20110210 arb minutes

ArB Agenda/Minutes

Agenda

• Call to order
• Approval of Agenda
• Approval of Minutes
• SAIF book Project Scope
• HDATA - Defered until next week - John Koish is unavailable
• SAIF IF for the SAIF book
• Other business and planning for next meeting
• Adjournment

Meeting Information

HL7 ArB Work Group Meeting Minutes Location: Telcon		Date: 20110210 Time: 4:00pm U.S. Eastern
Facilitator	Charlie Mead	Note taker(s)	Tony Julian
Attendee	Name	Affiliation
X	Bond,Andy	NEHTA
X	Curry, Jane	Health Information Strategies
.	Grieve, Grahame	Kestral Computing
X	Hufnagel, Steve	U.S. Department of Defense, Military Health System
X	Julian, Tony	Mayo Clinic
.	Koisch, John	Guidewire Architecture
X	Loyd, Patrick	Gordon point Informatics LTD.
X	Lynch, Cecil	ontoreason LLC
X	Mead, Charlie	National Cancer Institute
X	Ocasio, Wendell	Agilex Technologies
X	Parker, Ron	CA Infoway
.	Quinn, John	Health Level Seven, Inc.
.	Shakir, Abdul-Malik	Shakir Consulting
.	Guests
.	Haddorff, Rick	Mayo Clinic
.	Laskso, Lynn	HL7 Staff
X	Milosevic, Zoran	NEHTA
X	Pech, Brian	Kaiser Permanente
Quorum Requirements Met: Yes

Documents

http://wiki.hl7.org/images/2/2a/HL7_Project_Scope_Statement_v2011Jan_--_ArBSAIFBookv2.docx

Minutes

• Call to order
• Approval of Agenda
• Approval of Minutes
  • Move Approval - Jane/Steve
• HDATA - Defered until next week - John Koish is unavailable
• SAIF IF for the SAIF book
  • Cecil introduced the document in Australia
  • needs comments ASAP
  • Zoran Milosevic has comments he will forward -
  • Jane Curry - we all need to stop using 3 letter acronyms - it is hard to use them in various places. We need to use conceptual, etc. There were places where there was not enough rationale for the recommendations received. There was no dicsussion of meta-data.
  • Cecil Lynch - What type of meta-data?
  • Jane Curry - include information about information someone externally might not have, include mapping to external data at the logical level to meet the implementable specification.
  • Cecil Lynch - I need something more complete.
  • Jane Curry - I saw the IF viewpoint , not the information management viewpoint that lives in architecture 2 space when someone is driving down the stack.
  • Cecil Lynch - I will wait for your comments. To me the meta-data is the RIM - the framework for how you will apply things as well as data-types. In a platform-specific model I will use those concepts in a specific application. Since you know this now, I need some concrete ways you are thinking about meta-data.
  • Jane Curry - Comes from the semantic work of Robert Worden. We have meta-data that describes the specifications, with which a project may be unfamiliar. e.g. column descriptor, or a 11179 data dictionary, or a physically described criteria of a valid instance. You have to be able to interpret the difference between labels and definitions in a concept. We need to deal with the 2 architecture environment - relate to an external environment to the data they need.
  • Andy Bond - Recognize a class as meta-data, not bound to a usage.Need to position quaility classes. I will forward work Zoran and I did - forward to the list.
  • Zoran Milosevic: - That would be great - my major question is the scope - we are focusing on HL7 concepts. Will we support open-EHR - do we have a generic way of defining information?
  • Charlie Mead - This is the difference between the canonical framework, and the HL7 frameworks. It should specify models and data-types. You should be able to build a IG without binding to the RIM.
  • Zoran Milosevic - We wanted to have this canonical model- this is all captured in our information perspective. The document Andy refered to is rought.
  • Jane Curry - I was looking at it in the governance framework - the scope is bigger than any specification - pertinent to an ECCF Stack but outside of it. Managing semantics over time is the issue.
  • Ron Parker - I appreciate the use for organizations external to HL7. Would HL7 map to RIM-based?
  • Charlie Mead - It is instantiation.
  • Ron Parker - That is important. Still it might be possible to consume HL7 artifacts, which would require RIM mapping.
  • Charlie Mead - it does not make the work go away. If the canonical is defined, the importance of information model, data-types, and semantics. Does not make it easier, or go away.
  • Ron Parker - Challenge of those who dont like RIM are not bound to it - but should map to it. RIMBAA can stay in native mapping, but the RIM is not absolute.
  • Jane Curry - At the implementable level they have bound to RIM, data-types, etc. You have to investigat the code to understand it's meaning. That is the point I was getting at - to achieve the stairway to heaven, and dont have the mapping from to the same level as the design model.
  • Cecil Lynch - Mapping does not belong in the IF.
  • Jane Curry - Right - but you must map to create the interoperability environment - if you have labels that are not defined, you might not do it. Specially if you are infering the meaning of data from behaviour.
  • Cecil Lynch - careful not to make assumptions about mapping to something else. If you think you can map a 11179 model to a RIM environment, you are dreaming.
  • Charlie Mead - I agree - we live it every day at NCI. We have a better chance of meaninful uptake/understanding, we should say that HL7 is bound to RIM, etc, but allow those to come to mapping on their own.
  • Jane Curry - Understand the context of intended use -which many applications dont document. When they participate in iteroperability, they have to do so - specially sharing data for secondary use - must specify context as well as physical use. That is a governance issue. If you evaluate that you cannot participate in V3 or CDA because you can provide repeatability, etc, then that is all you can do - CSI is impossible.
  • Charlie Mead - Zoran and Andy have weighed in, I will look at it. Any other comments on this call? Commit to putting on the agenda for next week, get the comments to Cecil in time so he can address.
  • Jane Curry - Karen Smith is prepared to make the changes to conceptual, etc, but we have no money to pay her to do so.
  • Charlie Mead - The basic notion in the ECCF is to describe a complex system a proven strategy is to cross dimensions(ODP) with perspectives. Our perspectives are conceptual, logical, implementable, and we should lable them as perspectives. We are clear of viewpoints, and should be equally clear about the perspectives.
  • Zoran Milosevic - we need to nominate some criteria - what is conceptual, logical. Conceptual is for subject matter experts. Logical refines to put in IT representation. Implementable we are less clear.
  • Charlie Mead - We hav stormed over who owns those rows. The boundaries are not clean - Conceptual by domain experts and analysts, analysts help architects define the logical, architects work with developers to define implementable.
  • Jane Curry - It is a three schema architecture, that is 40 or so years old.
  • Charlie Mead - Some semblance of that is in NCI IF chapter 4. We can fold it in. We need to get comments to Cecil by Tuesday for discusson on Thursday. Jane and my comments are outstanding. Jane, can you do so by monday? Cecil, status of comments and second draft.
  • Cecil Lynch - yes, need comments before doing second draft.
• SAIF book Project Scope
  • Charlie Mead - The TSC wants the SAIF booklet to come to informational ballot in May, so we need it by late March or early April. I wrote statement and sent to Ron and Austin. At TSC Lynn shot it down since the Domain group has not approved. Real time, or email vote? Nice if we could get to TSC on Monday.
  • Steve Hufnagel - Seem straight forward.
  • The remainder of the call was crafting the PS.

MOTION: To approve the Project Scope statement for forwarding to the TSC. (Tony/Jane) (8-0-0)

Adjournment

Meeting adjourned at 5:05 Eastern

Tony Julian 22:20, 10 February 2011 (UTC)