UK Core Clinical and Technical Assurance Sprint 8 Documentation Pack

Issue 22

Context of Issue
Context:	UKCore-Observation-TobaccoConsumption (Profile), UKCore-TobaccoConsumption (ValueSet)
Type:	Modelling
Source:	UK Core Development Team
Scope:	Proposed In Scope

Issue	Proposal
tobaccoConsuption profile only allows valueQuantity, but SNOMED CT UK has codes such as 160603005 \| Light cigarette smoker (1-9 cigs/day) (finding)\|	Initial Proposal: Propose loosening up to allow valueCodeableConcepts, with a valueset to include all types of tobacco use. Main Points Raised in C&TA Calls: Ann Wrightson: Revisit purpose and usage contexts (what needs comparability?); minimise UX burden. Sian Musto: If quantities are captured, categories can be derived—understand why categorical terms are needed vs raw measures. Andrew Perry: Origin may be PRSB; will analyse GP usage stats for pre coordinated smoking concepts; note QoF complexity and that status ≠ single recording. Ian McNicoll: Clinical priority is smoking status (current/ex/never) as the key risk factor; allowing coded consumption is fine, but ensure status is addressed (e.g., separate profile/element). Kanthan Theivendran: Points to openEHR Tobacco Smoking Summary as a good reference for clinical data elements; AU FHIR uses archetypes to guide shapes. Charlie McCay: Primary fix now: correct the value set description (it wrongly says “level of consciousness”). Don’t over generalise—keep this profile focused on consumption; create separate value sets/profiles for smoking status and other perspectives; document intended use and misuse explicitly. Kevin Sprague: Agreed: update description; consider additional artefacts but avoid burdening implementers.

Issue

Proposal

tobaccoConsuption profile only allows valueQuantity, but SNOMED CT UK has codes such as 160603005 | Light cigarette smoker (1-9 cigs/day) (finding)|

Initial Proposal:

Propose loosening up to allow valueCodeableConcepts, with a valueset to include all types of tobacco use.

Main Points Raised in C&TA Calls:

Ann Wrightson:

Revisit purpose and usage contexts (what needs comparability?); minimise UX burden.

Sian Musto:

If quantities are captured, categories can be derived—understand why categorical terms are needed vs raw measures.

Andrew Perry:

Origin may be PRSB; will analyse GP usage stats for pre coordinated smoking concepts; note QoF complexity and that status ≠ single recording.

Ian McNicoll:

Clinical priority is smoking status (current/ex/never) as the key risk factor; allowing coded consumption is fine, but ensure status is addressed (e.g., separate profile/element).

Kanthan Theivendran:

Points to openEHR Tobacco Smoking Summary as a good reference for clinical data elements; AU FHIR uses archetypes to guide shapes.

Charlie McCay:

Primary fix now: correct the value set description (it wrongly says “level of consciousness”).
Don’t over generalise—keep this profile focused on consumption; create separate value sets/profiles for smoking status and other perspectives; document intended use and misuse explicitly.

Kevin Sprague:

Agreed: update description; consider additional artefacts but avoid burdening implementers.