Photo attribution: flickr (Pills by Jamiesrabbits). Modified to include picture inset.

Senior Research Fellow Adam Dunn, together with colleagues from Macquarie University and Boston Children’s Hospital, is proposing a global public registry for recording the competing interests of researchers. Here he explains why. 

Published research varies across a spectrum that at one end is simply marketing masquerading as genuine inquiry. It is for this reason that researchers are expected to declare any relationships that might have affected their impartiality when they publish their work.

The problem is that our system for disclosing competing interests is incomplete, inconsistent, and fragmented. Disclosure requirements differ across journals and the statements are often buried in PDFs hiding behind paywalls. As a consequence, we do almost nothing with competing interest disclosures after they are published, leaving readers in the dark about what to do when they encounter one.

I think we can fix this problem. Together with colleagues from Boston Children’s Hospital and Macquarie University, I proposed a global public registry for recording competing interests of researchers.

The proposal was published as part of a review we wrote for the inaugural issue of Research Integrity and Peer Review, and in an accompanying piece in Nature.

While the topic of competing interests is decades-old, there are only a handful of initiatives aimed at improving the consistency and accessibility of competing interest disclosures. A registry would extend beyond these to store a public record of financial and non-financial interests for every researcher with a unique identifier, like an ORCID. To be of value, the registry needs to be comprehensive, work in concert with bibliographic databases, and be well-supported by policy.

Some have advocated for journals to reject research if it is funded by companies in certain industries. For some, like tobacco, alcohol, and gambling, the risk of bias may be consistently high enough to warrant their exclusion, but for other areas like nutrition and pharmaceuticals, the answer is not as clear.

My view is that a registry could provide a more sophisticated alternative — a way to mitigate biases without driving an even larger wedge between industry and academia. By quantifying the uncertainty introduced by different classes of competing interests, we can create tools to communicate uncertainty to readers and take it into account when using the research to guide policy and decision-making.

The current prevailing opinion seems to be that any biases caused by competing interests are somehow magically mitigated once the competing interest has been written down at the end of an article. This is clearly not the case, and I think we can do much more than to demand transparency and then ignore what is disclosed.

If you’ve got an expert opinion or commentary you’d like to share with our readers, we’d love to hear it. Email thisweek@mq.edu.au with your suggestion.