Bioequivalence and Bioavailability Forum

Hi Bebac user,

❝ Could anyone explain how to conduct a sample size for replicated and partially replicated studies by fartssie tool ?

For (conventional, unscaled) ABE use the Bioequivalence, Replicate-sheet in FARTSSIE. You will note that the sample size of a 4-period full replicate is ≈½ of the 2×2×2 design. The one of a 3-period full replicate and a partial replicate are ≈¾ of the 2×2×2.
If you talk about reference-scaling (the FDA’s RSABE and the EMA’s ABEL), that’s not possible in FARTSSIE (see this thread). There is no analytical solution for power and therefore, the sample size. You need simulations (at least 100,000 for a stable result; see there). I suggest the -package PowerTOST.

Further information:

• The Vignette of PowerTOST.
  
• Articles about sample size estimation with examples:
  • ABE
    
  • RSABE
    
  • ABEL
    
  • ABE, RSABE, ABEL: A Comparison

Dear bebac user!
You can use older version (2.4 or 2.5)for sample size calculation using fully replicate/partial replicate design for HVD but not for NTI drugs. There is hardly any difference for HVD sample size outcome even when compared with PowerTost.
Regards,
Dshah