08-30-2016 06:09 AM
Hello everyone. I am conducting a project that requires Survival Analysis with Interval Censoring AND Time-dependent covariates. I understand that Proc LIFEREG can conduct Interval Censoring and Proc Phreg can deal with time dependent covariates, but neither of them can handle both.
The old way I was trying, is to do Data steps following Piecewise Exponential Model in the Allison's book, as a way to deal with time dependent covariates, then use (lower,upper) in the Model statement of Proc Lifereg, as a way to deal with Interval Censoring.
Now I found out Proc ICPhreg after SAS/Stat 13.2. But I read though the guideline and could not find any place mentioning time-dependent covariates.
Is my old way valid? or the new way, Proc ICPhreg is better for my purpose?
08-30-2016 07:45 AM - edited 08-30-2016 07:46 AM
thx. i was using my phone, now i am on my pc. but i still can't delete this one: i want to keep the one in SAS Procs