Bioequivalence and Bioavailability Forum

Hi K@K,

❝ incase of pilot studies, this two sequence can be used in circumstances like T2 formulation is dropped/cancelled.

What will be the basis of your decision to drop T2 – and not T1? In balanced (!) cross-over designs period effects mean out. This is not the case in your design.

Let’s consider an example. Below are the true (though unknown!) values of the formulations:

R   100
T1   95 (expected to be BE)
T2  140 (superior ≠ BE)

We have period effects of +20% (P1) and –20% (P2) and get* with the two designs:

1. T1 T2 R    95 168  80
T2 R  T1  140 120  76
R  T1 T2  100 114 112

T1/R  ( 95+ 76+114)/(80+120+100) =  95.00%
T2/R  (168+140+112)/(80+120+100) = 140.00%
   
   
2. T1 R  T2   95 120 112
R  T1 T2  100 114 112

T1/R  ( 95+114)/(120+100) =  95.00%
T2/R  (112+112)/(120+100) = 101.82%

Based on the results of design #2 why would you drop T2? Due to period effects it looks better than T1…

---

• No carry-over (or unequal sequence effects). Values are the means of arbitrary sized studies. Raw scale for simplicity.