reposted from: Listen to the first 9 minutes of Health Check from BBC World Service: http://www.bbc.co.uk/i/p00c9yfg/
Using the examples of two recent medical studies; use of aspirin to prevent cancer and the risk of using mobile phones when you are pregnant, Prof. David Spiegelhalter (who I heard giving a talk on biomedical statistics & statins at the June 2010 Cheltenham Science Festival) said that the studies to give more weight to were NOT single study laboratory non-human trials done by one group OR multi-factor related trials but INSTEAD large scale, peer reviewed, randomised controlled, systematic reviews undertaken by an authoritive group who have looked at the totality of evidence from a large numbers of people over many years and whose paper is backed up by a wide range of independent scientists.
No comments:
Post a Comment