Back to the Top
Dear all
My question is about BA/BE Anova. Imagine the following example:
A study was design as the table shows:
Week1
Week2
sequence TR
18 T
18 R
sequence RT
9 R
9 T
T means Formulation Test and R means formulation Reference. As we can
see the two sequence groups are unbalanced. This was due to 9 subjects
RT drop-outs.
What type of ANOVA should I perform ? If this design wasn't due to
drop-outs but only to a wrong design, the ANOVA conditions are the same,
or the drop-outs are treated as missing data ? What software should I
use to perform the BA/BE statistics in this cases ?
Thank you very much for your help
Best regards
Nuno Silva
Back to the Top
Dear Nuno,
From your email, I'm not quite sure whether or not partial data (i.e.
Phase I) is available for the subjects that dropped out of the RT group.
If so, then the FDA-recommended approach is to use all of your data
(including the dropout subjects) and employ Linear Mixed Effects
analysis to compute the BA/BE ANOVA. This capability has recently been
made available in WinNonlin Professional and Enterprise starting with v.
3.2. If, on the other hand, no data at all are available for these
subjects, then treating subject as a fixed effect will yield identical
results to the mixed effects approach. In which case, even earlier
versions of WinNonlin Professional or Enterprise may be used to perform
the analysis.
Please let me know if you have any further questions regarding
WinNonlin's statistical analysis capabilities, or if there is any way
that I may be of further assistance.
With Best Regards,
Mark
Mark R. Lovern, Ph. D.
Pharsight Scientific Support
5520 Dillard Drive, Suite 210
Cary, NC
PharmPK Discussion List Archive Index page
Copyright 1995-2010 David W. A. Bourne (david@boomer.org)