Freedom to some degree [🇷 for BE/BA]

posted by d_labes  – Berlin, Germany, 2009-04-20 10:30 (6258 d 23:10 ago) – Posting: # 3568
Views: 73,734

Dear bears, dear all

❝ You can try "options (digits=n)" in R console to increase decimals, where n the number decimal you like to splay. If not work, it can be the built-in option in the package.


Of course I had tried this.
Unfortunately all this doesn't help here.

❝ But as I can remember, we had validated 90%CI obtained from bears with those from SAS with replicate study using lme(). I will post the results here later. Basically we got the equivalent 90%CIs with SAS. Indeed, we know DenDFs obtained from bears are different from those using SAS. However, 90% CIs are the same.


Don't worry about this, because SAS and bear gave the same result (also with DDFM=SATTERTHWAITE or KR) as long as
  lme(y~tmt+period+sequence, random=1|subject, data=blabla)
and the Proc MIXED equivalent of that is used. The difference comes only into effect if one tries to implement the FDA model (which is mandatory I think, at least for studies aimed to the FDA regulated market).

My questions remains:
Do I do something wrong here? If yes, what?
If not, what are the chances that regulators accept the bear / R handling of replicate studies?
—
Regards,

Detlew

Complete thread:

UA Flag
Activity
 Admin contact
23,653 posts in 4,991 threads, 1,570 registered users;
224 visitors (0 registered, 224 guests [including 42 identified bots]).
Forum time: 09:41 CEST (Europe/Vienna)

The epistemological value of probability theory is based on the fact
that chance phenomena, considered collectively and on a grand scale,
create non-random regularity.    Andrey Kolmogorov

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