This wiki has undergone a migration to Confluence found Here
<meta name="googlebot" content="noindex">

MnM Minutes WGM 201201 San Antonio

From HL7Wiki
Jump to navigation Jump to search

Return to MnM Minutes for 2012

Sunday Jan. 15 Q3

(Llano - Ch: L McKenzie; Sc: G Beeler)

Attendees

George (Woody) Beeler Jr (Beeler Consulting LLC ), Jean Duteau (Gordon Point Informatics Ltd. ), Attila Farkas (HL7 Canada ), Sarah Gaunt (NEHTA ), Peter Gilbert (Covisint ), Hugh Glover (HL7 UK Voter #6 ), Austin Kreisler (SAIC - Science Applications International Corp ), Jay Lyle (Ockham Information Services LLC ), Lloyd McKenzie (HL7 Canada (Gordon Point Informatics) ), Rajan Rai (HL7 Canada ), Rik Smithies (NProgram Ltd ), Rene Spronk (HL7 Netherlands ), Andy Stechishin (Gordon Point Informatics Ltd. ), D. Mead Walker (Mead Walker Consulting ), Philip Wilford (nehta ), Grahame Grieve, J. D. Baleer

Agenda

Establish Agenda for WGM from Hot Topics and Other Items

Steps:

  • Reviewed the existing Hot Topics
  • Marked changes on HXIT for Associations and NullFlavor
  • Agreed to have L. McKenzie "clean up" the Hot topics list and bring his final changes to the Facilitator's Roundtable on Thursday evening for adoption.
  • Added a number of Hot topics to the agendas for this WGM (also outlined below)

Adjourned 3PM

Monday Jan. 15 Q1

(Regency East 2 - Ch: L McKenzie; Sc: G Beeler)

Attendees

George (Woody) Beeler Jr (Beeler Consulting LLC ), Lloyd McKenzie (HL7 Canada (Gordon Point Informatics) ), AbdulMalik Shakir (City of Hope National Medical Center )

Agenda

SAIF Artifact Definition - REVITALIZE the process

Discussion

Started with question of whether to choose to advance only a sub-set of the artifacts in the face of low adoption of "universal" messaging artifacts plus the needs of FHIR.

Suggest that the Information Model artifact definitions are usable in both contexts, although, in future, the requirement for creating RMIMs may be limited.

Simultaneously, there are clearly new artifacts and governance required to support FHIR.

What is role of behavioral/transactional models in FHIR??? Not immediately.

FAILED to MEET QUORUM

Monday Jan. 16 Q2

(Regency East 2 - Ch: G Beeler; Sc: L McKenzie)

Attendees

Woody Beeler, Lloyd McKenzie, Adel Glhamahalla, Rick Smithies, Raj Rajan

Agenda

Ballot planning

  • RIM R5
  • Core Principles R2 - when & what
  • Data Types R2+

Ballot Planning

RIM R5

  • R4 passed balloting in only 1 cycle
  • Number of changes this year will be small - primarily structural vocabulary
  • In this next release, a bunch of deprecated elements will be dropped
    • Will need to manage how this will affect publication for NE2013

Motion: Woody authorized to submit paperwork to initiate RIM R5 ballot Lloyd/Rik: 4/0/0

Core Principles R2 (when & what)

  • R1 is finished with one negative outstanding
  • Carry-overs from last time:
    • Behavioral models (still don't have this)
    • Negation indicators: List them all and how they should be used
    • isDocumentCharacteristic: what does it mean/how does it work
  • Can we do a full release or an Annex?

Motion: MnM authorizes Woody to prepare PSS and submissions to enable balloting of Core Principles R2, limited to the addition of documentation on negation indicators and isDocumentationCharacteristic.

