BA/BE Studies for Early Termination [Two-Stage / GS Designs]

posted by ElMaestro  – Denmark, 2015-09-22 15:35 (3109 d 22:20 ago) – Posting: # 15444
Views: 4,615

Hi balakotu,

❝ We are planning to conduct one "Replicate 4-period" bio-equivalence study with 70 subjects.

❝ After completion of clinical phase we will analyze only 40 subjects.

❝ If 40 subject’s data is shown bio-equivalent then we will stop the analysis of remaining 30 subjects. If the data is predictable then we will analyze remaining 30 subjects.


This sounds like a two-stage approach. If you allow yourself two chances of showing BE at the usual alpha of 5% then some regulators may protest due to overall type I error inflation. Check publications by Diane Potvin and coworkers for ways forward which retain control of the error.
In addition, I think you should use all available data: If you have data from 70 subjects, then I believe all 70 subjects should be analysed to avoid ethics issues.

"Replicate 4-period" in connection to two-stage designs is so far uncharted territory. I don't recommend trotting down that avenue.

Pass or fail!
ElMaestro

Complete thread:

UA Flag
Activity
 Admin contact
22,957 posts in 4,819 threads, 1,639 registered users;
79 visitors (0 registered, 79 guests [including 9 identified bots]).
Forum time: 12:56 CET (Europe/Vienna)

Nothing shows a lack of mathematical education more
than an overly precise calculation.    Carl Friedrich Gauß

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