With the one-year anniversary of implementation of the National Institutes of Health (NIH) Data Management and Sharing (DMS) Policy firmly in the rearview mirror, feedback that researchers are receiving from NIH staff is starting to become available. As researchers write their DMS plans, we recommend they keep these tips in mind.

This blog is condensed from a presentation given to the Federal Demonstration Partnership on December 14, 2023, as part of their DMS Pilot program. Presenters were the National Institute of Child Health and Human Development (NICHD), National Cancer Institute (NCI), National Institute of Mental Health (NIMH), and National Institute of Biomedical Imaging and Bioengineering (NIBIB).

In this blog post, we will cover some overall tips before going into specifics for each element of your DMS plan. Check out our additional guides on complying with the NIH DMS Policy and writing your plan in our Knowledge Base.

Overall Tips

• Researchers who used a template were more successful than those who did not. Here are some templates for DMS plans:
  • NIH Template​
  • DMPTool Template​
  • FDP Pilot Templates
• Watch out for contradictory information across plan elements.
• Explicitly list any data that will not be shared and explain why.
• Discuss all types of data that are included in the research plan portion of the grant.

Element 1: Data Type

• Be clear which data will be generated versus what will be shared. ​
• Include important details such as species/source, formats shared, amount, and metadata. ​
• For data not associated with publications, provide as much detail as you would for publication-associated data (repositories, data types, timelines, etc.). ​
• Use a table to organize information by data type.​
  
  
  
  
• Include an estimate of the amount of data.​

Element 2: Related Tools, Software and/or Code

• Provide names of the tools or software​ needed to access the shared data.
• Include how the tools can be accessed​: Is this tool commercially available or open-source? If it is paid, where can it be bought? Does it require a subscription? Include this information even for common tools.
• If a tool is custom-built, ensure there is an adequate plan to share the tool and software. ​
• If it’s easy to convert your files to a broadly accessible format (.xlsx to .csv), do so.

Element 3: Standards

• Maximize use of existing data and metadata standards, including formats, data dictionaries, ontologies, and models rather than developing new ones. We suggest working backwards from your chosen repository in order to find standards. Look through the repository documentation to see what standards your repository is already using.
• Describe standards you will use for each data type for the entire project.
• If no standards exist for a specific data type, indicate that in your DMS plan.

Element 4: Data Preservation, Access, and Associated Timelines

• Make sure to respond to the part of this element that requires information on how the dataset will be findable and identifiable. For this question, the NIH is looking for information on how your repository identifies datasets. Usually this is done via a DOI or an accession number.
• Make sure to include how long data will be available. Check your repository documentation for retention timelines.
• Though researchers no longer have to write separate genomic data sharing plans, the genomic data sharing policy is still in effect. Make sure to comply with all specified timelines.
• The repository you choose must be open to the public or the broader research community​. Lab websites or internal repositories are not permitted.​
• Data access and sharing cannot be controlled by the principal investigator (PI) or “by request.”​ Controlled access repositories, which provide some form of verification before data is shared, are allowed. Controlled access repositories do not rely on the PI to approve data sharing.
• Domain- or subject-specific repositories are preferred over generalist repositories.​
• Sharing only via publications and presentations is not permitted.

Element 5: Access, Distribution, and Reuse Considerations

• Explicitly state any reasons that data will not be shared​.
• Include specific information on or citations to any laws or policies cited.

There was no feedback specific to Element 6.

Budget and Budget Justification

• Your DMS budget will probably not be $0​. If nothing else, make sure to include costs for personnel who participate in DMS activities​
• Include a specific budget justification section on DMS​ activities, even if your budget is $0.
• Peer reviewers will not review the DMS plan, but they will review the budget and justification. So make sure your budget justification makes sense on its own.
• Make sure to budget for all the DMS activities you are agreeing to in your plan.

Getting Exceptions

• For any approach that deviates from what the DMS Policy “expects,” provide a justification; articulate why it's not possible or reasonable to meet specific expectations (e.g., to share certain scientific data by the end of the award, or to use a repository)​. 
• When citing laws, regulations, or policies as a reason to limit/delay sharing, cite the specific law/regulation/policy and/or invoke the associated protection​. Describe which data the limitation impacts and how it affects sharing (e.g., by causing a delay)​.
• When citing human-subject issues as a reason to limit sharing of de-identified data (e.g., due to small sample size), explain your Institutional Review Board’s role in that determination. 

If you need help writing your DMS plan, check out the other guides on the NIH policy at the FASEB DataWorks! Help Desk.

NIH Shares Feedback from Reviews of the First-Year Grant Applications

Conversation with Valerie Cotton, Deputy Director of Data Science and Sharing within the NIH’s National Institute of Child Health and Human Development, sharing learnings from the first year of implementing the NIH Data Management and Sharing Policy.