Not more than 2 sequences in full replicates [Design Issues]

posted by Helmut Homepage – Vienna, Austria, 2016-11-03 20:38 (3019 d 09:25 ago) – Posting: # 16770
Views: 5,901

Hi Mauricio,

❝ I am looking for a post in this forum that explain what is the best design to replicate a bioequivalence study.


Maybe there is one but the search function is slow those days. The data base seems to be corrupted and I didn’t have the time to repair it yet. :-(

❝ The 4x4 or 4x2? Normally, I use 4x4 this is a best choice? Or I can use 4x2 with same eficcacy?


To clarify: By 4×2 do you mean RTRT|TRTR and by 4×4 RTRT|TRTR|RRTT|TTRR or yet another one (e.g., TRRT|RTTR|TTRR|RRTT)? The notation used by PowerTOST is treatments × sequences × periods.

library(PowerTOST)
print(as.data.frame(known.designs()[8:9, ]), row.names=FALSE)

 no design    df df2 steps bk bknif   bkni                      name
  6  2x2x4 3*n-4 n-2     2  1   1/4 0.2500 2x2x4 replicate crossover
  7  2x4x4 3*n-4 n-4     4  1  1/16 0.0625 2x4x4 replicate crossover


Since both have the same degrees of freedom they should have the same efficacy.

But: IIRC, in the late 1990 Donald Schuirmann argued against more than two sequences in full replicate designs. See also Appendix B of the FDA’s guidance.

Dif-tor heh smusma 🖖🏼 Довге життя Україна! [image]
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes

Complete thread:

UA Flag
Activity
 Admin contact
23,376 posts in 4,912 threads, 1,665 registered users;
166 visitors (0 registered, 166 guests [including 10 identified bots]).
Forum time: 06:04 CET (Europe/Vienna)

There are sadistic scientists who hurry to hunt down errors
instead of establishing the truth.    Marie Curie

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5