Medical journal editors and publication bias

Wager and Williams have exposed cavalier attitudes to biased under-reporting of research among medical journal editors. 1 Only 55 (28%) of 200 websites they surveyed required clinical trial registration, and some senior editors felt that trial registration would do little to reduce publication bias. Nearly three decades ago one of us (IC) proposed outlawing the term "negative trial" and suggested that all well conceived and conducted trials-whatever their results-represent positive contributions to knowledge. 3 All medical journal editors should leave no doubt on their journal websites and elsewhere that they recognise the importance of publication bias, and make clear at www.alltrials.net that they and their journals support the demand for all clinical trials to be registered and reported.