Data Access Committee Review Standards (DACReS)

Regulatory and Ethics

Date: Thursday May 13, 2021

Time: 7:00 PST/ 10:00 EST/ 15:00 BST

Meeting Chair(s): Ted Dove, Vasiliki Rahimzadeh, Jonathan Lawson

Zoom details below.

Meeting Recording

Attendees:

1. Lindsay Smith
2. J. Patrick Woolley
3. Vaso Rahimzadeh
4. Edward Dove
5. John Clarkson
6. Sylvia Gayle
7. Mahsa Shabani
8. Thijs Devriendt
9. Brendan
10. Ann Novakowski
11. Lindsay @ OBI
12. Joel Minion
13. Sue Hui Chen
14. Giselle Kerry
15. Pilar Nicolas
16. Jonathan Lawson
17. Stacey Donnelly
18. Yasuhiro Tanaka
19. Laura Rodriguez
20. Stephanie Dyke

Minutes:

Aim of today’s meeting is to go over the comments we’ve received since the draft was released. Open for comment will end on the 31st of May. Suggested edits are very welcome. After the internal work group comment ends, a smaller group of volunteers will tidy up the draft, which will then be presented to the REWS call in late June/early July. Approval there will then see the draft put online for public comment over the summer period (~2 months). Once that period closes, the small team will work on another draft to come back to REWS for a final approval.

Aim to announce at plenary that the DACReS Policy is in place.

ED: There was a comment from Adrian in the preamble about secondary research purposes. He suggested in a semantic sense it might sound less important, several suggestions for alternative language. Any thoughts for a better formulation?

PN: If we say “data for research”? Is is necessary to say secondary?

VR: The term secondary use is common in US regulatory language.

LR: Would primary come through a DAC?

SG: I do agree that having secondary is not the best term, we would recommend secondary USE.

JC: Consistent with what we do at OBI, using secondary use.

ED: Other research uses, rather than secondary?

PN: In some cases, use of repository data could be the first use. Thinking in different models of repositories.

SG: Did the hospital not collect the data from patients to give the patient some result?

PN: Patients can donate samples to the repository, it is possible to directly donate.

SG: A word like additional uses?

PN: Secondary could have legal implications in some countries.

ED: There is general feeling that secondary is not appropriate here. Something like “additional uses” or “research purposes” without using the term secondary.

PW: Subsequent uses? If data originally is collected for a particular use, might be able to say a similar thing while avoiding problematic language.

ED: I think we have a path with subsequent.

SHC: One thing we might want to consider is to add this to the list of terms we have definitions for so we have clarity. People reading this might not have context.

PW: Remove reference to sequenced individuals?

SG: Agree that maybe we shouldn’t have that reference there. All of the data will have some increased privacy risk, not just genomic.

ED: We can broaden the language.

JM: Had trouble making sense of the sentence.

ED: Agree, reword as proposed.

ED: Comment about whether evaluation of data can be used for proposed project. My own response is that it’s bit of both, not exclusive. Not drafted specific to data minimization.

JL: Fundamental question for the rest of the document. Need to make a decision on whether DACs must, should, could or should not make their judgements based on whether the science is sufficient quality or worthwhile or valid. I think it’s a very difficult road to go down if DACs are required to determine the scientific quality. Could be accused of allow projects similar to those they have denied.

AN: Agree. It takes away from the objectivity of their function.

LR: Making sure the data requested can answer the question, which is different from an assessment of quality or priority.

ED: The purpose section was drafted more in a descriptive than normative sense. Meant to recognize that DACs can have many purposes but can depend on their constitution or remit, vs. getting into the essential criteria where it’s more normative. Is it a matter of eliminating some of these phrases?

SD: Who is going to be the judge. Something I might think is high quality another person might not. How to arbitrate that.

PW: Where does oversight fit here. Would it be worth adding another section about linking to oversight?

ED: I suppose part of the oversight is with reference. It’s a subjective criteria. What this policy is seeking to do is say a DAC might serve a secondary function in assessing research but it might not. If it does, it might have a role in assessing the quality.

AN: Not a fan of “high-quality research” term.

LR: Discrepancy between what academics could think is high-quality vs patient groups.

PN: Have some criteria or indications that make the DAC be aware of the quality of the research, but it is not the duty to evaluate the quality.

ED: Tweak the language rather than remove the phrase.

SD: It is a much broader mandate. As long as there is choice I suppose it seems fine.

To keep lift lighter, do it when you put data in repositories rather than when you lift it out.

JM: This reflects some of the experience of the metaDAC in the UK, where ethics were approved so far in advance, that research not foreseen is now being proposed.

ED: Two ways to think about this: 1) Add further nuance that there might be metadacs that this mandate does not apply to, and 2) strikethrough, it won’t add much to the point of point iii.

JM: Is there any other place in doc we can capture that a DAC is a living entity and revisits things over time?

ED: Yes!

JM: I really like the text, but is it better put in section 9. Bring this section up or put the strikethrough sentence here.

ED: RoR has been a big issue of discussion. Might be something to fold into section 9, thinking about potential for returning new data to the project and RoR as an issue.

ED: Thinking about avoiding some of the problematic language, might have a one sentence recognition that DACs might also have a mandate to oversee biospecimens. Also can tweak that second sentence but not talking about rigorous standards.

ED: Transparency. Making information publicly available. This was phrased in a normative way, but Stacey’s point is that it can be an extra administrative burden. Rather than seeing it available on the DAC website but putting it on the repository?

SD: Removes it from the DAC but not from others. I think it’s a great idea but puts burden on someone.

LR: The plain language aspect does not need to be specified, but something about use and purpose of the data still needs to be here.

ED: Obligation must be placed on the applicants. Maybe make it more clear that the burden falls on the applicant.

MS: Is there any mention of aggregated data on incoming requests?

ED: Good internal data to have in terms of number of applicants.

PN: Good transparent practice, but not on the same level as publication of approved transfer of data and projects.

SD(Dyke): comment on Transparency section: This is something that came up in the context of looking at how ethics boards were reviewing projects using data from open science environments, shared via DACs. Because participant consent and data use permissions are so central to research ethics review and boards struggling to know whether data from shared resources meet standards. Would be good to add something about having public statements about upstream ethics for data access in place to facilitate ethics reviews. Big studies, biobanks do this, something DACs could be responsible for.