I'm working on geometric least-squarens mean on STATA.
My data has 2 group (HTN +-) and 3 years follow-up data of proteinuria(upcr).
I wanted to calculate geometric least-sqaures mean % change of proteinura from baseline. (Step 1)
And next, I want to calculate geometric least-squares mean ratio between 2 groups. (Step 2)
I did the Step 1 by using xtmixed code.
gen lnpercent = ln(percentupcr)
xtmixed lnpercent htn##year || no:, var
margins, over (htn year) expression (exp(preict(xb)-100))
My question is step 2, calculating geometric least-squares mean ratio between 2 groups (+95% CI) .
Thank you in advance for your kind comments
gen lnpercent = ln(percentupcr)
xtmixed lnpercent htn##year || no:, var
margins, over (htn year) expression (exp(preict(xb)-100))