Blind men [Two-Stage / GS Designs]

posted by d_labes  – Berlin, Germany, 2019-06-05 12:23  – Posting: # 20317
Views: 2,191

Dear Astea,

» ... But for blind studies CVaverage (that is not distinguishing treatments) is not equal to CVpooled, how to deal with it? Is there a way to prove a possibility of inflation in blind parallel studies?

Not with the tools available in Power2Stage.
But others have done the Job for you.

Friede T., Kieser M.
Sample size adjustment in clinical trials for proving equivalence
Drug Information Journal, Vol. 35, pp. 1401–1408, 2001 doi:0092-8615/2001

Citing page 1407:
... the computation of the actual level of the one-sided tests can be done following the lines of Kieser and Friede (unpublished data; 2001). The basic idea is the same as in the unblinded case. However, calculations are more complex and will, therefore, be omitted here. In all situations considered in Kieser and Friede (unpublished data, 2001), the inflation of the significance level was smaller than 0.0001. Therefore, no adjustment of the nominal significance level is necessary for the two one-sided tests procedure in case of blinded sample size adjustment.

It's a little bit annoying that they reference themselves and with "unpublished data" but nevertheless ...

Regards,

Detlew

Complete thread:

Activity
 Admin contact
20,135 posts in 4,245 threads, 1,386 registered users;
online 12 (0 registered, 12 guests [including 6 identified bots]).
Forum time (Europe/Vienna): 00:01 CET

In the beginner’s mind there are many possibilities.
In the expert’s mind there are few.    Shunryu Suzuki

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5