BMJ Open recently published the results of a major EU-funded project (OPEN) investigating the problem of dissemination bias. Also know as publication bias, this is the distortion of the evidence base caused by selective or non-reporting of results. The authors concluded:
‘Despite various recent examples of dissemination bias and several initiatives to reduce it, the problem of dissemination bias has not been resolved.’
Many journals, including BMJ Open, have peer review processes that don’t judge importance, novelty or impact, making publishing so-called negative studies easier.
Still, the perception remains that authors find it harder to publish these studies or remain reluctant to do so. If this is true, how important is the terminology we use? Would more ‘negative’ results be published if we called them something else?
The term ‘negative result’ is used widely. The EU OPEN project used it (OPEN Stands for ‘Overcome failure to Publish nEgative findings’). It appears in journal titles, such as the Journal of Negative Results in Biomedicine. There’s a PLOS negative results collection. Professor Stephen Curry from Imperial College wrote about the importance of negative results for UK newspaper The Guardian.
BMJ Open encourages researchers to publish their studies, especially trials, regardless of results. It’s a premise of the AllTrials campaign, initiated by BMJ. In this spirit, we recently promoted this paper on Twitter. In it, Andrew Wilson and colleagues at the University of East Anglia report a randomised controlled study of maintenance sessions following pulmonary rehabilitation in patients with chronic obstructive pulmonary disease. They concluded “We do not recommend that our maintenance programme is adopted.”
To promote the paper and our willingness to consider this kind of study, we tweeted the paper with the comment that it was great that the authors were publishing their RCT ‘despite [a] negative result’.
The feedback from researchers was interesting. “Don’t talk about ‘negative results’ in trials” said Ruairidh Milne (@ruairidhm). “Best not to talk about ‘negative’ and ‘positive’ results in trials” said Shaun Treweek (@shauntreweek); “An important result rather than negative.” Andrew Cook (@ajcook) commented “I meet people who think trials which show now [sic] effect are a waste – calling them negative doesn’t help arguing how useful they are”.
All good points and we haven’t even covered that there are different types of ‘negative’ result (papers showing evidence of the absence of an effect or papers reporting an absence of evidence can both be labelled ‘negative’).
So – should we be thinking about what we call studies like these? Is the term ‘negative result’ just too… negative? Could it be influencing authors (or editors) unduly and if so what should we be calling these studies?