A rather bizarre (and in my view quite incorrect) point of view
that has been put to me recently is that for placebo-controlled
trials to be interpretable we require that the placebo 'response
rate' be constant from trial to trial.
I, however, maintain that it is precisely because we fear that the
placebo 'response rate' might differ from trial to trial that we need
concurrent controls and may find that historical controls are misleading.
It is proving difficult to persuade some of the truth of this.
However, I believe that 're-calibrating samples' are used by
scientists in all sorts of areas. (Of course, for reasons described
very well by Deming, recalibration based on a little information can
be worse than doing nothing at all, but that is another matter.)
I would be grateful for hints from list members as to applications in
laboratory sciences, where measuring instruments are recalibrated
using control samples. (In particular where standard protocols for
doing this exist.) This is my opinion is analogous to adjusting the
response under an active treatment by subtracting the placebos
response. (Having chosen, of course, a suitable scale to do this on.)
Suggestions to me and I will summarise for the list in due course.
Thanks in advance.
Stephen
Stephen Senn
Professor of Statistics
Department of Statistics
15 University Gardens
<http://www.gla.ac.uk>University of Glasgow
G12 8QQ
Tel: +44 (0)141 330 5141
Fax: +44(0)141 330 4814
email [log in to unmask]
Private webpage: http://www.senns.demon.co.uk/home.html
|