Date:         Fri, 22 Mar 1996 16:57:12 +0000
Reply-To:     Dr D Wang <vlak520@CS6400.MCC.AC.UK>
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         Dr D Wang <vlak520@CS6400.MCC.AC.UK>
Subject:      analyzing the survival rates using SAS CATMOD
Comments: cc: tt@soton.ac.uk

Hi Friends,
 
I am seeking your help for performing a survival analysis using SAS CATMOD.
It is well known that the survival analysis can be implemented by using SAS
PHREG procedure if individual survival times are given. Here I am talking
about another type of survival data--contingency table form like:
 
 Deaths, Months of Exposure by Age of Patient and Type of Heart Valve
-------------------------------------------------------------------------
                                    Type of Heart Valve
Age                               I                 II
-------------------------------------------------------------------------
<55           No. of Deaths       4                  1
              Exposure           1259              2082
 
55+           No. of Deaths       7                  9
              Exposure           1417              1647
-------------------------------------------------------------------------
 
Here age and type of heart valve are independent variables.
 
 
Theoretically the problem can be solved by log linear modelling assuming
that the deaths are independent Possion random variables. As I know, GLIM and SPSS
can be used to make such a log linear analysis but they have their disadvantages.
I would be extremely grateful if someone could tell me the model
specifications in SAS CATMOD Procedure to solve this survival rate problem.
 
Thank you in advance
 
 
Duolao