The double-edged sword of clinical data sharing

Joshua A Hirsch
Joshua A Hirsch

By Joshua A Hirsch

The Journal of NeuroInterventional Surgery (JNIS) recently featured a point/counterpoint regarding sharing clinical trial data. In fact, the participants were two of the journal’s founding editors.

At the heart of the discussion was a proposal from the International Committee of Medical Journal Editors (ICMJE) that was published in multiple lead journals entitled “Sharing Clinical Trial Data—A Proposal from the International Committee of Medical Journal Editors. ”

The ICMJE describes itself as a “small working group of general medical journal editors, whose participants meet annually and fund their own work on the Recommendations for the Conduct, Reporting, Editing and Publication of Scholarly Work in Medical Journals”. The premise of this proposal is that the editors believe there is an ethical responsibility, indeed, “obligation”, to share clinical trial data, because trial participants in a tangible way have been placed in potential harm’s way.

The proposal is that investigators would be required to share de-identified individual-patient data (IPD), underlying the results presented in an article published in a member journal. This requirement would include tables, figures, and appendices or supplementary material and be required to happen no later than six months after publication. The data underlying the results are defined as the IPD required to reproduce the article’s findings, including necessary metadata. Notably, the ICMJE proposal will require a specific indication of the author’s plan for data sharing as a component of clinical trial registration. This plan would include where the researchers will house the data and how other interested parties will be provided access to the data. The ICMJE proposes a one-year delay in implementing the proposal to minimise disruption to current efforts.

The editors of the New England Journal of Medicine (NEJM) have written a contemporaneous commentary that has received quite a bit of attention. Drs Dan Longo and Jeffrey Drazen raise concerns that people who were not involved in the generation and collection of the data may not understand the choices made in defining the parameters. Doubling down, they raise the notion that a new class of research person could emerge; people they termed “research parasites.”

Dr Joshua Hirsch supported the proposal and took aim at the notion of “research parasites”. He noted that Longo and Drazen use of this inherently negative term to make a point. Hirsch thought it would be helpful to remember that remora, oxpeckers and orchids have been, at times inaccurately described as parasites as well when in fact they provide commensal or symbiotic relationships to their hosts. He believes the metaphor is helpful in thinking through clinical data sharing.

Hirsch goes on to further argue that there are more ideas within each data set than can be explored by the people that are collecting it. Moreover, there are many more practitioners in the field of medicine with ideas than there are with the ability to generate large, randomised datasets. He points to potential research partners in private practice, who lack a strong research infrastructure. So-called “research parasites” might allow greater realisation of this academic potential through broader collaboration.

Hirsch’s final argument related to the moral imperative at work in clinical trials as exemplified by ischaemic stroke. Neurointerventional practitioners asked their vulnerable patients and families suffering from strokes caused by large vessel occlusion to participate in endovascular trials despite the fact that many lacked “personal” equipoise and believed that randomisation away from treatment put patients in harm’s way. In fact, the trials proved that this is the case. Going against our own beliefs we withheld what might have been lifesaving treatment by promoting participation in the trials. In the article, Hirsch states that neurointerventional scientists and clinicians honour that sacrifice by publicly sharing trial data to facilitate further study and commentary.

Dr David Fiorella takes a different point of view with respect to the ICMJE proposal. He asks, “Is clinical data sharing the future of clinical trials?” Reading through his response, it is fair to say that Fiorella believes the answer to be a firm and resounding no.

With brilliant flair, Fiorella turns the question back on the journals and editors that make up the ICMJE. He points out that the cost of open access publication is high and passed on to authors. If authors do not accept this cost, readers often need to pay a high fee in order to read the article. Fiorella posits that a moral obligation to study participants might include allowing the papers to be more broadly studied than the present model allows for. He thus calls for universal open access from publishing houses.

Fiorella goes on to point out that the ICMJE is creating an unfunded mandate for clinical researchers to create the mechanisms by which this data sharing occurs. He points out that the ICMJE does not offer any insight or suggestions in how to fund these requests.

Fiorella’s final argument relates to industry sponsorship of clinical trials in neurointervention. He imagines a situation in which industry sponsors would choose not to allow the publication of medical research in peer-reviewed journals, but simply submit to regulatory bodies and then use the data as they see fit to construct marketing materials. In these cases, such studies would then never be subject to the scrutiny of peer review and publication for public consumption. This IPD sharing mandate, he thus argues, could also provide a powerful overarching disincentive to industry to enter into collaborative clinical research initiatives with physicians.

Both authors recognise the legitimacy of each other’s point of view and are delighted that the JNIS provides a forum for robust discussion. We are hopeful that however this ICMJE proposal plays out, remora, oxpecker’s and orchids are made to feel welcome.


  1. Taichman DB, Backus J, Baethge C, Bauchner H, de Leeuw PW, Drazen JM, Fletcher J, Frizelle FA, Groves T, Haileamlak A, James A, Laine C, Peiperl L, Pinborg A, Sahni P, Wu S. Sharing Clinical Trial Data–A Proposal from the International Committee of Medical Journal Editors. N Engl J Med. 2016 Jan 28;374(4):384-6. doi: 10.1056/NEJMe1515172. Epub 2016 Jan 20. No abstract available. PMID: 26786954
  2. Longo DL, Drazen JM. Data Sharing. N Engl J Med. 2016 Jan 21;374(3):276-7. doi: 10.1056/NEJMe1516564.
  3. Hirsch JA. The future of clinical trials: data sharing. J Neurointerv Surg. 2016 Apr 15. pii: neurintsurg-2016-012434. doi: 10.1136/neurintsurg-2016-012434. [Epub ahead of print] No abstract available. PMID: 27084961
  4. Fiorella D. Is data sharing the future of clinical trials? J Neurointerv Surg. 2016 Apr 21. pii: neurintsurg-2016-012435. doi: 10.1136/neurintsurg-2016-012435. [Epub ahead of print] No abstract available. PMID:27102199

Joshua A Hirsch is at Massachusetts General Hospital, Boston, USA. He can be found tweeting @JoshuaAHirsch