2018-09-30 FMG WGM

HL7 TSC FMG Meeting Minutes Location: Constellation F		Date: 2018-09-30 Time: 12:30 PM U.S. Eastern
Chair:		Note taker(s): Anne W.

Quorum = chair + 4			yes/no
Co chairs	x	David Hay	x	Lloyd McKenzie
ex-officio	x	Wayne Kubick, CTO

Members		Members		Members		Observers/Guests
x	Hans Buitendijk	x	Brian Postlethwaite	x	Paul Knapp	x	Anne W., scribe
x	Josh Mandel	x	John Moehrke		Brian Pech
x	Grahame Grieve					x	Ewout Kramer

Minutes

• Discussion Topics
  • MOTION to defer FMG approval of publication of US Core until Grahame can identify and fix the bugs
  • MOTION to approve publication request for healthcare associated infection reports, R1
  • VOTE: All in favor
    • ACTION: Anne to add approval from last week and this approval to TSC e-vote
  • NIBs for round 3 hurry-up ballot
    • Scope of ballot was changes since may, although we received much more than that. The timeline for reconciliation is two weeks from today for normative content. Originally it was for STU content as well but will add more leeway to that, although it won’t be too much later due to content deadline, but this will allow focus on normative.
    • Patient resource had a handful of trackers. Quite a few from mappings that is a rendering issue. Those shouldn’t have been negative but it can be fixed. Lloyd reports it’s already fixed – was a CSS issue. Other is agenda tracker – not part of changes from May, so is in principle outside the scope of this ballot so it can be voted not related.
    • Observation: Hans reports two or three items have larger impacts. Lloyd: There is a breaking change but it is allowable due to the current rules – can defer to R5 and fix it then. We can’t make substantive change without going back to ballot. We can make substantive change between releases provided they fit with our defined rules. This is not a change we could make for R4 but it is a change that could be deferred to R5 for consideration and made then based on our rules. Josh asks why we can make breaking changes on a normative resource and when did we make those rules? That will be under discussion – rules were made in R2. Hans: A change like that affecting cardinality could affect other things as well. Can’t make a change like that from one version to the next. Lloyd: Elements can be replicated, extensions can be defined, and we find a way to address the issue within our constraints. Could deprecate it. Hans: If the conclusion from this is not to go back to ballot, you could create an anticipating extension for the next version and have the other deprecated. Paul: Or you could go back to ballot and clean it up now. Hans: There is a value attachment and media discussion that I’m working to unravel and some other clarifications. Three big ones.
    • Infrastructure: Grahame states it passed. A lot of comments, mostly editorial comments.
    • Conformance and Terminology ballot had a number of balloters who ignored the instructions to only ballot on changes. 50 of the 150 items are things that are nonsubstantive clarifications that we can block vote. About 10 are simply not related, not possible, not appropriate. Another 7 I’ve recommended as not persuasive because they involve reworks that are technically hanging off of something that changed in the last ballot. 3 tasks/2 issues need to go to committee. One may be not persuasive. In response to a comment in the last ballot cycle, vocabulary produced a variance regarding the rules for publishing, and there is strong consensus that the invariant is wrong. It’s published purely for the code generation. It received 2 negative ballot; the only rational response is to remove it, which is a substantive change that required going back to ballot. Could come up with an extension but nobody really likes that. It’s clearly related. One other issue – while investigating that issue, we found that the infrastructure ballot includes problematic text.
    • Lloyd: The challenge around reballoting is that we can’t roll publication past December. Lynn can have the ballot site turned around and we can send out a ballot notification kicking off the 30 days on the 5th of October. We will look at NIBs shortly. OO will need to decide quickly. Reviewed NIBS for TSC review this evening. We could open registration on the 6th, closing on the 5th of November, ballot would close on the 5th of December. Ballot would be a copy of the 2018 September ballot with just these changes. Wayne: Found something in essential requirements is 02.04.04 that says that people will be registered as continuing participants for subsequent ballot of the same content. So we could do this without a NIB. 02.04.05 says that any other situation requests a variance of process from the CTO/TSC. Questioning if we need a new consensus group so we wouldn’t need the 30 days. Lloyd: There is an existing process that we know works. TSC can entertain the possibilities of whether there are less painful mechanisms to make this happen. Grahame: My preference is that this ballot should be closed to existing participants. Discussion over preference for opening and how the ballot would be prepared. It is a lot less work for Grahame than other ballots.
      • Discuss timing issue with TSC
    • Paul asks about why OO has so little time to figure out what they need to do. Need to approve the NIB tomorrow for TSC to approve going to ballot. Paul asks why it has to be published by December. ONC is a consideration, and otherwise it won’t get published until summer. Discussion over when TSC could approve. Austin states we could approve general changes rather than specific ones.
      • Anne to add NIB review to Wednesday TSC lunch agenda
    • Reviewed Terminology NIB.
      • MOTION to approve: Paul/Grahame
      • VOTE: All in favor
    • Reviewed Infrastructure NIB. If we have to go back to ballot because of terminology, may as well clean up some of the other items.
      • MOTION to approve: Paul/Josh
      • VOTE: All in favor
    • MOTION to approve NIB for Observation: Hans/Paul
    • VOTE: All in favor
  • Discouraging or prohibiting IGs from balloting in the January cycle
    • Lloyd: Leaning is to discourage on the grounds that we’re changing the tooling for how IGs work in terms of using templates, and past experience suggests that tooling changes don’t necessarily flow easily. Will also be heavily focused on QA for normative. Grahame: Do we have any sense of how many we would have? Lloyd: SDC, maybe PRO, some that were draft this cycle that might want to move to STU, US Core R4. Grahame wants US Core to go. Thinks we should say that we won’t be using templates this cycle and go ahead and work on them. Lloyd: The other challenge is the short cycle. Ballot opens on 12/7, which means you’d have to pass QA by 12/2. NIBs due by 10/28. Would have to be structurally complete before the NIB. Discussion over QA for IGs re: timing. Must submit two weeks before official deadline for Grahame to do QA on the material. Austin: TSC is prepared to back up any management group that decides to reject a product going to ballot because it fails to meet quality criteria. Lloyd: So QA deadline will be 11/18.
    • Grahame: In January we’ll be doing R5 planning, so liaisons should relate to committees that they should prepare for having solid ideas for what they want included in R5 for the product roadmap plan. Grahame would like to see more of the technical infrastructure, patient summary to go normative. R5 would be published around August/September 2020. Later next year we will survey the market and ask them for commentary on what they think our timing should be. Need to start thinking about cycling more often, working towards a 3 month basis down the road.
• Adjourned at 1:44 pm
• Carry forward
  • Ballot review
  • FMG-assigned ballot items

Minutes

Back to FHIR_Management_Group

© 2018 Health Level Seven® International