Sent to me by Bob Davis:
http://www.guardian.co.uk/letters/story/0,3604,1511582,00.html
I would add one vital element to Ben Goldacre's sensible list of
requirements for understanding health risks (Risky business, June 20).
It is essential to convey the reliability of the evidence that
underpins the risk. Much risk-related research is based on observing
statistical associations between an exposure and an outcome. But this
kind of work is notoriously unreliable and can give only a snapshot of
work in progress.
For the research to deliver reliable information on risks, huge
numbers of people have to be studied. But most research is done on
only tiny samples of the population. Medical journals and journalists
need to explain just how secure the findings they are reporting really
are.
Dr Richard Horton
Editor, The Lancet
Goldacre's original at
http://www.guardian.co.uk/comment/story/0,,1510223,00.html
Guy
--
May contain traces of irony. Contents liable to settle after posting.
http://www.chapmancentral.co.uk
88% of helmet statistics are made up, 65% of them at CHS, Puget Sound