Dear all
I performed a longitudinal GEE model with an exchangeable correlation and
a logit link.
The outcome was visit(y/n) in a defined period of 12 months and each person
contributed a maximum of three observations to the estimation depending
if that person was seen in every one of three periods or not.
The format of the data is the following:
Id Visit_yn age other covariates
1 0
1 0
1 1
2 1
2 0
My original code:
xi: xtgee visit_yn age other_covariates, i(id) t(t_exam) fam(bin) link(logit)
corr(exch)
I have now seen in a univariate analysis that there are substantial differences
among the 6 centers and would like to adjust for this potential correlation
within people of the same center. I guess that a multilevel mixed model
would be suitable with center as the random effect.
I?m however not sure how this can be performed in Stata (probably using
glamm?)
I thought of something like
glamm visit_yn age other_covariates, i(center) link(logit) family(bin)
This does however not include the longitudinal design.
Could anybody help me with this?
Best regards
Oli
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
