20080214 arb minutes

Attendees

• Jane Curry
• Grahame Grieve
• Abdul Malik Shakir
• John Quinn
• Yangjian Bao
• John Koisch
• Charlie Mead
• Tony Julian
• Cecil Lynch
• Ron Parker
• Mead Walker

Agenda

• Roll call
• Minutes from last meeting
• Opening statement RE ArB responsibilities/authority (Chair)
• GE Issue (hopeful closure)
• Mission/Vision (???)
• ‘As is’ architecture (suggestions from Chair and discussion)
• Dynamic Model (update)
• Other Items

Minutes

• CM gave an update on the TSC discussions and reading the ArB minutes - What ArB needs to do is to stop asking permission and start doing stuff, and count on TSC to limit set.
• GG noted that the GE issue can be referenced here: http://www.hl7.org/v3ballot2008jan/html/domains/uvmr/uvmr_DataConsent.htm#RCMR_RM010001UV-rmi ... CM restated the GE issue: MnM position is that HL7 does not ballot terminology bindings except to structural attributes (more generally universal attributes). Terminology updates can happen outside of ballot, or through harmonization. It is not HL7's responsibility to be balloting terminology bindings ... about to issue this as a document and ballot. So in this case GE's issue is non substantive, because this is a non-structural binding to terminology. ArB needs to request clarification through MnM. Through a ballot, you need to specify a concept domain. AMS noted that the MnM committee has taken as an action item to clarify this issue. CM: Fair statement to GE ... Committee acted appropriately. MW: Perhaps we suggest not publishing as DSTU until the vocab is harmonized through harmonization process. CM: GE issue is an instance of inconsistency of balloted material. ArB should communicate directly with MnM about what the issue of inconsistency involves, and what the resolution criteria should be. Do this esp in light of QR's / Ballot Requirements to make sure that items like this don't go to ballot. Here is why this will not happen again. Can we generalize these into quality statements? CL: This was dealt with in part by Vocab Committee this last quarter. It is unclear whether the bindings in the ballot really _are_ universal. MnM needs to put down a step wise process on how this binding has to occur. CM: There has to be semantic consistency as well as agreement at the International level. This crosses responsibility boundaries between committees. JQ: Recommended sending it back to Foundation Steering Division for resolution. CM will communicate that there needs to be a deeper understanding between the Vocab Committee and MnM to work out what this resolution should be. JC : Universal Domains needed to be agreed to by the International representatives - there is a need for International representatives have the opportunity to have visibility of universal domain issues. CM: ArB needs to decide: do we think that a committee sent to HQ that appropriate vocabulary was harmonized. Do we say to the committee that when binding to something as universal, you at least need to specify a concept domain, and there needs to be some International validation and more internal control? Or do we say to GE that in spite of a botched process, that there is a larger QA effort underway, and vocab _was_ harmonized, so all is well. AMS: We want all of our specifications to refer to a domain that _exists_. The committee can satisfy the negative by having the concept domain added through harmonization. CM: Proposed ArB resolution: "ArB has reviewed the issue, agrees that there was a problem of timing, but that the committee has taken appropriate action (harmonization has occurred and domain has been added), and that GE should, de rigeur, withdraw its negative vote." JC: Publishing the results of the harmonization and following up with the committee should be part of the harmonization process. You can't simply publish without an appropriate notification to interested parties. GG: We should say that Balloting should stop until the processes, tooling, and structural issues are resolved. Also: Suggest that the committee did nothing wrong, and that the publishing process (tooling, timing, process) needs to be fixed. JQ / CM: GE's position is reasonable, but that the issue is resolved, and that TSC should resolve the larger publishing issue. CM will put out a proposed resolution.
  • As a convenience, here is what I take to be the relevant material:
    • ActInformationAccessCode, ActConsentType, ActInformationAccessContextCode, and ActPrivacyPolicyType are not defined within HL7 at this time.
    • Add these definitions to the appropriate vocabulary domains.
    • I cannot recommend something for DSTU that is not fully specified.
• CM: We need to base our mission / vision on the notion of architecture. And architecture should include the machinery of HL7. Also need to do the "to do" list for what ArB needs to fix (plus proposed solution). JK will set up a wiki page. Tony Julian