Add guidance on Observation.category to child health and pregnancy profiles

[RichardTON]

Observation.category can be used to support finding observations. Vital signs profiles provide such guidance. Something similar would be useful for child health and pregnancy observations.

Item 1 Add guidance on use of Observation.category to child health and pregnancy profiles of Observation in HL7 AU.

Only limited guidance can be provided as the values and use of Observation.category is contended in Zulip and by the OOWG

Relevant profiles are:

• AU Estimated Date of Delivery - this may be determined by an examination, by imaging, or by a calculation from an interview. This complexity is not yet ready for guidance.
• AU Gravidity - this may be determined by an interview, there is no suitable code in Observation Category Codes for this. See comment below.
• AU Last Menstrual Period - ditto
• AU Parity - ditto
• AU Smoking Status - social-history

~~Item 2 It might be worth having a new category code: obstetric-history~~ It will take much time to get agreement on any such category. Removed from this issue.

[RichardTON]

Make having the category required. The categories should be optional.

[dtr-agency]

@RichardTON request to add to sex assigned as birth and assertion of no relevant finding.

Don't necessarily agree that category it to be mandatory. I think the initial design would be helpful to have a patternCodeableConcept - that would be clearest to most people and is how we've done the majority of rules on must have this thing then can have others.

Item 2 is going to mean this ticket never gets closed - the one chosen is contentious, why put history, where is the definition, how would this be used, what use case is it serving.

[RichardTON]

Strategies

1. Slicing category by value:coding.system and coding.code. This prevents any instance of category that matches a slice from having alternate codings.
2. Using an invariant to test for an instance of category with the expected value. This allows alternate codings, but it will throw an error or warning and is not much use as guidance.
3. Slicing category by pattern:$this and using category.patternCodeableConcept to define the slice's discriminator values.

Only option 3 allows alternate codings and allows optional categories.

[RichardTON]

The use and preferred values of Observation.categories in Core are contested in Zulip. It is unclear what code from Observation Category Codes is suitable for observations that are based on an interview. The name of survey suggests that it is OK for a questionnaire, but its examples suggest otherwise. But it does not seem suited to an interview. Even exam seems unsuited.

[RichardTON]

Add guidance that the category social-history is relevant to smoking-status. Add no other guidance.

[RichardTON]

Category for

• sex assigned at birth - exam is a reasonable fit
• assertion of no relevant finding - no good fit in Observation Category Codes

[dtr-agency]

@RichardTON - this appears to have stalled. Please progress or close.

[RichardTON]

Progress on the meaning of survey - https://chat.fhir.org/#narrow/stream/179256-Orders-and-Observation-WG/topic/Use.20of.20Observation.2Ecategory/near/277834955

[brettesler-ext]

F2F comment - content was transferred to AU Core - as specific clinical content - propose create new ticket in AU core Jira and close this ticket

[hl7auadmin]

eVote: this is now being completely reviewed and actioned by the AU Core TDG business; Propose close.

[brettesler-ext]

Closed as no longer relevant https://confluence.hl7.org/display/HAFWG/2024-04-10+FHIRWG+Meeting+Minutes