MnM Minutes WGM 201005 Rio de Janeiro
- 1 Sunday,May 16, 2010 - Q3
- 2 Sunday, May 16, 2010 - Q4
- 3 Monday, May 17, 2010 - Q1
- 4 Monday, May 17, 2010 - Q2
- 5 Monday, May 17, 2010 - Q3
- 6 Monday, May 17, 2010 - Q4
- 7 Tuesday, May 18, 2010 - Q1
- 8 Tuesday, May 18, 2010 - Q2
- 9 Tuesday, May 18, 2010 - Q3
- 10 Tuesday, May 18, 2010 - Q4
- 11 Wednesday, May 19, 2010 - Q1
- 12 Wednesday, May 19, 2010 - Q2
- 13 Wednesday, May 20, 2010 - Q3
- 14 Thursday, May 20, 2010 - Q3 Incomplete model notes
- 15 Thursday, May 20, 2010 - Facilitator's Roundtable
Sunday,May 16, 2010 - Q3
- Agenda Topics Review
- Hot Topics Triage
PRESENT: Rene Spronk, Austin Kreisler, Patrick Loyd, Grahame Grieve, Lorraine Constable, Hugh Glover, Jean Duteau, Lloyd McKenzie, Woody Beeler, Gregg Seppala, Rik Smithies
- Hot Topics to discuss
- Object Identity
- Safe interpretation of data subset
- CMET process & Direct Model referencing
- Template versioning
- Cumulative Dose and Dose Check Quantity
- Incomplete Static Models
- Context Conduction in non-serialized models
- ACTION: Grahame will produce a webpage of the most common datatype problems and link to it from the M&M wiki.
- Context Conduction was decided to be dealt with outside of the WGM.
- ACTION: Woody will update the WGM schedule and email it to the list.
Sunday, May 16, 2010 - Q4
- Austin Kreisler, Jean Duteau, Woody Beeler, Gregg Seppala
- Tooling Overview For Facilitators
Given who was present, we changed the agenda to summarize issues and tasks for tools needed to develop and publish Ballot 2010Sep. Formal approval was not sought, but the results reflect a clear consensus of the discussants.
Desktop Publishing needs to support:
- Publishing Cycle
- Prepare Graphics, (clickable) from Visio
- Extract xml (like uvpa.xml) from DB
- Position source:
- Put files in right directories
- Clean temporary directories
- Run Generator
- Position interim output:
- Put files in right directories
- Run publish domain
- Provide "Scorched Earth" BAT files for Temporary files
- Document and expose Bat File for Static Model documentation merger migration
- Clean up for Publishing "Temp" files
- Define a step-by-step process using SVN
- Provide a reference PubDb with SC MT QI MI CI domains and a merge capability to replace prievious stuff; include corrected vtPubSections
- Domain merge is missing annotation types and not updating with "new" annotations
- Visio "Repair document" SCREWS the design (Switches all class codes to "ACT" and then tries to drop attributes that are not in Act.)
Monday, May 17, 2010 - Q1
- MIF2 Reconcilliation
- Lloyd McKenzie (chair), Ravi, Jean Duteau, Michael Steine, Tessa van Stijn, Woody Beeler (for last 15 minutes)
- Worked through the MIF2 comment spreadsheet.
- 17 - MOTION: Ravi/Jean
- 19 - MOTION: Ravi/Jean
- 20 - MOTION: Ravi/Jean
- 34 - MOTION: Ravi/Michael
- 36 - MOTION: Ravi/Jean
- 37 - MOTION: Ravi/Michael
- 39 - MOTION: Ravi/Jean
- 41 - MOTION: Jean/Michael
- 42 - MOTION: Ravi/Jean
- 43 - MOTION: Ravi/Jean
- 44 - MOTION: Ravi/Woody
- Typos - MOTION: Woody/Ravi
- Affirmatives - MOTION: Ravi/Woody
We will finish the rest of the Affirmatives either in another quarter or in a teleconference call.
Monday, May 17, 2010 - Q2
- Joint with Project Services
- Woody Beeler, Jean Duteau, Rick Haddorf, Ioana Singureanu
Rick and Ioana joined MnM to talk about an analysis that Project Services is undertaking to document and streamline the development and balloting processes as HL7 moves to SAIF. They took notes relative to the slide deck they presented. Our discussion focused on the current harmonization and review processes used for the RIM in MnM. We also noted that they should confer with Vocabulary relative to Harmonization and with V3 Publishing relative to ballot cycles, etc.
