I don't think that this is pie in the sky. There is a growing realisation (I can't understand why it took so long) that the key to deciding how big a study should be is the value of information. (Conventional power calculations are just ludicrous). There is some nice work by Peter Bacchetti and colleagues on this line. The same attitude applies in principle at least to any investigation - how much effort should I invest. Of course, there is a dange of infinite regress. How much effort should I invest deciding ho much effort to invest. How much time should I spend emailling discussions lists is another interesting question.
Yours, now returning to the academic coal face
Stephen Senn
________________________________________
From: Evidence based health (EBH) [[log in to unmask]] On Behalf Of Ben Goldacre [[log in to unmask]]
Sent: 03 November 2010 11:06
To: [log in to unmask]
Subject: Re: Too much bad research - conflicts of interest in academia
i suspect this is ludicrously pie-in-the-sky as a suggestion, but it would be interesting to see a health economic paper that calculated:
(1) the amount of time/money spent on efficacy research at the moment
(2) what is achieved (as uncertainties addressed and resolved, by numbers of patients affected)
(3) what could be achieved if it were spent more rationally.
b
On Wed, Nov 3, 2010 at 10:57 AM, Jon Brassey <[log in to unmask]<mailto:[log in to unmask]>> wrote:
Hi All,
I've just come back from the very good Evidence 2010 conference. I came away with lots of themes buzzing around in my head, but a few nicely sit together:
* There is too much research being pushed out, clinicians can't keep up to date.
* Much of the research is 'dodgy' (e.g. pharma involvement, not long enough, dubious outcomes, not patient focused etc)
The solution seems to be to stop the bloat of papers and do much less research that is better focused and has better outcomes and run for long enough. The phrase 'less is more' springs to mind! However, this solution seems to conflict with another strong theme from the conference - conflicts of interest in academia. Suggesting doing less research to academics appears to be similar to asking turkeys to vote for Christmas.
Is the solution to do less (but better) research or am I missing something? Also, given that this is ostensibly an academic list does the notion of academia being a problem/conflicted in EBM/EBHC meet with approval or horror? I appreciate I'm open to accusations of naivety (it hardly be the first time) but it the theme of academic conflicts was strong and came from multiple people (way more prominent than me)!
Best wishes
jon
Jon Brassey
TRIP Database
www.tripdatabase.com<http://www.tripdatabase.com/>
|