FHIR Chat · DocumentReference · implementers

Stream: implementers

Topic: DocumentReference


view this post on Zulip Lital Inghel (Oct 02 2021 at 12:40):

Hi I have a question regarding DocumentReference resource.

  1. It says "DocumentReference.content.attachment.url or DocumentReference.content.attachment.data or both SHALL be present".
    can the URL be GET [base]/DocumentReference/[id] that will bring back the document's data?

  2. what is the difference between:

  • $docref query with patient id, start date+end date, document type as parameters
    and

  • GET [base]/DocumentReference?patient=[patient]&type=[system]|[code]&period=[period]

  1. document reference-category presents one category-"clinical note", but can't we have other categories for documents that are not representing clinical notes?

thanks

view this post on Zulip John Moehrke (Oct 02 2021 at 12:55):

Yes the URL given in attachment is a GET based URL, security and privacy may apply.

view this post on Zulip John Moehrke (Oct 02 2021 at 12:56):

The IHE-MHD specification mandates the URL method as it is more flexible and more manageable.

view this post on Zulip John Moehrke (Oct 02 2021 at 12:56):

The .data method might be considered more helpful with small objects

view this post on Zulip John Moehrke (Oct 02 2021 at 13:00):

$docref is not a query, it is an operation. It is also US-Core specific, so not applicable outside the USA.
As far as I know it is exactly the same as you indicate.
I think the reason it is defined as an operation is so that the DocumentReference and objects can be "created" during the operation, where as a GET "in theory" can't create anything.

view this post on Zulip John Moehrke (Oct 02 2021 at 13:01):

You might be interested in the International standard (vs us-core) from IHE - https://profiles.ihe.net/ITI/MHD/

view this post on Zulip Eric Haas (Oct 02 2021 at 18:23):

Yes they do basically the same thing.

view this post on Zulip Michele Mottini (Oct 04 2021 at 01:31):

No, DocumentReference.content.attachment.url cannot be [base]/DocumentReference/[id] - that's the URL that brings back the DocumentReference itself, not its actual document data

view this post on Zulip Eric Haas (Oct 04 2021 at 15:13):

document reference-category presents one category-"clinical note", but can't we have other categories for documents that are not representing clinical notes?

The US Core DocumentReferences Category Value Set is a 'starter set' of categories supported for fetching and storing clinical notes.which means you can have other categories for your use case.

view this post on Zulip Nitin Lal (Oct 06 2021 at 01:58):

Hi Team, I had a quick question regarding FHIR questionnaireResponse. Our team wanted to implement an FHIR server that can handle submitting a large questionnaire response to a non-adaptive questionnaire. In the response to the questionnaire response can I return the contained questionnaire of the next section of the questionnaire even if it's not an adaptive flow

view this post on Zulip Lital Inghel (Oct 11 2021 at 13:25):

for document reference it says: 'DocumentReference.content.attachment.url or DocumentReference.content.attachment.data or both SHALL be present'.
since a patient might have many and complex docs, we'll have to use the URL.
the issue is that using this URL requires specific security permissions that not any FHIR user can have (since it access our file system). so only a specific type of user will be able to actually resolve this URL but other users will not. is this ok certification wise or we should find a way to resolve it? the user that is sending the FHIR request will not necessarily have the permission to get the URL data.

view this post on Zulip John Moehrke (Oct 11 2021 at 14:16):

Security/Privacy always can apply to all accesses.

view this post on Zulip John Moehrke (Oct 11 2021 at 14:19):

That said, the policies need to be reasonable and proper. A Binary resource can be protected just as well as anything else. Simple method is to apply the same policies that apply to the DocumentReference to the Binary pointed to by the DocumentReference.content.attachment.url. It is a simple search on the request for the Binary to find all DocumentReference that have a content.attachment.url equal to the Binary url. Once you have that, then you just apply the rules you would apply to that DocumentReference.

view this post on Zulip Michele Mottini (Oct 11 2021 at 19:12):

is this ok certification wise or we should find a way to resolve it?

Might pass certification, but it is surely NOT ok - the user being able to retried the DocumentReference but then not being able to get the referenced document itself is useless

view this post on Zulip Yunwei Wang (Oct 12 2021 at 13:28):

I asked similar question in the "SMART" stream.


Last updated: Apr 12 2022 at 19:14 UTC