Lloyd/Rik: 4/0/0

Data Types R2+

  • Looked at list of Data Type R3 issues
  • Consider doing a 2.1
  • We can bring any vocab changes other than null flavor forward into datatypes
  • We want people to identify what additional changes they want and how "urgent" a given change is.
    • Woody will update the page to be numbered and define a list of priorities from "blocking implementation" to "nice to have". When that's done, Lloyd will craft a small blurb for the HL7 Newsletter and send an e-mail to co-chairs list, MnM list, RIMBAA list, INM list and CGIT list

Monday Jan. 16 Q3

(Regency East 2 - Ch: L McKenzie; Sc: G Beeler)

Attendees

George (Woody) Beeler Jr (Beeler Consulting LLC ), Jean Duteau (Gordon Point Informatics Ltd. ), Robert Hausam (Hausam Consulting ), Julie James (Blue Wave Informatics ), William Ted Klein (Klein Consulting, Inc. ), Lloyd McKenzie (HL7 Canada (Gordon Point Informatics) ), Zoran Milosevic (NEHTA )

Agenda

Two Hot Topics

Role Class "Ontological"

Issue, coming from Vocabulary, problem found when attempting to satisfy Role relationships (found in context of common product model). Julie James brought forward a detailed analysis of the RoleClass hierarchy with the concern focusing on _RoleClassOntological.

The existing structure had a number of definition issues - SAME and SUBY appeared to be saying the same thing, yet were siblings, and generic is also tied into the definition of SAME.

After much discussion asked to see a harmonization proposal for March that restructures the hierarchy as:

ROL

_RoleClassAssociative
_RoleClassOntological
EQUIV (functionally equivalent entity)
SAME (with changes to make the fact that representations differ, but there is only one real-world thing that they represent)
SUBY (subsumed by) Modify the definition to make it clear that the original definition was to replace a record, and perhaps alter the print name to be "????"
GRIC
GEN
INST

Harmonization Proposal for QueryParameter.semanticsText attribute

Jean Duteau presented a revision of his parameterItem proposal related to this. The group critiqued it and encouraged him to submit the revision.

Tuesday Jan. 17 Q1

(Llano - Ch: G Grieve; Sc: G Beeler)

Attendees

George (Woody) Beeler Jr (Beeler Consulting LLC ), Jean Duteau (Gordon Point Informatics Ltd. ), Sarah Gaunt (NEHTA ), Adel Ghlamallah , Austin Kreisler (SAIC - Science Applications International Corp ), Lloyd McKenzie (HL7 Canada (Gordon Point Informatics) ), Zoran Milosevic (NEHTA ), Brian Pech (Kaiser Permanente ), Rik Smithies (NProgram Ltd ), Rene Spronk (HL7 Netherlands ), Andy Stechishin (Gordon Point Informatics Ltd. ), Grahame Grieve, Ewout Kramer

Agenda

FHIR - MnM Implications

  • Received an mini-tutorial in FHIR
  • Discussed a variety of methodology issues
  • FHIR will result in elimination of some artifacts and creation of new artifacts
  • Discussed issues re socializing and rolling FHIR out within HL7
  • Agreed that Beeler and others would look into the underpinnings of FHIR as it relates to shared development and publishing (with objectives of educating those like him who might have doubts.)
  • Agreed to craft SAIF Artifact Definitions for the core artifacts of FHIR

Tuesday Jan. 17 Q2

Attendees

Lloyd McKenzie (Chair), Mead Walker, Rik Smithies, Jean Duteau (scribe), Zoran Milosevic, Kathleen O'Connor, Bo Dagnall, Rene Spronk

Agenda

Two Hot Topics

Modeling Work Flow - Design Pattern

Mead, as part of building a Study Design RMIM, has come across the need to model work flow. He wrote up a document discussing the various attributes that can be used for expressing workflow. Lloyd commented that there are some additional ActRelationship codes that are not in Mead's document, eg. StartsAfterEndOf. These provide a little bit of overlap for some representations but they provide some extra means to represent things that Mead's document didn't consider.

We considered how to represent workflow with choices but no explicit criteria. There are probably implicit criteria but if they are not stated explicitly, how do we represent this. Mead would like one single ActRelationship type versus using a wider range of ActRelationship types. Lloyd suggested that we lay out the different options and indicate which are to be used in specific situations.

There are some scenarios in a workflow where there are relationships between activities that don't have critieria, i.e. A has to happen while B is occurring. Although these are parts of a grouping therapy act, there is a need to indicate the timing between those two, i.e. StartsAfterStartOf, EndsBeforeEndOf. There is also probably a need to use an ActReference to reference an act in a number of different places.

ACTION: Mead will take his initial document and write up the proposal for a pattern.