Monday, May 17, 2010 - Q3
- RIM Reconciliation
- Woody Beeler, Peter Hendler, Alexander Henket, Ioana Singureanu
Because there was a marginal quorum, we agreed not to reconcile Negative votes. We discussed all line items and agreed upon actions for the A-T A-S and A-C categories of votes. These are documented in the reconciliation spread sheet that will be posted after preliminary recommendations have been drafted for the Negative votes.
Monday, May 17, 2010 - Q4
- Core Principles Reconcilliation
- Woody Beeler, Victor Chai, Christhof Gessner, Ioana Singureanu, Michael Steine
The reconciliation focused on items submitted by Gessner and Singureanu, and which were in the categories supported by Beeler. The items reconciled are in the Reconciliation Spreadsheet.
After the WGM adjourns, Beeler will collect the reconciliation spreadsheet from T. Klein and merge his items with those from the M7M reconciliation. In addition, it was noted that the spreadsheet includes a number of line items from the RIM Ballot. These will be removed during the merger.
Tuesday, May 18, 2010 - Q1
- Grahame Grieve, Woody Beeler, Pedro J. Rodrigues, Michael Steine
- Core Principles Reconciliation
Core Principles Reconciliation
Because of marginal quorum, we only addressed a number issues assigned to Grahame Grieve and for which the negatives were being found "Persuasive" or "Persuasive with mod". These are flagged in the spread sheet with "TueQ1" in the grouping column.
Continued Reconciliation on Conference Calls
In order to continue reconciliation on a schedule that will permit G. Grieve to participate, we agreed to schedule these for an hour each Thursday at 4PM Eastern time, beginning June 3.
Tuesday, May 18, 2010 - Q2
- Reconcilliation Overflow
FAILED TO MAKE QUORUM
The designated chair and one other person were present.
Tuesday, May 18, 2010 - Q3
- Hot Topics - "CMET Process & Direct Model Referencing"
- Woody Beeler, Lorraine Constable, Lloyd Mackenzie, Ravi Natarajan, Rik Smithies
Discussed the questions surrounding CMET Management. General consideration is how to manage the CMET namespace, and assure proper "attribution level" assignments while placing the majority of the responsibility with the developing domains. Then the discussion morphed into the second half of the agenda.
Direct Model Referencing
There are currently two forms of model-to-model referencing (linking) - "stubs", as used to link wrappers to their contents, and CMET references. HL7 UK has asked that we add a third form in which the model being linked is designated directly by its identifier and, perhaps a version.
The discussion then proceeded to document the three forms of reference that would exist, with the following notes recorded:
Stubs: Reference (in the referring model): (uses MIF element "templateParameter") - these references are unbound until they are used in a dynamic model definition when they are bound to another model (wrapper or payload). The templateParameter includes a "name" (alias), "derivedFrom" (RimClass), "definingVocabulary" (classCode). (These elements provided to allow interpretation of the model without the IFC file).
The stub definition (in the interface file): (MIF element "stubDefinition") - The seclaration is characterized by: a name (alias), derivation (Rim Class) and a defining vocabulary (classCode), and a min (could be omitted) and a max model (eg. AI DMIM for control act wrapper and RIM for a payload)
Stub says you can bind (at some later time) any model whose characteristics "match" the declaration.
CMET - reference Appears in the source model as MIF element "commonModelElementRef". The reference is unbound until a specific ITS is invoked at which time the "IFC model" binds the refrerence to a particular model. The binding is by CMET name. Note, if no such binding can be found, the invocation of the ITS "fails".
The declaration in the IFC model is characterized by: a name (alias), derivation (Rim Class) and a defining vocabulary (classCode), and a min (could be omitted) and a max model (eg. AI DMIM for control act wrapper and RIM for a payload)
Direct Model Reference (requested addition) In the source model would be represented by a new MIF element "importedModelRef". The declaration would contain: a model id (as in packageLocation), a name (alias), derivation (Rim Class) and a defining vocabulary (classCode)
McKenzie moved, Smithies seconded to place "importedModelRef" into the MIF as soon as possible Item passed 3-0-0
- Beeler and Stechishin to start Hot Topic to further define CMET Management issues and future processes.
- The discussion extended to reference/binding management and the degree to which the model reference declarations are or ARE NOT constraining the actual binding. This discussion needs to be continued under a Hot Topic.
- Agreed that these three forms of reference will be documented in detail in the Core Principles by McKenzie.
