Best Practices:

• Choose a third-party data archive that is appropriate for the data being measured. Things to think about are data sharing expectations from the funder, whether similar data are already available in the archive, the funding stability of the archive, and whether the data will have appropriate security and access control.
• Choose a third-party data archive that makes the data as widely available as possible. If Jordan’s data were less sensitive, a repository like OpenNeuro would make the data more easily available.

Things to Avoid:

• Hosting data on your own website

Value Set Definitions: 

• Yes
• No

Value of Use Case Example:

Yes - Jordan has already decided to use the NIMH Data Archive as her third-party repository.

Discussion of Use Case:

 Jordan has chosen to use the NIMH Data Archive because she wants to analyze her newly acquired data along with data already in NDA. Sharing data in NDA will also bring her into compliance with the data sharing expectations should this research be funded by NIMH (NOT-MH-19-033). Jordan has also noted that NIMH expects a certain set of clinical measures to be collected for all of their funded studies (NOT-MH-20-067), so she is planning on adding those measures when she collects data.

One additional factor in Jordan’s decision to use NDA is the “study” feature in that repository.  NDA creates “collections” of data that are related to a single research program.  Generally, collections are associated with a single grant award, either from NIMH, one of the other NIH Institutes and Centers using the repository, or from non-federal funding groups. The study feature will allow Jordan to create a data set containing the exact data set used in a publication. The data in a study can all come from her collection or can come from more than one NDA collection. Studies allow others to run different data analysis pipelines on the exact data Jordan has already published on.  

All of the data in NDA is stored in a commercial cloud provider with appropriate security. Data access is restricted qualified investigators as determined by a Data Access Committee (DAC) convened by NIMH.  The DAC also checks whether users requesting access are planning on using the data in a way that is consistent with the informed consent.  In addition to storing the data, NDA provides limited cloud computational credits. That feature of the archive should be useful to a user like Jordan.