An alternative approach to evaluate the effect of an intervention at different time-points in experimental design is provided by Fitzmaurice et al (2004). In this approach a GEE or mixed model analysis i performed in which all measurements are used as outcome (including the baseline measurement). The procedure could be written as:
gen t2=2.time
gen t3=3.time
xtgee sysbp t2 t3 1.therapy#(c.t2 c.t3) , i(id) corr(exch) robust
Is there any way to calculate contrasts from this procedure?
gen t2=2.time
gen t3=3.time
xtgee sysbp t2 t3 1.therapy#(c.t2 c.t3) , i(id) corr(exch) robust
Is there any way to calculate contrasts from this procedure?
