2017-06-19 FGB concall

HL7 TSC FGB Meeting Minutes Location: cGoToMeeting ID: 136-494-157		Date: 2017-mm-dd Time: 4:30 PM U.S. Eastern
Facilitator:		Note taker(s): Anne W.

Quorum = Chair plus 2				yes/no
Co-Chair/CTO		Members
	Regrets	Lorraine Constable	x	Grahame Grieve		David Hay
x	Dave Shaver		Ewout Kramer		Cecil Lynch (ArB)	x	Lloyd McKenzie (FMG)
x	Wayne Kubick	x	Calvin Beebe	.
observers/guests
		x	Anne W., scribe

Agenda

• Roll call -
• Agenda Review
• Approve minutes of 2017-05-22 FGB concall
• Action Item review:
  • Grahame to write up lifecycle management for FHIR
  • Monitor governance issues that arise, noting any that are FHIR-specific
  • Provide regular governance reports to SGB
  • Review FHIR registry document when it comes forward
• Discussion
  • Governance principles for FHIR lifecycle management
• FMG update -
• Methodology update-
• Review precepts, associated governance points and risks
• Next steps: Define and Resource other Governance processes.

Minutes

• Roll call -
• Agenda Review
• Approve minutes of 2017-05-22 FGB concall
• Action Item review:
  • Grahame to write up lifecycle management for FHIR
  • Monitor governance issues that arise, noting any that are FHIR-specific
  • Provide regular governance reports to SGB
  • Review FHIR registry document when it comes forward
• Discussion
  • SGB discussed state machine alignment with FHIR last week. Wayne suggested SGB and FGB should work together to look at consistent naming. Lloyd: We should look at why the differences are and provide explanations. Wayne: We should also look at what's the gold standard moving forward with how we deal with these issues. Don't want to keep redefining common concepts; should have a clear path forward. Lloyd: That makes the presumption that ten years from now we won't come up with better insight. V3 was our version of better at the time; FHIR is our version of better now. Calvin: That does leave a challenge though for people trying to bridge the gap - it creates chasms. Lloyd: When we come up with a new specification, by definition there are going to have to be some chasms. If it was just minor tweaks, we wouldn't have a new specification. Must be sufficient difference to justify radical change. That means that conversion between the old and the new is going to at least in some areas be challenging.
  • Calvin: The problem here is C-CDA on FHIR advocates itself as using both families in one product but it doesn't span with easy mappings. Implementers are thinking it should be easy but it isn't. Wayne: If we identify and define the business concepts appropriately it will be easier. Lloyd: In V3 we had strict alignment with business concepts and there were good things and challenging things that came out of that. FHIR has been looser on that for a variety of reasons. Wayne: What proactive things can we do? Lloyd: Having an expectation for mappings means that it's not up to the community to figure it out. It also means that the WG has to be sufficiently clear in what they're doing in the new spec to correlate to the old spec to avoid misalignment or describe why misalignment is necessary. Wayne: That sounds good, as well as providing awareness of the misalignments. Where do we document these mappings within HL7? Lloyd: The mappings should live within one product spec or the other. In the C-CDA profile there are some mappings (element). Discussion over mappings at value set level - need to look at that.
  • Calvin: If we can't have abstract domain models that deal with standardized business concepts, that's going to cause problems. If we're modeling the domain concepts they ought to hold up against different product families. Lloyd: FHIR deliberately tries to be different than V3 in many places - we don't want some of them named the same. In the case of Status, the differences are driven by the fact that there is a fixed state model on one side. Precept could be that when we can align we should, and when we can't we should document it. Lloyd: Where does that should fit in the hierarchy of shoulds? If we had started with that, things would look a lot different. Happy to have it listed as something that should be thought about, but don't want alignment between standards to trump utility and appeal to implementers of that new work. Grahame: It's about the way you pose the question. The FHIR spec must have a mapping to CDA Act statuses - the question is how do you optimize real world usage. Goal is consistency but can't make guarantees. Lloyd: We need to be cautious about the language with which we frame this.
  • Calvin: When we have these value sets for common business concepts they shouldn't have to be too radically adjusted. Levels of abstraction add complexity. Calvin: We can explain that there are record statuses and clinical statuses, but what's really causing heartburn when clinical statuses don't line up. We did force the map by errata for allergies and problems because it was a patient risk not to. Grahame: The domain level functionality should be consistent and might require changes on both sides; each individual issue needs to be considered; we would be happy to receive comments around functional issues with mappings.
    • ACTION: Wordsmith this for next time and then send to SGB
  • Governance principles for FHIR lifecycle management
    • Grahame has drafted. It covers governance principles around changes with version updates.
• FMG update -
  • FMG has been focused on Connectathon tracks and a few PSSs. Will be reviewing and approving quality criteria this week.
• Adjourned at 5:05 pm Eastern

Next Steps

Actions (Include Owner, Action Item, and due date)

Next Meeting/Preliminary Agenda Items 2017-07-03 FGB concall

Back to FHIR_Governance_Board

© 2017 Health Level Seven® International. All rights reserved.