2017-10-17 - SNOMED on FHIR Planning Meeting

Welcome, introductions and apologies

@Former user (Deleted) and @Dion McMurtrie

•  General introductions and welcome

Review existing and planned work

@Former user (Deleted) and @Dion McMurtrie

• Members to report on existing and planned work with using SNOMED in HL7 FHIR

• HL7 International and SNOMED International to report on existing and planned work with using SNOMED in HL7 FHIR

Collaborative deliverables and plans

@Former user (Deleted) and @Dion McMurtrie

• Discuss the proposed scope of collaborative deliverables, for example:

  • What format will the deliverables take?

  • Are we providing general guidance for SNOMED CT use, or specific guidance for each resource (or both)?

• Consider what existing or planned work can be leveraged

• Who will be doing (and coordinating) the collaborative work?

• Collaborative platforms for sharing the work as it progresses

• Communication and meeting requirements

Implementation Guidance Principles

@Former user (Deleted) and @Dion McMurtrie

• Discuss the principles we should follow in developing the collaborative deliverables.

• For example:

  • Is the goal to provide homogeneous population of the resources for interoperability or permissive guidance to cater for a greater degree of localisation?

  • Do we restrict value sets (in a computable way) to ensure that the same information cannot be represented in 2 different ways - for example:

    • The same semantics can/can't be represented both precoordinated and in separate fields

    • The same semantics can/can't be represented in more than one resource (e.g. Condition, Observation, Family History, Allergy Intolerance)

  • How widely do we require/recommend SNOMED CT to be used across the data elements within each resource, versus permitting other code systems to be used? E.g. vital signs, statuses

  • Which are the following are potentially in scope?

    • Restricting existing SNOMED CT value sets already bound to a FHIR resource to subhierarchies that are more semantically appropriate to the relevant data element

    • Defining new SNOMED CT value sets for data elements that do not currently use SNOMED CT

      • Unless they have a required binding to another code system?

    • Restricting the cardinality of data elements (e.g. body site) to reduce ambiguity

    • Splicing data elements to more closely align with the SNOMED CT concept model (only where necessary to avoid ambiguity?)

    • Providing mappings between FHIR value sets (e.g. statuses) and SNOMED CT codes (including context)

    • Providing SNOMED CT expression templates to allow implementations to transform between different ways of populating the same semantics in a FHIR resource

Progress work

@Former user (Deleted) and @Dion McMurtrie

• Progress specific work related to SNOMED CT use in FHIR artifacts, for example the SNOMED CT Implementation Guide for the Observation Resource

Next steps

@Former user (Deleted) and @Dion McMurtrie

• Discuss next steps, including future meetings and work items