What is the difference between using variables fitted and variables crossed?
For example when a subject can receive different DOSE at different visit (patient 1: DOSE 1 at VISIT 1 and VISIT 2 then DOSE 3 at VISIT 3 while patient 2: DOSE 1 at VISIT 1, DOSE 2 at VISIT 2, DOSE 3 at VISIT 3...).
See bellow for an example of the SAS code that I would use:
class patient dose visit;
model CHANGE= BASELINE DOSE VISIT [DOSE*VISIT or DOSE(VISIT)];
repeated / subject=patient type=TOEP;