From: bvatul@ufl.edu
Subject: Covariate: statistical vs clinical
Date: Thu, 18 Jan 2001 22:42:07 -0500

Hello All

Could some one please clarify my query:

1. How to distinguish between statistically significant and clinically significant covariate?
2. In actual clinical situations where there are many confounding variables (concurrent medications etc) how reliable are covariate estimates in these cases?

Thanks a lot

Atul


*****


From: "Piotrovskij, Vladimir [JanBe]" <VPIOTROV@janbe.jnj.com>
Subject: RE: Covariate: statistical vs clinical
Date: Tue, 23 Jan 2001 09:52:29 +0100

Atul,
1. In the drug development context, a fixed effect of a covariate is clinically relevant if it changes prescription recommendations (e.g., dosing per kg body weight instead of an identical dose for everyone, or different doses for males and females, etc.)
2. I don't think there is a universal answer to this question: it depends on your specific situation. Concomitant medication issues are perhaps most complicated.

Best regards,
Vladimir
----------------------------------------------------------------------
Vladimir Piotrovsky, Ph.D.
Janssen Research Foundation
Human Pharmacokinetics (ext. 5463)
B-2340 Beerse
Belgium
Email: vpiotrov@janbe.jnj.com