Bioequivalence and Bioavailability Forum 02:52 CET

Main page Policy/Terms of Use Abbreviations Latest Posts

 Log in |  Register |  Search

balakotu
Regular

India,
2018-04-24 07:30

Posting: # 18726
Views: 1,136
 

 Marginally failed study for Europe [Study As­sess­ment]

Hi,
I have conducted a full replicate BE study for one of the highly variable product for Europe regulatory.
Based on the reference product variability, relaxed 90% Confidence Intervals is 75.18 to 133.01 for Cmax.
Study marginally failed with 90% upper CI data of 133.18
Is there any way to justify Europe regulatory authority(ies) to accept this study data? (DCP submission)

Regards


Edit: Category changed; see also this post #1 and #4. [Helmut]
Yura
Regular

Belarus,
2018-04-24 12:31

@ balakotu
Posting: # 18727
Views: 941
 

 Marginally failed study for Europe

Hi balakotu
1. Emissions for R-R are not considered and, therefore, it is possible that equivalence limits can be reduced.
2. Do not assessed the alpha and size of CI may be greater than the 90% CI.
Those. It may be more significant difference than it seems.:-|
Regards
Helmut
Hero
avatar
Homepage
Vienna, Austria,
2018-04-29 23:46

@ balakotu
Posting: # 18743
Views: 796
 

 Upper 90% CL 0.17% > U

Hi balakotu,

please give complete information: Sample size (if unbalanced: number of subjects per sequences), observed GMR, and 90% CI. Was it a 4-period (TRTR|RTRT) or a 3-period (TRT|RTR) design? Target power and GMR assumed in study planning?

» […] relaxed 90% Confidence Intervals

<nitpick>

You mean: Expanded (acceptance) limits

</nitpick>

» Is there any way to justify Europe regulatory authority(ies) to accept this study data?

Are you talking about ‘bending the rules’ and convince them accepting it? Chances are pretty low (patient’s risk >0.05). Furthermore, some European statisticians are already aware of the potential inflation of Type I Error in reference-scaling, which might be the case with your CVwR of 38.9%.

Cheers,
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. ☼
Science Quotes
Activity
 Thread view
Bioequivalence and Bioavailability Forum |  Admin contact
18,914 posts in 4,036 threads, 1,283 registered users;
online 15 (0 registered, 15 guests [including 13 identified bots]).

Statistics is, or should be, about scientific investigation
and how to do it better, but many statisticians believe
it is a branch of mathematics.    George E.P. Box

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