Tuesday, May 18, 2010 - Q4
- Hot Topics - "Template Versioning" & "Cumulative Dose and Dose Check Quantity"
- Calvin E. Beebe, Woody Beeler, Keith Boone, Sarah Gaunt, Christhof Gessner, Hugh Glover, Tom de Jong, Rik Smithies, Rene Spronk
Template binding in context of implementation guides
Issue raised from Structured Documents in regards ability to bind templates by both an ID and a version, because this would make management of the bindings simpler.
Proposed solution is provide both a templateId and a versionDate. Then the Reference to the template would be the templateId (required) and an effectiveDate (optional). If date is provided, the binding would resolve the the version of the identified template that was effective on the date specified. If not date is specified then the binding is "dynamic" to the template version in force on that date.
Noted that this is how value set binding is being managed in HL7.
Within an Implementation Guide, one should further have ability to define a date as part of the implementation guide which date would bind all dynamic references as of that date.
Although there was general agreement that this process would work, it was not advanced as a solution because thjere is uncertainty as to how to represent this combination in an instance, where the II data type contains both a "root" and an "extension."
Cumulative Dose and Dose Check Quantity
There was extensive discussion of this, but no documentation was provided. The discussants agreed to propose changes to these descriptions of these attributes that would include an example as to how to represent this.
Wednesday, May 19, 2010 - Q1
- Hosting Joint with Vocabulary - MIF Reconciliation
- Woody Beeler, Heather Grain, Stan Huff, Ted Klein, Beverly Knight, Lloyd Mackenzie, Rita Scichelone] , Stefen Sebutsch
MIF passed Informative Ballot. Discussed the role of "harmonization" as a feedback/correction path for flaws in model content, methodology, etc.
- Need to identify material that is "out of scope" for next ballot
- Need to consider sub-setting content by level of detail (at some point in future)
- Agreed to desire to get currently agreed-to material through and Normative to have a stake in the ground
- Agreed to ACTION: joint review around 2nd week in July to identify strategy for advancing to the September Ballot
- Both groups will have a 1-hr reconciliation sessions. (M7M new weekly conference calls at 4PM Eastern, starting 6/3; Vocab calls will be one hour of their normal bi-weekly call schedule
- Will Revise the form based on McKenzie updated checklist
- Write a contextual manual - for proponents
- Will Clean up the "harmonization home page"
- remove obsolete tools,
- add a correct tool for mark-up;
- change coverpage/proposal to initialProposal/finalProposal
- Enforce content "complete" at first deadline. Second deadline for correction, not for new content.
- Retain the Technical Review and feedback between deadlines
- Need assure representative attendance.
- Representative Quorum - representatives from
- MnM*, Vocab*, Structured Docs, International*, Proponent Work Groups, DESD, SSDSD
- Mandatory representatives from the * above plus the proponent for the item under review
- A Minimum of 6
- No one voting for two groups
- Mandatory attendance must be fulfilled by separate individuals for each required role.
Wednesday, May 19, 2010 - Q2
Joined Joint with Vocabulary
Wednesday, May 20, 2010 - Q3
Hot Topics: "Object identity" & "Safe interpretation of subsets of data"
Grahame Grieve, Woody Beeler, Lloyd Mackenzie, Rene Spronk, Victor Chai, Sarah Gaunt, Gaby Jewell, Ravi Natarajan
Object Nets and Object Identity
Discussion concerning Managing Object Identities. Object identity
- Without an identifier on an object, it's very difficult to manage the object. So identifiers should never be constrained out
- even if the static model constrains an association out, it may still be actually in the object model because of context conduction or past exceptions
- Role.id is a muddle - not clear what the scope of the identifier is. R2 differentiation of business vs object identifier is helpful but does not resolve the problem. Rene will update Object identity to describe R2 on this matter. Recommend pushing adoption of R1.1 as much as possible in this regard
- Entities - some are identified by code. Propose that Vocab create an property of a code system to say whether a code is also an identifier (Appellation)
- Participations and ActRelationships - do not have identity. In general there should only be one association between an act and a (act|role) with the same type and effective time
Safe Interpretation of RIM Data
- Lloyd - Short answer: anything that's required in the applicable model.
