FHIR Chat · Cardinality of QuestionnaireReponse.identifier · questionnaire

Stream: questionnaire

Topic: Cardinality of QuestionnaireReponse.identifier


view this post on Zulip Morten Ernebjerg (Jun 04 2021 at 09:36):

Unlike in most other resources, the cardinality of the identifier element in QuestionnaireResponse is 0..1 rather than 0..*. The label for this element says "The Unique id for this set of answers" so this seems to be a very deliberate choice. What was the reasoning behind this restriction?
FYI @Johannes Oehm .

view this post on Zulip Lloyd McKenzie (Jun 04 2021 at 13:25):

No one's ever had a use-case for multiple identifiers on a single response. Do you have one?

view this post on Zulip Morten Ernebjerg (Jun 04 2021 at 14:10):

No, I was just curious :smile: I (maybe incorrectly) assumed that the general approach was allowing multiple identifiers and this seemed to be a deliberate divergence from that "rule". I suppose I thought a general argument for 0..* would be that a resource might get passed around to different organizations that all want to attach their own identifier (or orgs may have e.g. an old and a new identifier systems and want to attach one of each). And that seemed to apply to QuestionnaireResponse as well.

Is it rather that there's a general principle of aways picking the "narrowest" cardinality consistent with explicitly known use cases when defining a new resource & that gives 0..1 for QuestionnaireResponse.identifier?

view this post on Zulip Lloyd McKenzie (Jun 04 2021 at 14:24):

We try not to make things repeat just on the possibility "maybe someone could possibly do this" but instead are driven by actual use-cases, because adding repetitions adds costs. However, if there's a reasonable likelihood of multiple repetitions existing, then we'd boost the cardinality.

view this post on Zulip Lloyd McKenzie (Jun 04 2021 at 14:24):

Certainly if anyone is aware of a use-case for multiple business identifiers for the same response, we'd love to hear it.

view this post on Zulip Morten Ernebjerg (Jun 04 2021 at 14:38):

Ok, thanks for the explanation @Lloyd McKenzie !


Last updated: Apr 12 2022 at 19:14 UTC