Act in DEF Mood

We have DEF mood for Acts and we have 'isCriterion' and there is confusion over the use of each. Each of them provide a means to present a template of an activity. Although there is a distinction between them, it is a hard distinction to pin down. We came up with "An ACT in DEF mood defines what an Act is allowed to be. An ACT with 'isCriterion' defines an Act that is being looked for."

Tuesday Jan. 17 Q4

(Llano - Ch: L McKenzie; Sc: G Grieve)

Attendees

Woody Beeler Lloyd McKenzie Andy Stechishin John Roberts Dale Nelson Pete Gilbert Adel Ghlamallah Grahame Grieve

Agenda

Templates and Models:

1. How do you indicate in a parent model, which notes are eligible to have templates applied. And for parent models which are already constrained (ie. Being used in an implementation guide) how do you indicate which specific templates are valid. Is the mechanism different for open versus closed templates.

2. For specific templates, how do you indicate what element on a parent model they are applied to.

3. In an instance, how do you identify which templates have been applied – ie which constraints are declared to have been applied.

4. Show examples of all answers.

5. Are there any tooling changes needed in the current RMIM Designer to apply 1 or 2. What about MDHT and SDM?

Discussion

1. invoking a template:

  • can only invoke an template at a rim entry point since templates can only be defined starting at rim entry points
  • can't limit which templates are valid or applicable
  • can specify a list of possible templates in the MIF - this is not exclusive, but can list some possible templates that are appropriate.
    • requiredTemplateGroup: "Identifies a set of templates which must be true for this attribute or association. If multiple sets requiredTemplateGroups are specified, then the complete set of templates in at least one of the groups must be true"
    • supportedTemplateFroup: "Identifies a template that is known to be applicable for this attribute or association. Essentially this is providing a 'hint' about possible templates that can be used. However, it does not require the use of any of the listed templates and is not necessarily an exhaustive list of templates that could be valid."
    • visio doesn't support this
  • mechanism is the same for open and closed templates

2.

  • template must constrain a particular class in another model
  • how you express that depends on how you express the template
  • Pretty much actually done in text at this point, though can be done in the MIF

Much discussion ensued from this point, involving the history of the template and the way templates are used.

3.

  • see the template spec for use of templateId - in the instance there should be a list of templates on the element.

4.

Example for a template id:

   <observation>
     <templateId root="2.16.840.1.113883.10" extension=" REPC_RM000103"/>
     
     <derivationExpr>Sumscore</derivationExpr>
     <effectiveTime value="200601191211"/>
     <value value="3" />
     <component>
       <observation>
         
         <value xsi:type="CO" code="1" codeSystem="2.16.840.1.113883.2.6.15.1.1"/>
       </observation>
     </component>
     <component>
       <observation>
         
         <value xsi:type="CO" code="1" codeSystem="2.16.840.1.113883.2.6.15.1.2"/>
       </observation>
     </component>
     <component>
       <observation>
         
         <value xsi:type="CO" code="1" codeSystem="2.16.840.1.113883.2.6.15.1.3"/>
       </observation>
     </componen>
   </observation>


5.

  1. - could do visio tooling changes to support those two MIF elements (moderate amount of work)
  2. - very large amount of work - (though would love to have it) (some pieces of it exist in the SMD)
  3. - n/a/
  4. - N/A


Summary

  • MIF does support the requirements from the template spec, but is not used for the templates we have actually published (via Structured Documents)
  • Providing the tooling to allow Structured Documents to use the MIF would be a significant challenge


Templates and Modeling - Hot Topics

Wednesday Jan. 18 Q1

(Regency East 2 - Ch: G Beeler; Sc: L McKenzie)

Attendees

George (Woody) Beeler Jr (Beeler Consulting LLC ), Carmela Couderc (Siemens Healthcare ), Peter Gilbert (Covisint ), Monica Harry (Gordon Point Informatics ), Robert Hausam (Hausam Consulting ), Wendy Huang (Canada Health Infoway Inc. ), Stanley Huff (Intermountain Healthcare ), Beverly Knight (HL7 Canada ), Andrea MacLean , Lloyd McKenzie (HL7 Canada (Gordon Point Informatics) ), Sandra Stuart (Kaiser Permanente )

Agenda

  • Vocabulary submission quality
  • Language code issues
  • Defining value sets as part of a static model

Host Vocabulary

Vocabulary Submission Quality

Problem with the quality of some harmonization submissions. Issue is that proposals aren't necessarily vetted by someone who understands vocabulary rules.