- More general answer: in the absence of a conformance context, nothing can be ignored. You can only ignore things because the conformance statement instructed the RIM author that these things must be ignorable
- It's generally safe to interpret an act in the absence of it's context because of moodCode - except for moodCode Opt (if the mood code is opt, the act is incomplete and the owner must be considered)
- You always have to consider the intent to decide how the act must be interpreted. Recommend to consider the actRelationships that point at the act - depends on your intent, and consider the type they have as to whether the context they suggest is relevant to your intent
- It's really hard to interpret any data in the absence of the specification because of the poor quality of our existing models. This is especially apparent in CDA R2 profiles where the choice of Act Relationships is so limited
- you must consider associations established by context conduction
- We can't advise on what could be generally ignored. It does depend, but there's nothing that we can say for sure that can be ignored
Thursday, May 20, 2010 - Q3 Incomplete model notes
Reviewed Grahame's comments on Incomplete Static Models and developed and approved resolution
Discussed Order interaction for the creation of a document open issue from Rene. Agreed that using an id attribute on the definition was a good way to reference an external template. The actual construction of a template that provides constraints on vocabulary, cardinalities, etc. in RIM terms is still a problem, but that's outside the scope of the issue raised.
Discussed issue from Rene about ability to have a generic query that includes a parameter identifying what message type to return in the response. Only way to do this at the moment is to have the payload for the response interaction contain a choice of the allowed payload models; or to have the payload be a model representing the whole RIM. (Though conformance on the latter would be nearly impossible.)
Thursday, May 20, 2010 - Facilitator's Roundtable
- Harmonization Discussion
- MnM Summary
Woody (chair/MNM), Bob (StructDocs), Jean (Pharmacy/ITS/scribe), Hugh (Medication), Patrick (OO/InM/CS), Lorraine Constable (visiting), Peter Hendler (visiting), Gunther (visiting), Christoph (Healthcare Devices), AMS (RCRIM/CIC), Cecil (ArB Liaison), Ted (Vocab), Rene (RIMBAA), Wendy (PA Vocab), Gregg (PA), Lloyd (MnM)
- Discussion of Harmonization Process (see Wed Q1 discussion)
- Upcoming Dates: July 4 - Coverpages due; July 25 - Final proposals due; Harmonization - August 3-6 (noon-4pm Eastern)
- progress with template-related tooling and repository
- looking for projects for using template tools
- talked with InM about green-CDA
- doing internal QA of their topics and will be bringing vocabulary proposals from their topics
- Question: when documenting attributes, are there some that do not need to be documented?
- Answer: Yes, any fixed-value immutable attributes do not need to be documented.
- will be balloting a document about Micro-ITS
- CMETs were reballotted based on the Common Product Model
- Question: Online ballot contains old CMETs while the downloaded ballot contains new CMETs. Does the ballot stand or do we need to reballot?
- ACTION: (Publishing) Provide a detailed note to Don and Woody about the error so that they can look into it.
- Question: Can we move the Medication message pieces into Pharmacy domain and keep the POME identifier?
- ACTION: (Publishing) Don and Woody will look into this to see how it would work.
- Question: Do HMDs have any purpose?
- Answer: No, they do not.
- Looking at Clinical Statement. It has been a "super-model" that is hard to maintain.
- Same trouble as Clinical Statement. How do we maintain the Composite Order "super-model"? How do we link up with Common Product Model? Behavioural Model is moving forward with storyboards to come up with some pretty simple concepts for explanatory purposes. Also dealt with ordering a document.
- Balloted a DAM and received no negative comments. Some of the Affirmative comments suggested substantive changes. If they make changes, they can limit the ballot to just those things that have changed.
- Draft of the Information Framework was presented
- Question: Are there HL7 code system valuesets that are in CDA?
- Answer: Yes, because it is an Realm-specific IG, they have not been bringing them forward.
- ACTION: (StructuredDocs) We will be determining how to bring those forward to Harmonization.
- For Universal bindings to other than CS-bound systems, approval of harmonization becomes a recommendation to the International Affiliates Council. We are waiting for the IAC to determine how they want to handle Universal bindings.
- Batch Wrapper balloting - the comments we received may result in dropping some of the attributes of the Batch classes.
- Dealing with the concept of an Integrated Registry that was brought up from Canada Health Infoway. Looking at Personnel Management, Patient Administration, and some balloted service specs shows wide variance.
- Example of issues:
- Desire for the query placer to be able to shape the query response
- PM has a query to return a formatted text report (consider Document Ordering)
- Subscribing to receive notifications (old v2 pub/sub)
- Request message to change the state of the information object (as opposed to creating a bunch of Trigger Events to make specific state changes).
- How to determine if a request was successful? (there appear to be two different patterns to do this)
- ACTION: Gregg will enter these as MnM Hot Topics.
- Reviewed the week's work (see the above minutes)