01-28-2015 01:14 PM
Please, I want to obtain adjusted event rate using genmod procedure. For two groups this is quite easy using the lsmeans statement. My outcome is a count of event which follows Poisson distribution.
proc genmod data=me;
model outcome=group age weight/link=log dist=poisson
Where group variable has two groups code 1 if group=0 and 0 if group=1;
Now, I have only one arm, but I will still like to obtain the adjusted rate (i.e. adjusting for age and weight). Does any one know if this is possible in the genmod procedure? I will also like to know if similar adjusted estimate is possible using the GLM procedure as I have another variable that is continuous.
02-02-2015 12:39 PM
The intercept is the source for the adjusted rate in a single arrm trial. You "adjust out" for the continuous covariates, and get the "rate" at the mean values of the covariates.