NIH Data Management & Sharing Policy 2023

The National Institutes of Health (NIH) instituted a new Data Management & Sharing Policy (DMSP) on January 25, 2023. This library guide provides guidance and resources for University of Minnesota stakeholders.

Overview

The National Institutes of Health (NIH) implemented a new data management and sharing policy (DMSP) on January 25, 2023. High-level aspects of the policy are below:

  • Requires submission of a Data Management & Sharing (DMS) Plan with the grant application, and compliance with that Plan as it was approved by NIH program staff (plans are reviewed, not scored).
  • Applies to new applications and renewals submitted on or after January 25, 2023 (does not apply to any applications prior to that date).
  • Applies to all research funded whole or in part by NIH that generates scientific data, including clinical data.
  • Specific offices and institutes (e.g., National Institute of Mental Health) may have additional DMS Plan requirements.
  • Use Form H and submit DMS Plan in section "Other Plans."
  • DMS Plans should be 2 pages long.
  • DMS Plans can be updated throughout the lifecycle of the grant.

Resources

Get help

  • UMN Research Data Services, a partnership between the Libraries and CLA's LATIS, offers education, training, and consultation on data management and data sharing across all disciplines. See contact information in the left pane.
  • For DMS Plan review, please send a google document draft to data@umn.edu, and we will review it within three days. We may request to meet with you to discuss the initial review, and will need two to three days to schedule it.

Find a repository

Steps for finding a data repository 

  1. Check your Notice of Funding Opportunity or with your NIH Institute, Center, or Office (ICO) to determine whether either encourages the deposit of data into a specific data repository. 
  2. Explore the list of NIH-Supported Data Sharing Resources to see whether a repository listed under your ICO, subject area, or model system is a good fit for your data. 
  3. Share in a general data repository or DRUM.

UMN data repository memberships

Data Repository for the University of Minnesota (DRUM)

  • DRUM is a free public-access repository with no associated curation or deposit fees.
  • DRUM meets many of the NIH and OSTP recommended features of data repositories. 
  • However, DRUM is not suitable for all kinds of data. See limitations below: 
    • Human Participant Data Policy: data should be non-sensitive, de-identified, and have clear participant consent for open-access sharing. 
    • Data Collection Policy: data should be owned by the depositor or depositors must have clear rights to share, and documentation should be sufficient to understand the data. 
    • DRUM is not a good fit for large datasets (individual files over 3GB and submissions over 30GB total), which are difficult to both submit and access through DRUM. 

Repository comparison

 
Features DRUM ICPSR OpenICPSR

Harvard Dataverse

 Dryad

Open Science Framework
Public-access sharing yes no yes yes yes yes
Offers controlled/ restricted access no  yes yes yes no yes
Who controls access requests n/a repository repository depositor n/a depositor
Allows custom terms of use no no no yes no no
Fee for data deposit no no for UMN no no  no for UMN no
Fee for data access no for non-members no no no no
Allows blind peer review no no no no yes yes

Generalist Repository Comparison Chart maintained by NIH

Budget for data sharing

NIH DMSP Budgeting Resource prepared by UMN Research Data Services and Sponsored Projects Administration.

Forecasting Costs for Preserving, Archiving, and Promoting Access to Biomedical Data from National Academies of Science, Engineering, and Medicine.

COGR Review of the Final NIH Policy for Data Management and Sharing: Budgeting and Costing from Council on Governmental Relations' NIH Data Management and Sharing Readiness Guide.

Human participant considerations

De-identification

When collecting data from and with human participants and communities, Element 5.C. of the DMS Plan requires that you describe protection of participants including de-identification of the data. Be explicit in the process you will use to address direct and indirect identifiers.

  • Direct identifiers should be completely removed from data. This includes the 18 identifiers described in the HIPAA Safe Harbor Method and any other information that directly ties to an individual.
  • Indirect identifiers require close examination for variables - that when combined with other variables, datasets, or publicly available information - could re-identify participants.
  • The process of data curation usually involves some level of inspection for direct and indirect identifiers. Not all data repositories have data curators, and if they do, curation may not be a free service.
  • Note that a de-identified dataset is not anonymized. When crafting DMS Plans, IRB applications, and participant agreements, using language such as "de-identified" or "confidential" is preferred.

Resources

Informed consent language

When collecting data from and with human participants and communities, Element 5.A. of the DMS Plan requires that you describe how informed consent will be obtained for data sharing, and if there will be any access restrictions to the data related to consent. Be explicit in the language you will use in the consent form given that the Certificate of Confidentiality (issued to all NIH awardees) requires explicit consent for data sharing. Include:

  • How the data will be processed before it is shared, including de-identification methods
  • What data will be shared (and what data will not be shared)
  • Where the data will be shared (name the specific repository)
  • How the data will be accessed (publicly available or restricted to specific requesters)
  • Who will grant access (the repository or the PI)

Resources

Templates and examples

NIH template and sample plans

NIH DMS Plan template in DMPTool 

  • Log in to DMPTool using your UMN credentials to see DRUM-specific language and submit your DMS Plan to Research Data Services for review.

NIH institutes, centers, and offices may provide more specific guidance 

UMN examples (forthcoming)

DRUM boilerplate language

  • Review DRUM policies to ensure your data can be shared in the institutional repository.
  • This language may be copy/pasted into Element 4.A. of the DMS Plan and/or you can split the text below into Elements 4.A., 4.B., 4.C., and 5.B.

"The data will be shared via the Data Repository for the University of Minnesota (DRUM), an open access, publicly-accessible, institutional repository. DRUM has been certified since 2017 by CoreTrustSeal, an international community-based organization that recognizes sustainable and trustworthy repositories. Curators review submissions and work with data authors to comply with data sharing requirements in ways that make data findable, accessible, interoperable, and reusable (FAIR) - including, but not limited to, file transformation and metadata augmentation (Dublin Core is the metadata standard). DRUM commits to 10 years of long-term preservation using services such as file migration (limited format types), off-site backup, bit-level checksums, and Digital Object Identifiers (DOI) for archival citations. The DOI exposes data to online discovery tools like Google Scholar and Web of Science Data Citation Index."

Open Science Framework boilerplate language

  • Note: Be sure to use the template without hyperlinks for your DMS Plan, as links are not allowed.

General DMP Checklist

  • Note: this is a general data management plan checklist useful for NSF, DOE, and other national agencies. It is not specific to NIH's DMS Plan requirements.

Frequently asked questions

Click on a topic below to explore specific sections of Research Data Services' FAQ resource.

Last Updated: May 19, 2023 1:46 PM