Steve,
I think you may be referring to an early trial comparing carotid endarterectomy with standard medical care in patients with carotid stenosis. I apologize that I can't remember the name of the trial. Perhaps if you look up the NASCET trial, the earlier trial will be referred to in the introduction/rationale.
Best regards,
Janet E. Martin, BScPhm, PharmD
Project Leader
Evidence-Based Prescribing Initiative
London Health Sciences Centre
London, Ontario
CANADA N6A 4G5
[log in to unmask]
>>> "Doggett, David" <[log in to unmask]> 06/15/01 02:18PM >>>
Dear Steve,
I don't know the specific article you are looking for. However, I hope you
are not going to make the simplistic argument that intention-to-treat
analysis is always the only analysis that should be done. Awhile back there
was a series of articles on this in an issue of Statistics in Medicine
(1998, Vol. 17, articles on pp. 249, 251, 371; discussion on p. 387; also
see Demissie K et al., J Clin Epidemiol, 51:81-91, 1998). The consensus of
these was that intention-to-treat analysis and non-compliance analysis
answer different questions, regimen effectiveness and treatment efficacy,
respectively. Both of these results are valid and useful, but for different
purposes. There is no one right way; so the recommendation is that,
whenever possible, both types of analysis should be carried out, and the
different meanings of the two results should be discussed.
The same issue arises in screening studies. A policy maker may want to know
the effectiveness of screening with noncompliers included. However, a
consumer wants to know what the efficacy will be if they comply.
Interestingly, in cost-effectiveness analysis of screening, noncompliance
can sometimes be ignored, because the noncompliers will degrade
effectiveness, but at the same time may decrease costs, thereby having no
effect on the incremental cost-effetiveness ratio. This is only true if
screening increases effectiveness and also increases costs. So the effect
of intention-to-treat analysis on screening cost-effectiveness is situation
dependent.
David L. Doggett, Ph.D.
Senior Medical Research Analyst
Health Technology Assessment and Information Services
ECRI, a non-profit health services research organization
5200 Butler Pike
Plymouth Meeting, Pennsylvania 19462, U.S.A.
Phone: (610) 825-6000 x5509
FAX: (610) 834-1275
http://www.ecri.org
e-mail: [log in to unmask]
-----Original Message-----
From: Simon, Steve, PhD [mailto:[log in to unmask]]
Sent: Friday, June 15, 2001 12:05 PM
To: [log in to unmask]
Subject: Surgical example of Intention to Treat
I seem to recall a recent publication that showed why Intention to Treat is
a better approach to analysis. It involved some surgical intervention versus
a non-surgical intervention and there were a fair number of patients who
died prior to getting surgery (which is a form of non-compliance). I'd like
to be able to cite this example in an article I am writing, but I just can't
seem to track it down.
Can anybody help?
aTdHvAaNnKcSe (thanks in advance).
Steve Simon, [log in to unmask], Standard Disclaimer.
STATS: STeve's Attempt to Teach Statistics. http://www.cmh.edu/stats
Watch for a change in servers. On or around June 2001, this page will
move to http://www.childrens-mercy.org/stats
|