Complex sample analysis with proc genmod

[ChuksManuel]

Hi SAS procedure community,

I have a survey data that i want to calculate the relative risk. I know i can implement this in SAS proc genmod using dist=binomial and link=log option. However i want to account for the complex sample design (design weight, strata and cluster). 

I saw the %surveygenmod macro which is the sas version of svyglm in R. The macro isn't already in SAS. Can anyone direct me on how I can get this uploaded in my SAS environment for use?

Thanks

[SAS_Rob]

My suggestion would be to take the same approach as the example in the link below.  You would just change the GENMOD code to reflect the DIST=BIN and LINK=LOG options.

SAS/STAT Poisson Regressions for Complex Surveys

[StatDave · Accepted solution]

You should be able to do this by using PROC SURVEYLOGISTIC to fit a logistic model and then use the NLMeans macro as shown in this note.

[ChuksManuel]

Thanks!