Bioequivalence and Bioavailability Forum

Dear Sir,
In our two-treatment, single-period parallel bioequivalence study of Iron Sucrose Injection, effects for treatments and drug formulations as factors in the statistical model.

Please see the below SAS code for Two-Way crossover design.

Without group effect SAS code for two way crossover design

PROC GLM
class sub form seq per;
model Cmax=seq sub(seq) form per;
test h=seq e=sub(seq);
lsmeans form/stderr pdiff singular=.5 out=cmax_form;
lsmeans per/stderr  pdiff singular=.5 out=cmax_per;
contrast 'T vs R' form -1 1;
estimate 'test-ref' form -1 1;
run;

Group effect sas code for two way crossover design

PROC GLM
class group seq subject period form;
model cmax=period group seq  group*seq subject(group*seq) period(group) group*form form /CLparm Alpha=.10;
     Test H=group E=subject(group*Seq);
     Test H=seq E=subject(group*seq);
     contrast 'T vs R' form -1 1;
     Estimate 'Test - Reference' form 1 -1;
     Lsmeans group/stderr pdiff E=subject(group*seq);
     Lsmeans seq/stderr pdiff E=subject(group*seq);
     Lsmeans period(Group)/Stderr Pdiff ;
     Lsmeans form/Stderr Pdiff singular=.5 out=cmax_form;
     Random subject(group*seq);
run;

So, please suggest me which changes should be made for parallel design with and without group effect.