Can Data Sharing Become the Path of Least Resistance?

"The year 2016 could be the year when medical research converges on data sharing as a universal standard, if recent events, reflected in several PLOS Medicine articles this month, are a good indication. Attaining that standard, however, may take a little longer.

Even in morally straightforward cases, data sharing can encounter roadblocks, as discussed in a recent Policy Forum by Vasee Moorthy and colleagues at the World Health Organization [1]. WHO convened a consultation in September, inviting scientists, medical journal editors, representatives of industry, funding organizations, and government “[i]n recognition of the need to streamline mechanisms of data dissemination—globally and in as close to real-time as possible” in the context of public health emergencies. Specifically, the consultation sought to prevent the kind of delays in data sharing that may have impeded resolution of the 2014–2015 Ebola crisis. Editors attending the consultation—representing BMJ, Nature journals, New England Journal of Medicine, and the PLOS journals—were called on to address a concern that data sharing in an emergency could lead to subsequent rejection of research by journals, on the grounds of prior publication.

The editors responded with a consensus statement agreeing that journal policies should not hinder data sharing to mitigate global public health emergencies: “In such scenarios, journals should not penalize, and, indeed, should encourage or mandate public sharing of relevant data" [2]. A subsequent Comment in The Lancet expressed support for data sharing in public health emergencies by authors from major research funding organizations [3]. The International Committee of Medical Journal Editors (ICMJE), meeting in November 2015, lent further support to the principles of the WHO consultation by amending ICMJE "Recommendations" [4] to endorse data sharing for public health emergencies of any geographic scope: “In the event of a public health emergency (as defined by public health officials), information with immediate implications for public health should be disseminated without concern that this will preclude subsequent consideration for publication in a journal.”

Public health emergencies present an ethical imperative to share data from studies of all kinds. Clinical trials, which involve an obligation to make the most of information obtained from volunteers who assume the risk of participation—particularly when this information which may affect the health of much wider groups—provide another example of ongoing progress in data sharing. After considering issues of particular relevance to clinical trials, including topics presented in the Institute of Medicine report of January 2015 [5], editors of the 14 ICMJE member journals, including PLOS Medicine, jointly published a proposed plan last week that, if adopted following public comment, will require data sharing for all clinical trials that will be published in the many journals that endorse ICMJE recommendations [6]. The plan would provide a major impetus for sharing data, reminiscent of the 2004 ICMJE requirement for registration of clinical trials [7]. Implementation would mean that researchers wishing to publish clinical trials in any of the large number of participating journals, including some of the most influential venues for clinical trial reports, would need to work with their institutional ethics committees to ensure that, going forward, informed consent language permits sharing of deidentified individual participant data (IPD)...."