Commentary

Comments on the 2025 ICMJE disclosure update and what it means for digital health validation

On the expanded definition of 'relevant relationship' in the March 2025 recommendations

Lars Henriksen, PhD
Published April 2, 2025

The International Committee of Medical Journal Editors released its latest revision of the Recommendations for the Conduct, Reporting, Editing, and Publication of Scholarly Work in Medical Journals in March 2025. Most of the changes are minor editorial. One, however, has direct consequences for researchers publishing validation studies of commercially available digital health tools: the expanded definition of what counts as a “relevant relationship” for disclosure purposes.¹

What changed

The 2023 wording asked authors to disclose financial and non-financial relationships “that could be perceived to influence, or that give the appearance of potentially influencing” the work. The 2025 wording retains that language but adds an explicit clause covering relationships with the developer or vendor of a product evaluated in the work, including — and this is the notable addition — access arrangements that granted the authors unusual pre-release or privileged access to the product.² The intent, as we read it, is to close a gap that has been quietly widening in the digital-health literature, in which research groups occasionally receive pre-release access to a vendor’s system for evaluation, do not treat this as a financial relationship (because no money changes hands), and therefore disclose nothing.

We think the update is overdue. Pre-release or privileged access is a form of resource exchange that materially shapes what a research group can evaluate; it is qualitatively different from purchasing the same version of the product the public has access to. Whether one calls it a financial relationship is, in our view, the wrong question; the right question is whether a reader could reasonably want to know, and on that test the answer is yes.

How we are implementing this

The Initiative’s editorial policy was already stricter than ICMJE on this point — we disclose any pre-release access, any vendor-provided API credentials beyond what the public can obtain, and any correspondence with vendor technical staff during the evaluation period, in a dedicated section of each validation study.³ The 2025 ICMJE update is aligned with this practice, and we do not expect to change our own procedures. For groups that have not been disclosing this material, the update provides a clear public benchmark that makes continued silence harder to defend.

A note on the “access channel” question

One subtlety the ICMJE update does not fully resolve is how to disclose the channel through which public access was obtained. If researchers purchase a subscription to an app through a consumer app store, does the relevant receipt need to be disclosed? The answer, in our reading, is no — purchasing a product through the same channel any consumer would use is not a “relevant relationship” in any meaningful sense.⁴ If, however, the researchers received a subscription free of charge through an academic-access programme, that is a relationship, and on the 2025 reading it must be disclosed.

What we expect to see in the literature

Two downstream effects seem likely. First, disclosure sections in validation studies will get longer; we anticipate an average increase of two to three sentences in the COI section of affected papers over the next 18 months. Second, editors at journals that adopt ICMJE verbatim — which is most of the major clinical journals — will begin rejecting or requesting revision on submitted manuscripts that do not meet the new bar. We would encourage journals in the nutrition and digital-health space that have been slower to align with ICMJE to take this revision as a prompt to do so.⁵

References