TSD statistical model - with multiple sites [Two-Stage / GS Designs]

posted by d_stat – Slovenia, 2021-07-20 15:58 (1005 d 05:21 ago) – Posting: # 22482
Views: 2,802

Dear Helmut,

Thank you for sharing these references and valuable comments.

❝ BTW, the EMA’s modification of the model was shown to be irrelevant.

And if I deducted correctly, this helps that at least for FDA statistical model for TSD we can therefore omit interaction term and always combine stage data :-) - as per Karalis model looks like this (1):
Stage 1: ‘sequence’, ‘period’, ‘treatment’ and ‘subject(sequence)’ all fixed
Stage 2 (EX): ‘sequence’, ‘treatment’, ‘stage’, ‘period(stage)’ and ‘subject(sequence × stage)’ all fixed - subject(sequence × stage) can also be random, since it produces same result.

We will conduct study on multiple sites, so it adds complexity to the statistical models to be used:
Stage 1 model: Sequence, Treatment, Site, Period (Site), Sequence*Site, Treatment*Site as fixed and Subject (Sequence*Site) as random
Stage 2 model: Sequence, Treatment, Site, Stage, Period (Site*Stage), Sequence*Site, Treatment*Site, Stage*Site, Sequence*Site*Stage as fixed effects and Subject (Sequence*Site*Stage) as random - and based on (1) we can omit Treatment*Site*Stage term.

❝ It is implemented in the [image]-package Power2Stage since April 2018.

Indeed, we have used [image]-package Power2Stage calculations when discussing approach with the FDA. These packages are lifesaver :clap:
Regardless FDA still requires us to submit simulations on the validated model to justify our "specific" TSD approach. We still need to figure out what this means.

❝ I recently faced a deficiency letter of a European agency where a study (passing BE with ‘Method B’ already in the first stage) was not accepted. Passed BE with the exact method as well…

But 'Method B' success in Stage 1 means your were already within the BE limits with even wider intervals (i.e. even smaller patient risk)! Cannot image why someone would reject this? :confused:


  1. Karalis V, Macheras P. On the Statistical Model of the Two-Stage Designs in Bioequivalence Assessment. J Pharm Pharmacol. 2014; 66(1): 48–52. doi:10.1111/jphp.12164.

Complete thread:

UA Flag
 Admin contact
22,988 posts in 4,825 threads, 1,655 registered users;
36 visitors (0 registered, 36 guests [including 7 identified bots]).
Forum time: 21:20 CEST (Europe/Vienna)

The whole purpose of education is
to turn mirrors into windows.    Sydney J. Harris

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