Dear all,
I am trying to design a clinical trial with 3 arms: Arm A, Arm B and Arm C.
The study has to be crossover and the primary comparison will be between
treatment arms A and B.
The secondary endpoint is the comparison between arms A and C.
No comparisons should be performed between arms B and C.
So, instead of a 3x3 crossover design they proposed a 2x2 crossover (AB/BA)
and all of the patients will receive treatment C only at the final visit.
Hence, the study will end to be a 2x3 crossover study: ABC/BAC. (Treatment C
will always be the last one)
Although it sounds logical, I am not familiar with such design.
Should I analyze it as 2x2 Crossover (AB/BA - primary endpoint)? And then
what about A vs C?
Do you have any comments?
How would you call the above mentioned design? Are you aware of any
references?
Thank you very much
Evie
P. S.This design has a number of other practical advantages (nothing to do
with stats) and the client will follow it any way.
|