Bioequivalence and Bioavailability Forum

Hi Helmut,

❝ “We have to pay for both studies. So which is the ‘ideal’ sample size of the pilot?”

this thread is extremely interesting!

❝ n <- as.numeric(sampleN.TOST(CV=as.numeric(CVCL(CV=CV[j], df=df[k], side="upper",

❝        alpha=alpha)[2]), targetpower=target, theta0=GMR, print=F, details=F)[7])

I might indeed read your code wrongly, but if I get it right the line converts the expected CV to a kind of "almost worst-case"-CV within reason as defined by alpha. I can see why, but it makes the resulting curve somewhat pessimistic. "On average" (pardon!) the total sample sizes will tend to be lower. On the other hand we also argued a few times here that observed GMR from a pilot must be taken into account. It gets complicated...

How bout something along these lines:

1. Define a true CV and true GMR and n.pilot and a fut. criterion ("We will not conduct the pivotal if obsGMR is more than 10% different etc. or if total sample size is so-and-so.").
   
2. Simulate a pilot trial w. true CV and GMR.
   
3. Extract obsCV and obsGMR, calculate n.pivot sample size.
   
4. If fut. criterion is met, bummer, go back to 2.
   
5. Simulate a pivotal with true CV and true GMR.
   
6. Find out if it is bioequivalent, record the stats.
   
7. Repeat from pt. 2 many times.

Playing around with alpha might be relevant too.