Details
-
Change Request
-
Resolution: Persuasive with Modification
-
Medium
-
Structured Data Capture (SDC) (FHIR)
-
current
-
FHIR Infrastructure
-
Extractable Questionnaire - Definition
-
-
Brian Postlethwaite/Paul Lynch: 4-0-0
-
Clarification
-
Non-substantive
Description
See discussion on zulip: https://chat.fhir.org/#narrow/stream/179255-questionnaire/topic/Json.20Structure.20for.20Repeating.20Complex.20Data.20Type.20in.20resource
While extracting complex data types that are repeats (e.g. HumanName under a Patient), it is difficult for the extraction tool to interpret the answer to a questionnaire if the questionnaire does not explicitly have groups set up for repeated complex types. For example, if there're two fields respectively mapped to a given name and a family name, it is unclear to the extraction tool if they should be nested under the same HumanName or two separate HumanNames.
This is an issue because the extraction spec doesn't really specify what the questionnaire should look like in cases like this. Here we propose making the use of groups in such cases mandatory (so if you want to have one human name which has a given name and a family name, define a group for the human name and nest the two questions under it. and if you want to have two separate human names one with only a given name, another with only a family name, define two groups for the two human names).
The tightening of the spec here would allow unambiguous extraction of fhir resources from questionnaire responses.