Outliers - this one is for you, zizou [Regulatives / Guidelines]

posted by ElMaestro  – Denmark, 2021-10-31 08:01 (288 d 00:59 ago) – Posting: # 22662
Views: 843

Hi all,

some mentally deranged weirdo published a paper in the DIA journal Therapeutic Innovation & Regulatory Science. Apparently the paper makes inference about the type I error when outlier removal takes place as described in the Canadian guidance. The author writes:

"A user named “zizou” asked a question on an internet forum in 2019 [6]; paraphrased, the question was: Wouldn’t such an outlier exclusion practice result in decreased variability without affecting the expected geometric mean ratio of Test and Reference and thus lead to increased type I error (patient’s risk, risk of declaring bioequivalence for a bioinequivalent product)?"

If I am not mistaken the author is referring to this post from zizou.

The conclusion is pretty clear - zizou was entirely correct in expressing that concern. And the author is grateful :ok: :clap:. The ball is in HC's court now and the final word on this matter has not been said.

Pass or fail!

Complete thread:

UA Flag
 Admin contact
22,301 posts in 4,667 threads, 1,585 registered users;
online 3 (0 registered, 3 guests [including 2 identified bots]).
Forum time: Monday 10:01 CEST (Europe/Vienna)

Science… never solves a problem
without creating ten more.    George Bernard Shaw

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