More things not to do

Medical journals are supposed to be biased towards publishing positive studies. Doctors are people of action. So journals like to please them by giving them things to do. The journals also like it because reprint sales of a study showing a positive effect for a drug can bring in tens if not hundreds of thousands of pounds. The BMJ, however, seems to have cornered the market in papers showing things that don't work, and this issue is a good example.

A large randomised controlled trial from Australia shows that eradicating Helicobacter pylori is not an effective …