  • Action: Woody to update harmonization submission form to capture "vetting MnM facilitator" and "vetting Vocabulary facilitator". The former is required for all RIM changes, including structural vocabulary. The latter is required for all vocabulary changes, including structural vocabulary.

At some point, "vocabulary facilitator" may be required to have training, but for now, we'll settle for a name to blame/shame

  • Action: Woody to provide Vocabulary with a list of code systems under MnM control (structural & datatypes vocabulary)
  • Action: Woody to add bug tracker item for publishing to expose this in the ballot

Language Code

Ted is going to work on the issue where we have two different sets of OIDs that could be used for language codes (one for all ISO639 codes, and one for each of the 3 types). The single OID is referenced in the ISO spec

The actual RFC reference can't change until Datatypes R2+.

  • Action Wendy to add requested change to list of Datatypes R3 issues

How to Define Value Sets as Part of a Static Model

We need a way to be able to communicate a value-set definition as part of an instance. E.g. Sending a survey.

For study design, there can be a need to send value set definitions that aren't enumerated.

Best solution is to represent the value-set definition as a CTS as an ED within the instance

Action: (done) Lloyd will add to the Datatypes R2+ a requirement to introduce a new property to CD of "valueSetDefinition" with a datatype of ED intended to convey the CTS 2 value set definition. (with proper mime types, could do other syntaxes such as SNOMED refset format too.)

Harmonization requirement

Motion: Lloyd/Beverly 10/0/0 Workgroup health should only be based on committee harmonization representation if they would have "standing" as defined in the harmonization rules. Specifically: - vocabulary WG, MnM WG, SD wG - Proponent Work groups - Any Work Group that has content in the current Ballot Cycle or that has an approved project developing content for future DSTU or Normative ballot

Workgroups *without* standing, as defined above, should not be penalized in Work Group health for non-participation in harmonization. Furthermore, one person can represent more than one group provided they participate in all groups they represent.

Thursday Jan. 19 Evening

Attendees

Woody Beeler (chair), Abdul-Malik Shakir (CIC), Rene (Marketing/RIMBAA), Andy (Publishing/Tooling/ITS/Templates), Ted (Vocabulary), Amnon Shabo (Clinical Genomics), Lloyd McKenzie (M&M), Hugh Glover (Pharmacy), Julie (Patient Safety), Lorraine Constable (ArB), Wendy Huang (Patient Administration), Patrick Loyd (Clinical Statement), Erin Holt (PHER), Joginder Madra (PHER), Rob Hausam (O&O), Jean-Henri Duteau (Pharmacy/M&M), Mead Walker (Patient Safety/RCRIM), Norman Daoust (Anatomic Pathology)

Agenda

Revise Harmonization "Rules and Expectations of Participation"

TSC has adopted a rule that all workgroups besides T3SD must participate in Harmonization. On Wed Q4, MnM and Vocabulary took a motion which was presented to the TSC chair. It will be discussed at the next TSC meeting.

Schedule 2012 Harmonization

  • For the May ballot, March 13-16, 2012 (February 12 / March 4 deadlines)
  • For the September ballot, July 10-13, 2012 (June 10 / July 1 deadlines)
  • For the January ballot, November 13-16, 2012 (October 14 / November 4 deadlines)

