I am trying to run a Cox proportional hazard model. I am using proc phreg, i think that I am doing this right, but was wondering if i could have some help making sure what I think I'm doing is actually what its doing.
I want to look at disease and see if there are any incidence differences between the group that took S1 and those that took N1 (it comes out to 4 groups being yes or no for either) I have created dummy variable for S1 and N1, and I want to automatically stratify based on age, sex and smoking.
/*creating a status and censor for disease to look at the effect on subgroups (1= has the disease)*/ data status1; set status; if dis eq 1 then status= 1; if dis eq 0 then status= 2; if dis eq 1 then censor= 1; if dis eq 0 then censor= 0; run; /* S11= took S1, S10= no S1, N11= took N1, N10= no N1*/ proc phreg data= status1; model status* dis(0)= S11 S12 N11 N10 age gender smoking; run;
How would I word something like this, and if I changed the order of S1 and N1 how would my wording change on what I've done? Would this read them all seperately or because S1 is before N1 would it take N1 into account? I have made a different dummy variable for the four total groups (none, S1only, N1only, both) would this be more accurate?
Any help that you could give me with this would be great. I just want to make sure that the analysis is really doing what i think it is. Thanks!!