Editors’ note: Today’s post is by Matt Cannon, Associate Director for Open Science Programs at Taylor & Francis.
Earlier this week, STM released a document with voluntary recommendations designed to help publishers provide clearer data sharing guidance for medical and health science researchers when they publish in academic journals. The document is the output of a Task-and-Finish group (TFG) that brought together individuals from multiple scholarly publishers and incorporated input from industry experts via two workshops. While policies for research data sharing have accelerated in momentum, this resource addresses specific information related to medical and health sciences research — a much-needed step for this community beyond existing generic guidance.
Reducing Complexity and Improving the Consistency of Data Policies
Since 2013, the number of journal and publisher data sharing policies has increased rapidly, often in response to requirements for data sharing and data management by funders and institutions. In principle, these policies aim to guide authors on the minimum requirements for sharing the research data underpinning their articles.
Initially, when these policies were being created, publishers often took inspiration from communities where data sharing was common practice — such as genomics — where the benefits of data availability had already been realized, leading to specific examples of data types and key repositories to guide authors. Additionally, as publishers developed policies in parallel (and often in isolation), there was a lot of variability across journals, causing confusion and additional burden for researchers in understanding and meeting these requirements.
Where Generic Data Policies Fail Medical and Health Researchers
The need to go beyond general guidance was illustrated clearly by a 2023 survey by Open Pharma, which brings pharmaceutical companies, publishers, and other stakeholders together to promote open research practices in medical publishing. Respondents reported that medical journal data sharing policies were not fit for purpose: companies said that they often did not have enough specific information to understand journal requirements when preparing their submissions. Some reported that policy ambiguity was actively influencing where they chose to publish; others described delays to their timelines for sharing results.
What STM Recommends
STM has been a long-time supporter of collaborative approaches towards research data sharing (see examples in Learned Publishing, D-Lib Magazine, and STM’s Joint Statement on Research Data). This new effort extends that work by focusing specifically on how medical and health research publishers could communicate to prospective authors with specific language and appropriate detail to be able to support responsible data sharing in their domain.
Using input from workshops with publishers and pharmaceutical companies and the structure of the RDA Data Policy Standardization and Implementation IG recommendations & outputs, three publisher representatives (Matthew Cannon and Rebecca Taylor-Grant for Taylor & Francis and Rhiannon Meaden for OUP) created this new voluntary guidance. The document includes best-practice examples of journals giving very specific information to researchers on how to define, share and store their medical and health sciences data.
The aim is that publishers or journal managers can use the new documentation to support journals in being more specific with their guidance to authors. The document has headings for each data policy feature, with explanatory text and a best-practice example.
Key topics include:
- Defining what counts as research data
- Handling exceptions for sensitive or identifiable data
- Facilitating access to data during peer review
- Selecting repositories, including controlled access
- Managing data embargoes
- Improving Data Availability Statements
- Specifying data formats and standards
- Clarifying licensing requirements
- Helping authors understand the FAIR principles
Three Areas Where Clearer Guidance Could Make the Biggest Difference
Through the workshops, there were consistent themes in the feedback from attendees: lack of clarity around what data was required to share; how to have a discussion with the journal/publisher when authors felt they had a valid exception to a data sharing requirement; and being able to use specific repositories. Therefore, the following sections of the documentation are expected to have the biggest opportunity to improve the experiences for authors when submitting and publishing their research.
- Defining what data means in the context of the journal/portfolio
Clearer definitions help researchers recognize their data and explain what is meant by research data. Publishers can provide examples related to the journal’s subject area and consider common research methods. For some medical journals, the publication of research conducted via clinical trials is relevant, and this should be acknowledged and specific information provided. As an example, see BMJ’s Data Guidelines https://authors.bmj.com/policies/data-sharing/.
- Exceptions to sharing
One of the common concerns about data sharing in medical and health science is the need to protect patients and service users. Journals can give guidance about how sensitive data should be handled and where a journal has a mandatory sharing policy — how exceptions can be discussed, claimed if necessary, and made transparent via the Data Availability Statement. See, for example, PLOS Medicine https://journals.plos.org/plosmedicine/s/data-availability.
- Data Repositories
Some of the key feedback that was received in the sessions with Pharma companies was that they may have specific repository requirements. For example, using controlled access repositories, or new repositories created to manage clinical study data. However, these are rarely given as examples in publisher repository guidance. Publishers can check that their recommended repositories list is relevant to the data types in the data definition, including being aware of key domain repositories — such as this NIH list or these guidelines from F1000 Research.
How the Community Can Use the Recommendations
This work has been shared at RDA conferences, and with representatives from the working groups, members of the STM Open Research Committee, and data policy experts from the community.
With the guidance now being shared publicly, publishers are encouraged to make use of it to improve the guidance that they give to authors about data sharing through journal web pages and data sharing resources – ultimately leading to more data being available to increase trust in published results.
If you’re in London next week, come learn more at the RDA conference on October 6, 2026.
Author’s note: AI-assisted editing tools were used to help revise wording and structure for clarity and house style.
