Hi all

New to this forum. I'm working on a longitudinal mixed model and wanted to ensure the model is specified correctly. Below are the model specifications:

Dependent variable: Severity-of-illness score
Independent variables: Study intervention (treatment vs. placebo) ; Time (0,24, 72 hours) ; study site (14 sites as categorical variables) ; time*intervention interaction term
Outcome of interest: Effect of the intervention on severity-of-illness score over time (i.e. the time*intervention interaction term)

The model is currently specified as:
xtset subject_ID time
mixed score time##intervention i.site||subject_ID, var residuals (ar, t(time))
contrast intervention##time

This then gives a chi2 and p-value for the interaction term

Appreciate any thoughts!

Sam