This wiki has undergone a migration to Confluence found Here

2018-05-18 SGB WGM Agenda

From HL7Wiki
Revision as of 18:31, 29 May 2018 by Annewiz (talk | contribs) (→‎Minutes)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

back to Standards Governance Board main page

HL7 SGB Minutes

Location: Salon 19, Wurzburg

Date: 2017-05-18
Time: 9:00 AM
Facilitator Paul/Lorraine Note taker(s) Anne
Attendee Name

x Calvin Beebe
x Lorraine Constable
Russ Hamm
Regrets Tony Julian
x Paul Knapp
Ewout Kramer
Regrets Austin Kreisler
x Wayne Kubick
x Thom Kuhn
Ken McCaslin
x Rik Smithies
x Sandy Stuart
x Guest: Grahame Grieve, Ted Klein, Kathleen Connor, Galen Mulrooney
Quorum: Chair + 4


  • Agenda review
  • Action Items
  • Discussion Topics
    • Vocabulary Issues/Precepts
    • Governance over OIDs and URIs
      • From Ted: Some HL7 members have mis-assigned some OIDs. We (HL7) have to make some decisions over level of governance of the identifiers used in our standards, such as OIDs and the canonical URIs in FHIR, and it seems this needs to be put on the agenda for the Cologne SGB meeting.
    • Large architectures coming in
    • Short term/long term planning for September/January
    • SGB Communication Plan
    • Normative balloting/continuous maintenance in FHIR


  • Agenda review
  • Action Items
  • Discussion Topics
    • FHIR note to cochairs regarding balloting
      • Reviewed Lloyd’s draft to guidance for committers on normative content. Grahame notes that balloters submit gforge comments and then link to the wrong content, so then it has to be marked as not related. Grahame states that Karen said that you can’t redirect to the appropriate content because it’s Normative. Grahame notes that committees are instructed not to ignore the issue but create a new tracker. Some people voted on entirely the wrong ballot.
        • ACTION: Check with Karen regarding why you can’t redirect – is it an ANSI or ER guideline
      • Discussion over substantive changes/timeline. Need to be clear on non-substantive. WGs should have examples. Paul: Question of substantive is not if you’ve implemented, it’s whether people have had the opportunity to review and comment. Calvin: It is an interesting question we could take up with ANSI. Can we make changes based on methodology that we’re operating with no? Need to have a clear understanding of each situation.
      • Number 2 on the list needs clarity. Added explanatory note from SGB
    • Normative balloting/continuous maintenance in FHIR
      • Once R4 is out, FHIR will start working on R5 after a break, although the community may request updates sooner. We expect that FHIR will keep rolling. Continuous maintenance means you’re always accepting comment and publishing on a predictable schedule. Grahame thinks we should move towards minimizing the importance of the ballot, with more focus on the continuous maintenance side of the process. Don’t want people to wait for ballot before they comment. Thom: It would move to something more like a change proposal process. Paul: If you go to continuous maintenance, everything is open for comment. Wayne: The thing to figure out is how you handle balloting. Calvin: We need to figure out what it means to do continuous maintenance in a normative state. Wayne; We need to propose how we want to handle maintenance for FHIR once it’s normative and then run it by ANSI. We are grappling with the same thing regarding UTG.
        • ACTION: SGB to look at continuous maintenance for Normative standards. Invite Karen.
    • Vocabulary Issues/Precepts
      • Ted: We’re looking at unifying vocabulary across standards. OO has completed the process with one of their projects. CIMI has been refusing to compromise on coming up with a unified set. FHIR has some issues as well. There needs to be a mechanism to get people to unify their vocabulary. Calvin: How do we decide what vocabulary is the base? We need something that’s abstractly metric that measures all these things. For new content we make sure that if two new things come in, they’re harmonized/unified. But the older the stuff is, the more intractable the issue is. CIMI issues are going to have to be escalated to the TSC. Ballot-bound stuff is another issue. Have been working with Conformance on how you constrain value set content. C-CDA has introduced more complexity. FHIR was folded into harmonization a year ago with non ballot-bound content. Problems are rising because people are recognizing the problems because standards are mixing. Need to unify the governance across the product lines. Trying to drum up funding to have people specifically start working on the problem. Ted is trying to put into place some processes to avoid more instances of this problem. For Vocabulary that is bound required is normative. Lorraine suggests looking at normative required bindings to start. Paul: The role of Vocabulary within the organization is to stop us from hurting ourselves, but there is a lack of empowerment. Ted: Governance issue is that there is duplicative work. Paul: We need to make a decision if we’re going to fix it now with FHIR or later with a new standard. The stuff in FHIR STU has to get cleaned up before it goes Normative.
    • Governance over OIDs and URIs
      • From Ted: Some HL7 members have mis-assigned some OIDs. We (HL7) have to make some decisions over level of governance of the identifiers used in our standards, such as OIDs and the canonical URIs in FHIR, and it seems this needs to be put on the agenda for the Cologne SGB meeting.
      • Ted recommends that every place people see code systems, there should be an identifier. Discussion over how this manifests across the standards. MK: Where do I go to for the ultimate source for the code set of a data element? Ted: That’s a difficult question because we allow binding to mean different things for different purposes. The binding in an IG tells you the value set and gives you contextual information. The value set tells you what the code system is. Every vocabulary item in FHIR will have an identifier locator to the authoritative source. In the design of UTG, there will be a documented pointer that says “for this item, here is the source of truth location.” Calvin: We need to work toward social engineering, perhaps. Ted: We’re looking at adding new metrics to incent people to move towards the process. Calvin: We need to reward and incentivize compliance with this. TSC should resolve this so people move towards one thing. We need to be creative and consider positive reward. Need to address identification, cleanup, and avoidance.
        • ACTION: Ted will be decreasing his support in about 15 months – need TSC to consider vocabulary tooling and succession planning. Need to look at near term succession planning across strategic WGs.
      • Do we require that people clean up their identifier errors? It’s a technical errata.
        • ACTION: Create precept regarding addressing identifier errors
      • What is the best way to deal with people using OIDs incorrectly? We have no governance over how people use them. People use them inconsistently and also use the wrong ones. People are assigning unregistered OIDs, not standing up access points for them, etc. Complex issue; not easily fixed by a step in the review process. Problems were fewer when OIDs were free. People need to understand that they can hurt the community by not managing the identifiers correctly. Lorraine suggests having a guidance document with best practices.
    • Issue from PH regarding spreadsheet reconciliation
      • PH took items out of tracker, reconciled in spreadsheet, and didn’t put them back into the tracker. One question is when they applied the changes from the reconciliation. In the event that they didn’t apply the changes before it went back to ballot, that is an issue. If they simply forgot to add things back into tracker, it’s just an administrative issue.
        • ACTION: Paul to ask PH for clarification on the status and ask for a plan for addressing the issue
    • SGB Communication Plan
      • Where do we put our materials? How do we make it clear to people what they should do with our material? Drafted initial plan.
    • Large architectures coming in
      • Carry forward
    • Short term/long term planning for September/January
      • Communication plan – now
      • Continuous maintenance – investigate between now and Baltimore
    • RDAM
      • Carry forward

Meeting Outcomes

Next Meeting/Preliminary Agenda Items

© 2018 Health Level Seven® International. All rights reserved