Facilitators' Reports

  • Anatomic Pathology - working on creating a Domain Analysis Model related to specimens and other related things with identifiers. In conjuction with Clinical Genomics and O&O
  • CIC - Working on various Domain Analysis Model. One that will be balloted on Schizophrenia and another on Trauma Registry. Had question about how guidelines for DAMs will be created.
  • Patient Safety - Regulatory products revolving around reports. No ballotable material in the next cycle.
  • RCRIM - Specifications on trials, will be building a clinical document for study protocols.
  • Pharmacy - Balloting a R2 Medication Orders and Medication Dispense. Mainly will include institutional use cases as well as updating to new RIM and Clinical Statement compatibility. Producing a DAM on Medication Profile. Working on an Implementation Guide for using Pharmacy in different architectural environments. Project with SOA on using Pharmacy material in services. Project to look at FHIR Pharmacy resources. Many topics will be going into NE2012. OOC in early March in Oslo, Norway. Work on IDMP ballot is continuing with another ballot in May and will include a number of Harmonization Proposals.
  • O&O - Common Product Model will have a new ballot for the main structure and CMETs. This will be in Datatypes R2. The NamePart value set is okay to add for Medication name parts. Some Harmonization Proposals will be arriving.
  • PHER - Immunization Normative Ballot - no harmonization proposals are expected. Two CDA IGuides will be balloted.
  • PA - Encounter Link model will be reballoted.
  • ArB - Moving from balloting the SAIF IG to supporting workgroups.
  • Clincal Genomics - working with NCI's Life Sciences DAM but their DAM is research-focused. Balloting Clinical Genetic Testing IGuide and two CMETs.
  • Vocabulary - CTS2 was a focus and there will be two CTS2 ballots, CTS2 requirements and CTS2 IGuide. Also discussed how to continue the relationship with IHTSDO. Worked on the Terminology Authority project. Core Principles ballot has completed so they will be working on the Binding Syntax definition. TermInfo R2 project was kicked off. Facilitator Training will be worked on to provide mentorship and the ability for new facilitators to be produced.
  • ITS - ITS R1.1 has been reconciled. The other main concentration was on FHIR.
  • Templates - Joint project between Templates/ITS/MnM/StructDoc about the use of templates in CDA.
  • RIMBAA - Tool Evaluation of Software Development tools related to HL7 v3. Discussion of FHIR resulted in the discussion of a RIM path, i.e. a path used to identify a specific RIM attribute. RIMBAA is looking into defining a systems architecture for RIM-based applications.

M&M Meeting Highlights

  • Discussed using Sunday Q3/Q4 for actual MnM work, i.e hot topics, instead of scheduling to facilitate participation on these topics.
  • RIM R5 and Core Principles R2 will be balloting.
  • Datatypes R3 issues should be looked at to identify priorities. Datatypes R2.1 may occur whilch will consist of urgent "easy" changes.

Discussion of FHIR Talks

  • Agreed that there is a need for a number of talks about "What is FHIR?"
  • Vocabulary will request a terminology-focused talk about FHIR
  • Sun Q2 and Sun Q4 will be scheduled for FHIR talks - a Doodle poll will be produced to determine the best use of these quarters We will also list it in the Tutorial section as a free tutorial.

Friday Jan. 20 Q1

Attendees

Lloyd McKenzie (chair), Jean Duteau, Rene Spronk, Patrick Loyd, Woody Beeler

Agenda

MnM Business and Planning Mtg

  • Revised the Mission and Charter, the SWOT, the 3-Year Plan, and the Decision Making Practices. MOTION: To approve the revised documents. (Loyd/Duteau - 3/0/0)
  • Created the Draft Agenda for WGM 201205
  • Submitted the Room Requests for WGM 201205

Datatypes R2 Ballot 4 Reconciliation

  • DT R2 ballot 4 has a clear scope restriction:
    "This is the fourth membership ballot for Release 2 of the Data Type - Abstract Specification."
    "One substantiative change has been to this ballot, the removal of the CD.isCompositional property. This change is the only part of the document that is open for comment in this ballot. Other comments may be made, but will most likely be deferred to data types R3."
  • We currently have documented negative votes from 16 voters, on 5 different line items. Noneof the line items relate to the item specified in the scope limitation. Voters represented: Kaiser (6 voters, 1 item) , VA (1 voter, 2 items), and Siemens (8 voters, 2 items).
  • In Q3 Sunday meeting for the September 2009 WGM (Atlanta) we discussed that this did not need face-to-face resolution since all of the comments are out of scope, and will be deferred for R3. (Documented ONLY in notes of G Grieve and recollections of G Beeler). However, MnM FAILED to ever formally adopt this position in order to close reconciliation.

MOTION: MnM agrees that the action to find the comments "Not Related" and to defer to R3 should be adopted. We will notify the voters of a meeting in three weeks to take the proposed (Not related) action on these items, should they wish to speak to it. Then on that date (yet to be determined) we will "complete reconciliation" and ask for withdrawal. Should they not wish to withdraw we will recirculate ballot to affirm the decision with the ballot pool, per GOM.(Woody/Jean - 3/0/0)