Researchers could acquire data and publish themselves if firms refuse

If companies won’t publish invisible and abandoned trial data, researchers should take matters into their own hands, scientists at Johns Hopkins University, US, propose.

Unpublished data acquired through freedom of information requests, litigation or otherwise available to academics should be formally published to better inform medical decisionmakers, say Peter Doshi and colleagues, writing in the British Medical Journal. This would ideally be done by companies who ran the trials, but if they refuse, could be done by researchers.

Doshi and colleagues set out a series of steps in their proposal for ‘restoring invisible and abandoned trials’, calling on researchers to publish data they may already be holding, and journals to accept manuscripts including this kind of data.