Date: Sat, 23 Aug 2008 17:36:00 +0200
Reply-To: Stefan Pohl <stefan.pohl@ISH.DE>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Stefan Pohl <stefan.pohl@ISH.DE>
Subject: Discrete time survival model with gamma frailty
Content-Type: text/plain; charset="us-ascii"
I have discrete time survival data which can be incomplete, i.e. left
truncated and/or right censored.
I have transformed my data in person-period data and estimated a simple
binary regression model.
I used proc genmod/proc logistic to estimate the parameter estimates without
assuming frailty (neither individual nor shared).
I want to take into account for frailty assuming a gamma distribution. There
are ideas that the frailty distribution has to be updated if the data can be
left truncated. At first I want to go the naive way, i.e. I do not want to
update the frailty distribution.
Are there any SAS-macros or SAS-Procs (GLIMMIX?) which I can use?
Thank your for your help.