Optional field for IRB information?

[sappelhoff]

Has there been any consideration of an optional field
in the description for IRB information or is this something that goes into
the How-to-Acknowledge? Most journals require authors to provide IRB
information for any data that they use in a publication.

Originally posted by @VisLab in #372 (comment)

[effigies]

The README file is completely free-form and seems an appropriate place to put IRB information. If there's a database of IRBs that might be linked to in a semantic-web sort of way, then a field in the JSON metadata makes sense.

[yarikoptic]

IMHO I think it might be worth considering free form value for a field in dataset_description.json . Special treatment could be done by tools if value found to be a URL or match some specific id pattern

[effigies]

Well, I can find a database of IRBs from the US Dept of Health and Human Services, so a machine-parseable field seems reasonable. Also looks like terminology for ethics committees varies internationally, so a generic name might be better than IRB.

And is this meant to encode the committee itself, the committee-approved protocol or both?

[PeerHerholz]

+1 re a more generic name (based on experience from EU).

I would vote for encoding both, thus only creating one additional field. The "how" is maybe not that straight forward though. What about something like <committee>_<committee-approved_protocol_id>? The naming scheme of the committee would however be another thing as some are university-wide, some specific for an institute/department, etc. .

[satra]

this field should be allowed to be a list. many datasets comprise data collected under different IRB protocols.

[VisLab]

-1 for README

I am just going through the process of uploading a study to openNeuro. The README's are a mixed bag, and curators are unlikely to think about IRB info unless they see a field for it. This information may not be easy to find after a few years.

[effigies]

@PeerHerholz

I would vote for encoding both, thus only creating one additional field. The "how" is maybe not that straight forward though. What about something like <committee>_<committee-approved_protocol_id>? The naming scheme of the committee would however be another thing as some are university-wide, some specific for an institute/department, etc. .

I would leave it free-form, and leave room for structured data. Then if schema.org standardizes how to describe ethics committees/protocols, we'll be able to accommodate direct inclusion of those objects.

[PeerHerholz]

Totally agree! How would one incorporate @satra's point? A list of str?

[effigies]

Can we just say a list, without specifying the content?

[PeerHerholz]

Would be fine with me. I have no strong opinions wrt this and I guess we would have to gather some examples before we can see what's suitable anyway.

[VisLab]

Has any consensus been reached on this? I think it is very important to have the provenance of the data if people are going to be able to use it in analysis publications. Here is an example:

Army Human Research Protections Office (Protocol ARL-20098-10051, ARL 12-040, and ARL 12-041)

I think a string would be fine. Every institution has their own format.

[sappelhoff]

Re-reading the thread it sounds like there is an agreement about adding a new OPTIONAL field to the dataset_description.json.

To move forward we need a dedicated person to make a concrete proposal in the form of a pull request to the specification.

This concrete proposal can then be further refined, explosed to the larger community, and eventually merged into the spec.

[effigies]

My impression is that response is positive and the concerns are:

1. Allowing multiple protocols.
2. Avoiding US-centric language
3. Not over-prescribing format

So I'd propose the following addition to dataset_description.json:

| ResearchProtocols | OPTIONAL. A list of research protocols approved by relevant ethics committees. |

And add to the example:

{
  ...
  "ResearchProtocols": ["Army Human Research Protections Office (Protocol ARL-20098-10051, ARL 12-040, and ARL 12-041)"],
  ...
}

Open to an alternative if we want "Review" or "Ethics" to be more easily searchable. We can also list country-specific acronyms, such as IRB (USA), REC (UK), MREC (Netherlands), CPP (France), REB (Canada) and HREC (Australia) if that would be more helpful. (These are the ones available on Ethics comittee.)

---

Barring any objections or amendments, I can make the proposal. Or if someone has a strong alternative preference, they can open their own PR with my blessing.

[satra]

@effigies - I would separate protocols from approvals. There can be another list of "ResearchProtocols" - a la protocols.io. I think this should be made specific to "EthicsApprovals"

regarding country specific things, most people in the DOD sector use the term HRPO rather than IRB. so i would stay away from country-specific names.

[effigies]

Do we want to simultaneously add protocols then? Or separate proposals?

[VisLab]

I like the term "EthicsApprovals" Can we just add that with a recommended string as argument.

[effigies]

Proposal at #412.