S-plus for BE test

["He Sun 301-827-5919 FAX t-" <SUNH@cder.fda.gov>]

    Dear S-Plus users:

    A practical code for conducting bioequivalence analysis (BE) on maximum
    concentration (cmax) for a two-way crossover trail using SAS PROC GLM
    could be:

    Proc glm data=data1;
    class subj seq per trt;
    model cmax=seq subj(seq) per trt/ss1 ss2 ss3 ss4;
    test h=seq e=subj(seq);
    estimate 'trt 1 vs trt 2' trt 1 -1;
    lsmeans trt/s;

    Could anyone on the list please advise me how to perform the same
    analysis using S-plus? Or please provide me the proper reading resource
    to study.  When I use glm or aov function in S-plus, I cannot get the
    same result, especially for the type 3 error output.

    Thanks a million and Happy Holidays!

    Henry Sun
    Pharmacometrics
    FDA
    e-mail  SUNH@cder.fda.gov