At 14:30 28/02/2008, you wrote:
>On Thu, Feb 28, 2008 at 9:30 AM, Paul Caldwell <[log in to unmask]> wrote:
> >
> > agreed. often used in 2 ways-like antibiotics "give me something
> to make me
> > well now" or as placebo for docs who do not have access to CMHTs etc.
> > interestingly, the data is not knew and i am surprised that
> unpublished data
> > was allowed to remain so for so long. perhaps ben goldacre has a
> point: that
> > publsihing trials should be compulsory and not discretionary on the
> > trialist.
> >
> >
>
>There is a simple solution to this that all the regulatory agencies
>have declined to implement: only pre-registered trials should be
>admissible when applying for a licence. All trials would therefore
>have to be registered before starting (in case they produced good
>results), so the licensing agencies could immediately see what
>proportion of the total trials had produced positive results.
>
>Cheers Geoff
I thought that's exactly what the FDA had done?
Publication bias still remains, and who wants to publish a study that
an unheard of chemical entity doesn't work?
That only becomes newsworthy AFTER a drug is in widespread use.
OTOH the meta-analysis is probably not reliable either and is not
relevant to UK general practice where the treatment options are
different to those available in the studies.
This is a long running saga. It's also deeply rooted in human
behaviour. The dangers of resorting to nicotine in times of stress,
the habit of "getting plastered" when life is hard seem to me to be
far more dangerous than the prescriptions that GPs
issue for anti-depressants, only some of which are ever used. In GP
land these do represent BOTH useful treatments for many patients, and
tokens of care with acceptance of the degree of human suffering - a
stark alternative to the minimalist "pull yourself together" approach
(though taking responsibility for self is of course the end point we
all aim for).
In any case happy to work with patients own decisions, so long as
they don't ask me for something that doesn't exist / isn't accessible.
Julian
|