Skip to end of metadata
Go to start of metadata

Attendance can be found Here

Co-Chair and Key Participant information can be found on the Agenda found Here

Discussion items

9:45Value Set Expansions guidance

Provision of practical guidance on value set expansion is needed.

This has been a joint project for some time but the way to resolve the issues have changed.  As the project approach must change the existing statement of the use case and information required should be retained, but the project will be closed.  This project may be picked up again when we get to handling binding in FHIR in R5.

Motion:  Project 1325 be closed motion by Rob McClure, Seconded by Sandra Stuart.

12 Yes   0 No,  3 Abstain 

Present 15 plus chair    Motion carries

Rob McClure to talk to Dave Hamill to get the project closed and archived and update the confluence page indicating that the project is archived but that this information may still be of value.

Clinical Practice Guidelines for FHIRRob McClure

The desire is to support consistency when implemented for clinical use.

Description profiles, differentials and snapshots.  Based on your environment you will have more than one way to include content such as snomed.  There are two different ways of describing the same content.  this supports being able to process appropriately using multiple machine processable mechanisms that will return the same value set content - that is useful but can risk interoperability.  

The exposure of the calculation method which returns the list of concepts was also considered.  We already have value sets that have more than one version but return the same membership e.g. an enumerated list, and expression based list - neither are wrong they are just two methods of achieving the same end.  We need to be able to say that a version is deprecated, but we also need to be able to say which of possibly more than one method is active.  This is a messy situation and we are seeking a practical solution.  This concept is incorporated into the definition of a value set expansion.

We need to recognize that snomed is the only code system that can handle expressions effectively so we need to have alternative methods depending upon the capacity of the code system representation. 

As a future release of UTG (post R1) we will have to deal with surfacing (in publishing of terminology) the historical releases to support ongoing implementation.  There is a possibility that there could be more than one way of expressing something.  At the moment the profile only offers one way to do X.  This does not rule out people using the compose option.

Motion:  Rob McClure moved that the CPG expression based profile  be taken to the FMG as a proposal to include that profile into core and ask for their guidance on including an expression based approach to value set definition in core.

Seconder: Ioana Singureanu

Vote:  7 Yes, 3 No, 6 Abstain

Present 16 plus chair  This passed but barely and will require further discussion

Rob McClure the CPG expression based profile  be taken to the FMG as a proposal to include that profile into core and ask for their guidance on including an expression based approach to value set definition in core.

Profile for CPGValueSet

This creates an enumeration for the value set expansion.  the CPG will hopefully be removed and could be improved to reflect what it is - rather than cpg-CachedValueSet it could be   ValueSetDefinitionProxy

cardinality of the compose should not be 1..*  - it should be 0..*  This may reflect a problem in tooling. Added as  FHIR-25930 - Getting issue details... STATUS

Motion:  Rob McClure moved that the cpg-cachedvalueset be renamed as ValueSetDefinitionProxy profile, and be taken to FMG with a request that the profile be included in core.

Seconded Jim Steele

Vote:  16 Yes,  0 No,  0 Against

16 present plus chair.  Motion passed.

analysis of conformance 

Alexander Henket Alexander  Final review of analysis of conformance will be completed by ICG and sent to Ted/Carol when complete.

  • Withdrawal or reaffirmation of CTS2 spec

held over to next quarter

  • Note FHIR (GG) interest in exploring additional FHIR binding types. Plus impact of use of expansions for a single value set definition that change over time and impact on use for data queries with data recorded over time.

Is there a binding process fully documented - additional binding types are needed to handle dynamic changes in code system content.

A different set of rules may be needed as we deal with code system changes such as ICD in a record where these concepts are now represented by SNOMED.  Identification of the data which is permitted at different points in time and how this is to be managed in V2.   This discussion looks at how to implement the theoretical concepts from core principles in a practical way.

further discussion next quarter

Action items

  • No labels