[Skip to Content]
[Skip to Content Landing]
December 7, 2005

Contradictions in Highly Cited Clinical Research—Reply

JAMA. 2005;294(21):2695-2696. doi:10.1001/jama.294.21.2696-a

In Reply: I fully agree with Dr West that misinterpretation of P values may be at the root of several contradictions of postulated research findings. Indeed, medical research is increasingly conducted with very low prestudy odds of a true finding. This is a major problem in the molecular era of biomedicine that has been pointed out by several methodologists.1,2 I recently extended this framework3 to conclude that, under the current circumstances (considering typical prestudy odds, bias, number of teams working on a topic, and study design features), most published research findings are probably false.