ITS Concall Minutes 20120807

Return to: ITS Main Page > ITS Meeting Minutes > 2012

Return to ITS Meeting Minutes

ITS Teleconference - August 7, 2012

Present: Paul Knapp, (PK), Brian Pech (BP), Dale Nelson (DN)

Chair: PK

Scribe: PK

Agenda

1. Role call and agenda
2. Approve Minutes of previous meetings July 24
3. Update to ITS Guide posted to ballot
4. ID/IDREF for CDA R3
   • PK will update from discussions with past ITS co-chairs
5. Action Item Update
   • Datatypes 1.1
     • Last action: Publication request done - April 24, 2012
   • Structures 1.1
     • Last action: Publication request (project 335) - July 10, 2012
   • ITS Guide
     • Last action:In ballot for September 2012 - July, 2012
   • Web Services
     • Last action: Vassil Peytchev has reported he can assist with Web Services at Vancouver WGM (will update in July/August) - May, 2012
   • InM Abstract Transports
     • Last Action: Reconciliation proceeding on joint InM/ITS call - July 10, 2012
   • R2B schemas
     • Call scheduled for July 13, 2012 to determine
   • Wiki page update:
     • Last action: AS presented the work that he and Yunwei Wang had completed over previous week, updates should be complete by next meeting - February 7, 2012
   • ISO Datatypes release 1 - outstanding negatives
     • Last action: PK will chase down outstanding negatives- January 31, 2012
6. Review of Product Briefs
7. Next Meeting
8. Adjourn
9. Adjourn

Called to order at 4:00 pm Eastern

No Quorum:

Discussion Item: CDA ID/IDREF

PK reported that Structured Docs (SD) had requested at the Vancouver 2012 WGM that the xml contract of ID/IDREF be used to handle references to 'objects' in CDA R3 instances. He also reported that, from discussion with Charlie McCay, that ITS had previously considered and rejected the use of ID/IDREF due to a minor degree on tooling support, and to a major degree based on the need to have the construct of 'references' handled within the HL7 methodology, then implemented via the different serializations, rather than being handled directly via a particular serialization without the benefit of an abstract HL7 modelling construct.

The issue is how to capture a 'reference' rather than the concrete 'instance' of a model or model snippet in a manner which can survive serialization and de-serialization through different markup languages. Currently 'references' are held in RP data type in V2, and in the 'reference' attribute of an 'ED' data type in V3 (eg. Model -> xml -> model -> JSON - > model). Also, we need to understand if this work is scoped only to ITS, or involves M&M, given the need to identify the abstract concept, and/or Tooling, given the need to detect target instances and produce renderings.

While it is potentially possible to have 'tooling' detect 'ED' instances which just use 'reference's and to render these in xml as 'xml IDREF' references, the ID to which this reference refers is generally an HL7 ID which is an element of type 'ID' which has a root and optional extension attribute - NOT an 'xml ID' attribute. Unless we can also detect and appropriately render 'xml ID' attributes we will have 'refs' but not the 'ID' to which they refer - unless we simply add the 'xml ID' attribute to all potential 'ID's which may significantly bloat the instances.

We will continue the discussion next week.

Call finished 5:10 Eastern