FHIR Chat · Questionnaire as StructureDefinitions? · implementers

Stream: implementers

Topic: Questionnaire as StructureDefinitions?


view this post on Zulip Diego Bosca (Mar 12 2020 at 22:05):

I haven't seen more than a single Questionnaire profile as StructureDefinition. Are Questionnaires supposed to be defined as instances of Questionnaire and not profiles? Is there any other Resource that has to be used in this way?

view this post on Zulip Lloyd McKenzie (Mar 13 2020 at 04:34):

The only time you'd profile Questionnaire is if you wanted to define constraints on a set of questionnaires. E.g. "All questionnaires we create must have labels on the questions and groups, and the Quantity and Reference data types are prohibited for answers". Individual questionnaires are represented as instances, not profiles.

view this post on Zulip Lloyd McKenzie (Mar 13 2020 at 04:34):

(And that's the same way all other resources are used - you only profile the resource if you want to define rules for a set of instances.)

view this post on Zulip Diego Bosca (Mar 13 2020 at 08:36):

I get that, but seems like same could be said for things like Observations, if you had a sibling ObservationResponse Resource where you put the answers...

view this post on Zulip Eric Haas (Mar 13 2020 at 22:33):

The proper analogy would be obsdef and obs

view this post on Zulip Diego Bosca (Mar 16 2020 at 08:16):

but people actually create profiles on observation...

view this post on Zulip Lloyd McKenzie (Mar 16 2020 at 13:21):

Sure. But you don't do that to define a single instance of Observation, you use it to define a class of Observations. You can define profiles on Questionnaire too - to define classes of Questionnaires. And you could define profiles on QuestionnaireResponse - but that would be to say things like "QuestionnaireResponse must have subject.reference, not to enforce the Questionnaire. (Otherwise, what's the point of having Questionnaire?)


Last updated: Apr 12 2022 at 19:14 UTC