Last week, I made a post based on a Pharma Times story regarding the scientific/medical literature being distorted by publication bias - i.e negative findings weren't published as often as positive - which is problematic when you are describing the effects and side effect of a drug that you're seeking approval on. Both sets of results are pretty germane to the conclusion of whether it's a safe and effective treatment or not.
Today, In the Pipeline, published another reminder of this publication bias and a link to a classic paper on the subject from John Ioannides. Derek Low notes in his post that as publication bias goes - at least large Phase III studies are difficult to fake. You usually get a fairly accurate depiction in a large controlled trial.
Still it is good to remember the caveats offered by Ioannides (outlined in the In the Pipeline post) regarding the liklihood a study's results are believable or not. Take my word for it, or maybe you'd like to buy a nice bridge?
Posted by Bruce Lehr October 19th 2010.
Comments