Here's some more speculation on Dave Sackett's comment about editors.
Dear Professor Sackett,
Some time ago, you pointed me in the direction of Dave Simel. I have read a
paper of his(1), which I found, was much more than interesting. I already
knew that what we say is significant depends upon an almost (frequentist)
statistical view of life - or results. The idea in a paper by Dave Simel and
his colleagues that there are what he calls indeterminate, intermediate and
uninterpretable results seems to me to argue against the idea of a
positive/negative cutoff of the sort that editors have to make. I understand
that there has to be a limit to what is published (even if this is simply
determined by space), but are not negative results just as important,
especially if one takes a Bayesian perspective on life?
I do not believe that the age of this paper (12 years old) make it any less
relevant. There are others, such as Alvan Feinstein, who seem to endorse
what is said here. I remain sceptical that 'evidence' can only be positive.
We cannot ignore the negative or the indeterminate, intermediate and
uninterpretable results, just because we - or editors - choose to. What is
perhaps more important is whether the results will change practice. This is
not intended as an attack on you or any other editors merely a reflection on
the limits to EBHC.
Paul Kamill, MRCGP MPH
Bradford
Reference List
1. Simel DL, Feussner JR, DeLong ER, Matchar DB. Intermediate,
indeterminate, and uninterpretable diagnostic test results.
Med.Decis.Making. 1987;7(2):107-14.
-----Original Message-----
From: Dave Sackett <[log in to unmask]>
To: Balvinder Kaur <[log in to unmask]>
Cc: Arturo Marti-Carvajal <[log in to unmask]>;
[log in to unmask] <[log in to unmask]>;
[log in to unmask] <[log in to unmask]>; [log in to unmask]
<[log in to unmask]>; [log in to unmask]
<[log in to unmask]>
Date: 14 January 1999 16:24
Subject: Re: Negative Results.
>some great comments and exchanges on this.
>
>can i add my tuppance as an editor.
>
>although research has shown that the predominant reason that "negative"
>studies don't get published is that their authors hold them back (at every
>step of analysing them, persenting them, abstracting them, and submitting
>them for publication), i frequently reject "negative" studies for my
>journal (ebm).
>
>the reason is that they aren't really "negative" (in the sense that the
>new Rx clearly doesn't work).
>
>rather, they are merely "indeterminate" (the study is so small
>[under-powered] that even a clinically-significant effect would lie within
>a confidence interval that included zero). the conclusion of such a study
>has to be: "RxA might work but, then again, it might not," and i don't
>consider that conclusion worth asking my busy clinical readers to wade
>through.
>
>hope this helps.
>
>cheers
>dls
>...........................................................................
.
>Prof David L. Sackett
>Director, NHS R&D Centre for Evidence-Based Medicine
>Consultant in Medicine Editor, Evidence-Based Medicine
>Nuffield Department of Medicine, University of Oxford
>Level 5, John Radcliffe Hospital, Oxford OX3 9DU, England
>Phone: +44-(0)1865-221320 Fax: +44-(0)1865 222901
>Email: [log in to unmask] WWW: http://cebm.jr2.ox.ac.uk
>...........................................................................
.
>
>
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|