Skip to end of metadata
Go to start of metadata

Meeting url: https://join.freeconferencecall.com/vocab

Date

Time

3:30PM US EDT (UTC-4)

Attendees



Planned Agenda Topics

  1. Roll call and Agenda check (Ted)
    No changes.

  2. Announcements (all)
    Harmonization completed yesterday.   All except one proposal were approved unanimously, the other one was withdrawn.

  3. Action Items Check  (Ted)
    No updates.

  4. Atlanta WGM Agenda Planning (Carol Macumber)


  5. Value Set Canonical URL issue (Rob McClure)
    Rob set the following to the list and we need to get these issues resolved ASAP:Vocab,
    On today’s Structured Docs call Rick ran through the updated C-CDA on FHIR STU update to be published and the few value sets included are all in VSAC, except one that also includes DICOM - DIRSectionTypeCodes - but this one is actually simply additional DICOM codes added to the one that is in VSAC - DIRSectionTypeCodes LOINC. The issue: the listed canonical url in the FHIR spec for all these value sets in the about to be published updated spec is different than the canonical url provided in the FHIR resource that VSAC publishes and VSAC is considered the source of truth (SOT)

    Issues to discuss:

    1. The problem of "multiple canonical urls for a value set that we actually want to be the same thing" that has arisen is not unique to situations where VSAC is in the mix. We should have a policy around this,
    2. But what is a bit unique here is that this problem can arise when we use a terminology service that is not under our control as a source of truth. 
      1. We certainly can tell CCDA on FHIR to to use what ever the SOT indicates as canonical url no mater what it is if that is our approach. 
      2. Or perhaps we might ask that the SOT structure value sets they host for us - as is occurring with VSAC CCDA value set - to use a particular url. I don’t think VSAC can do that but it could be an expectation we state. If we did this we might want situations like this to use a UTG-style canonical url.
    3. We also must consider the impact of making changes in canonical urls and how that should be managed. 

    Vocabulary WG Assertaions:
    A.  There is only one single "canonical URL" for a value set (definition).   A different canonical URL is thus identifying a different value set.   Note that there may be additional URLs for a value set, but these are not known as the "canonical URL" and flagged as the "preferred" URL for the value set.
    B.  The multiple identifiers of different types for a particular value set will be listed in NamingSystem resources; for each type of identifier, only one will be indicated "preferred".  For type URL, the one marked preferred can be considered the proper one to use as the canonical URL.
    C.  When there are multiple identifiers for a value set, if there is more than one of a particular type, then the way that we will indicate which one of those of a particular type will be indicated with our use of NamingSystem.
    D.  Code Systems also may have multiple identifiers, and some of those are considered "preferred".    The preferred code system identifier of type uri will be considered the canonical URL for the code system.
    E.  For the 'external' code systems (published by organizations other than HL7, such as X12), we must develop advisory guidance (probably in Atlanta at the joint HTA/Vocab sessions)
    F.   In the meantime, what has been done already can be found at https://standard.x12.org/Home/Default/005020.   (note that this will require a login to view)
    G.  We must develop a strategy for the URLs for the base and for the versioning and make it available to X12 (and others) as guidance recommendations from HL7.
    H.  This must culminate in what must be put into the IG for the value set, since IP restrictions prevent the content of the value set being published in the IG.
    I.  To be continued...



  6. UTG Update (Ted)
    Resolved all the initial issues with the imported content.
    We have changed the architecture of proposal preparation and approval, now will be using BitBucket.  We will use browser pages with rendered content for proposals in flight.

  7. Harmonization Item (Ted)
    On yesterday's harmonization call meeting, the proposal from SD for MediaType was withdrawn by Austin (Rob M was traveling and unable to present it).  There may have been a variety of concussions and misunderstandings related to it, and we need to determine what should actually be done here for this.  In summary, the V3 code system MediaType is an "Internal/External" code system, which in V3 means that the external (non-HL7) source at the time it was needed had neither an identifier for the code system that was usable in the HL7 published data types available, and/or the content was buried or in a format that was not readily accessible to the HL7 community.   The subset of codes that were deemed critical for HL7 users was created as an "EI" HL7 code system with only these codes.  For this type of code system, we havre hitherto NOT added content to the code system if it is not in the external source; the copy in the HL7 artifact is merely a "convenience copy" of a subset of the external content.    This policy seems to haver caused some issues which came to a head yesterday and we should discuss on todays call.

  8. Glossary Project (Heather)
    No one here to speak to this project, item tabled.

  9. FHIR items (Rob Hausam)
    Tabled, out of time.

  10. Concept Map Task Force Report (Rob Hausam and Carmela)
    Tabled, out of time.

  11. Gender Harmony Project Update (Rob McClure) 
    Tabled, out of time.

  12. Vocab Facilitator Review (Heather Grain)
    No one here to speak to this project, item tabled.

  13. Liaison Reports
    a. ISO (Ted)

    Next TC215 meeting in Daegu, Korea November 4-8.  Ted was just at the annual TC249 Plenary in Bangkok; next one not yet finalized. Rob Hausam will also attend to cover for Ed and Wayne.

    b. LOINC (Ted/Rob H)
    Next meeting combined Clinical and Lab combined LOINC to be held in Barcelona October 22-24.  New LOINC release done a few weeks ago, version 2.66, download available.

    c. SNOMED (Rob H)

    July 31 there will be a new release, along with an update to the free subsets.

  14. New Business -


  15. Next call items
    Grahame has requested that SNOMED has major issues with HL7 and SNOMED with a face to face meeting.    Link to Zulip: https://chat.fhir.org/#narrow/stream/179244-snomed/topic/Possible.20meeting.20Amsterdam.20Nov.2019
    Rob Hausam will be working on FHIR Terminology Services implementation.


Call adjourned at 5:03PM EDT UTC-4. The next main Vocabulary WG call scheduled for August 1, Ted will not be available for this call, we will need chairs to run and attend it.  

Action items

  • Ted Kleinreach out to Rob Snelick and Frank Oemig for topics for Atlanta Thursday Q1
  • Ted Klein will send a message to the chairs list to organize the next call.