÜËðº¦1ÓÐ0ÎÞ
8 1 1 1
180 1 2 0
232 1 2 0
452 0 1 0
58 1 1 1
540 0 2 0
220 1 1 0
63 1 1 1
195 1 2 0
76 1 2 0
70 1 2 0
52 1 1 0
13 1 2 1
490 0 2 0
676 0 1 0
18 1 2 1
300 1 2 0
496 0 1 0
560 0 1 0
210 1 2 0
63 1 1 1
528 0 1 0
396 1 2 0
365 0 1 0
23 1 2 1
作者第二步引入x3取代x2进行观察
另外《医学统计学》人民卫生出版社 孙震球主编 有cox 分析的例子及SAS程序
