Louis52
☆    

2018-04-18 20:18

Posting: # 18698
Views: 1,434
 

 BE parallel studies with more than 2 arms [Power / Sample Size]

Hello,

Did anybody encounter a situation with a BE study with more than 2 arms (i.e. Ref 1, Ref 2, and Test)? I understand it might sound weird but it can be a real case. Is there any adjustment for the sample size calculation in this instance? Any other adjustments, like for the power, etc?

Thanks


Edit: Category changed; see also this post #1. [Helmut]
ElMaestro
★★★

Denmark,
2018-04-18 20:55

@ Louis52
Posting: # 18699
Views: 1,191
 

 BE parallel studies with more than 2 arms

Hi Louis,

» Did anybody encounter a situation with a BE study with more than 2 arms (i.e. Ref 1, Ref 2, and Test)? I understand it might sound weird but it can be a real case. Is there any adjustment for the sample size calculation in this instance? Any other adjustments, like for the power, etc?

This is a fairly common scenario.
You can dimension it conservatively by assuming the worst metric and its CV.
Multiplicity adjustment could be very relevant in this case (you want BE for test against both refs), but there is no obvious way to do it because things can be somewhat correlated. If you look at them separately and assume no correlation then you can easily power both tests to 90% and expect about an 80% chance of success if your assumptions are correct. And so forth. Note also for EU you are often removing the irrelevant data serially, but you aren't doing so for the US analysis. I am not aware of anyone having ever qualified a healthy way to deal with this aspect at the sample size stage (but I am aware of people less closely connected to real life having done -or should I say 'tried'- such a thing).
How does it look at your end, is the between-unit CV (total) for the potencies different between the two refs?

if (3) 4

x=c("Foo", "Bar")
b=data.frame(x)
typeof(b[,1]) ##aha, integer?
b[,1]+1 ##then let me add 1


Best regards,
ElMaestro

“(...) targeted cancer therapies will benefit fewer than 2 percent of the cancer patients they’re aimed at. That reality is often lost on consumers, who are being fed a steady diet of winning anecdotes about miracle cures.” New York Times (ed.), June 9, 2018.
Louis52
☆    

2018-04-23 21:20

@ ElMaestro
Posting: # 18724
Views: 1,035
 

 BE parallel studies with more than 2 arms

Probably the most conservative is the one to use. Indeed, no big difference between the CVs.
Thanks!
Activity
 Thread view
Bioequivalence and Bioavailability Forum |  Admin contact
19,405 posts in 4,122 threads, 1,323 registered users;
online 5 (0 registered, 5 guests [including 4 identified bots]).
Forum time (Europe/Vienna): 05:07 CEST

[Those] who have an excessive faith in their theories or in their
ideas are not only poorly disposed to make discoveries, but they
also make very poor observations.    Claude Bernard

The BIOEQUIVALENCE / BIOAVAILABILITY FORUM is hosted by
BEBAC Ing. Helmut Schütz
HTML5