Consent for participation and use of data in the clinical research

[JohnMoehrke]

You should have a definition of how the Consent resource would be used to capture the act of the patient giving consent. This consent both for their active participation (if appropriate), as well as consent for use of the data for that purpose of the clinical research project.

• Consent for active participation is not well developed in FHIR at this time. The committee is looking to help someone willing to work with them to create the first examples. From these examples improvements in the elements in the Consent resource could be done.
• Consent for data use (privacy) is more well developed in FHIR at this time. The critical aspects is that you would have a clinical-research PurposeOfUse, and a named project specifically. There would likely be a period of time that the data would be gathered, a period of time when the data would be used. a filter criteria of what kind of data is relevant to the project (or exclusion of not relevant so as to avoid unnecessary sensitive topics).

[mattkoch614]

@lengfelj @cjcuster @daihugh Is Consent something important to include with respect to our data exchange scenarios?

[cjcuster]

Sorry in advance if I misunderstand how it would be used...Do the sponsors need to receive consent info from the site? I would expect the site to do the consenting, and the required filtering, internally. Does the sponsor then need to consume that info on demand, in a resource? If so, I can see it being useful for the sponsor.

[JohnMoehrke]

I was thinking that the researching organization needed to get consent (both participation consent, and privacy consent). This Consent documentation would need to be maintained as proof for the life of the clinical trial, and data retention requirements of the clinical trial. They would present that consent to the clinical practice as proof that they are authorized to receive the data. The interoperability impact is this 'presenting of the consent as proof of authorization'. This is why the elements of the Consent resource are useful. Also, the Consent resource includes access rights rules (Consent.provisions) that can make clear what data the patient did authorize access to.

Note there is a new change request to add a standard method for a requesting organization to indicate the location of the Consent that that requesting organization believes authorizes the request. This has not yet been put into FHIR, so it will be in R5. But the method can be used with any version of FHIR, as the method will be an http X header. see 19312 https://gforge.hl7.org/gf/project/fhir/tracker/?action=TrackerItemEdit&tracker_item_id=19312&start=0

This seems most clean to me. I do recognize that other models are possible. You can certainly choose a different model, and not include this one. The model must be clear to the reader of the IG. Even if no model is chosen, this must be clear.

[mattkoch614]

I'm going to close this for now because Consent, at least at this time, is outside of the scope for this IG. It might be added later, and we can take these points into consideration.