A new systematic review has found that only 53% of clinical trials are published in journal articles. The review included 39 studies which looked at how many trials were published after being entered on a clinical trial database or approved by an ethics committee. The authors of the review also found that trials with positive findings were three times more likely to be published than those with negative results and that research funded by governments was twice as likely to be published as research funded by industry.

The review did not look at how many trials had reported results via clinical trial registers rather than in journal articles. Previous research has found that results posted on trial registers contain more complete information than journal articles but only 22% of trials report their results promptly on a register in this way, even when required to do so by law.

Most national registers are set up in a way that prompts people filling them out for the essential items of information that should be included in a summary of results from the trial. This is why AllTrials is calling for reports that cover these details, whatever their format (a journal article or a summary or a grant report for example), to be added to the trial registrations, not just published elsewhere, and for fuller reports (such as the clinical study reports used in product license applications) to be made available when they are produced.

The review was written by researchers with the OPEN Project (To Overcome failure to Publish nEgative fiNdings), an initiative funded by the European Commission to gauge the impact of publication bias on medicine and to recommend ways of reducing that bias.