Details
-
Change Request
-
Resolution: Persuasive
-
Medium
-
Using CQL With FHIR (FHIR)
-
1.0.0-ballot
-
Clinical Decision Support
-
ELM Library
-
8.11.1
-
-
Chris Moesel / Juliet Rubini : 23-0-0
-
Clarification
-
Non-substantive
Description
I expect that an ELM Library instance must provide an elmXmlContent or elmJsonContent (or both). Since both these slices are optional, however, a library with neither would still conform to the profile. Consider adding an invariant that requires one or the other (or both) to exist.
To make it easier for implementers, I suggest defining the invariant in terms of the contentType rather than slice membership. E.g., something like content.where(contentType = 'application/elm+xml' or contentType = 'application/elm+json').exists() (note: I haven't tested that fhirpath, but hopefully it's close enough to illustrate the idea).
There is also an expectation that within each content, the data is either serialized or pointed to via a url, but this is not formally enforced. Consider adding invariants to each slice to require that data or url exists.
Attachments
Issue Links
- is voted on by
-
BALLOT-58376 Affirmative - Paul Denning : 2024-Jan-FHIR IG CQL E1 STU
- Balloted
-
BALLOT-60948 Affirmative - Mark Kramer : 2024-Jan-FHIR IG CQL E1 STU
- Balloted
-
BALLOT-60985 Affirmative - Yunwei Wang : 2024-Jan-FHIR IG CQL E1 STU
- Balloted