20101003 arb cambridge wgm minutes

ArB Minutes

Meeting Information

HL7 ArB Work Group Meeting Minutes Location: Cambridge, MA		Date: 20101003 Time: 9:00am U.S. EDT
Facilitator	Ron Parker	Note taker(s)	Tony Julian
Attendee	Name	Affiliation
X	Bond,Andy	NEHTA
X	Curry, Jane	Health Information Strategies
X	Grieve, Grahame	Kestral Computing
X	Julian, Tony	Mayo Clinic
X	Koisch, John	Guidewire Architecture
.	Loyd, Patrick	Gordon point Informatics LTD.
X	Lynch, Cecil	ontoreason LLC
X	Mead, Charlie	National Cancer Institute
.	Nelson, Dale	II4SM
X	Ocasio, Wendell	Agilex Technologies
X	Parker, Ron	CA Infoway
.	Quinn, John	Health Level Seven, Inc.
.	Shakir, Abdul-Malik	Shakir Consulting
.	Guests
.	Haddorff, Richard	Mayo Clinic
X	Hufnagel, Steve	U.S. Department of Defense, Military Health System
.	Koehn, Marc	Gordon point Informatics LTD
.	Laakso, Lynn	Health Level Seven International
.	McGaughey, Skip	.
X	McKinnon, Bruce	Jassco Systems
X	Pech, Brian	Kaiser Permanente
.	Peres, Greg	CA Infoway
.	Robertson, Scott	Kaiser Permanente
.	Rospide, Eddy	Albany Medical Center
.	Smith, Karen	HL7 Technical editor
X	Thompson, Cliff(via telcon)	OntoSolutions LLC
X	VanDerZel, Michael	UMCG
.	Wrightson, Ann	HL7 UK
Quorum Requirements Met: (yes/No)

Agenda

Agenda Topics

• Sunday Q1
  • Call to order
  • Approval of Agenda for quarter
  • Approval of agenda for WGM
  • Approval of Minutes
  • Update from TSC (Ron/Charlie) (30-45 minutes)
  • Set direction for the week.
  • Arb Membership Criteria Review
  • SAIF Glossary Review
• Sunday Q2
  • Call to order
  • Approval of Agenda for quarter
  • Behavioural Framework (John Koisch) ???

• Sunday Q3
  • Call to order
  • Approval of Agenda for quarter
  • Information Framework and IFIG (Cecil Lynch)
  • Coherence ???

Supporting Documents

1. List any/all documents to be provided at the meeting, including their URL if applicable. For a WGM, indicate if hardcopies will be supplied or attendees should print off copies themselves

Minutes

Minutes/Conclusions Reached:

Sunday Q1

• Call to order
  • Called to order at 9:15am U.S. EDT.
• Approval of Agenda for quarter
  • Approved by affirmation.
• Approval of agenda for WGM
  • Approved by affirmation.
• Approval of Minutes
  • Approved by affirmation.
• Update from TSC (Ron/Charlie) (30-45 minutes)
  • Reviewed Roadmap - committe renamed to Strategic Initiatives Committee
  • Publishing facilitator will coordinate our Glossary work with Publications
  • Concern was raised that many of the HL7 leaders are consultants
  • Jane Curry agreed to be publishing facilitator
    • Discussion was held on what level the glossary should exist, as well as other HL7 content. Publishing/Tooling long term goal is to not only create a single ballot package, as well as custom view package - where all the links work.
    • We need a business process as well as a set of tools to make it real. Some people recognize the need, some do not.
    • There is a number of people who just want to get their job done, and not be concerned about the enterprise.
  • Tutorials need some form of validation
    • e.G. Security COnsiderations Cookbook is tutorial with very "Authoritative" language

MOTION: Approve Jane Curry as Publishing Facilitator: Unanimous

• • Arb to ensure the definition of what constitutes "Substantive change" is done
    • Woody suggested we delegate and include more than one other WG (MnM + ??)
    • Should include versioning/sunsetting issue as well.
    • Was a role of the old ARB.
    • Ended up clarifying rules to prevent ARB from being court of appeals for Substantive Change.
    • Original document defined substantive change as anything that breaks wire format
    • With the advent of V3 xml change in wire format was not necessarily substantive.
    • Grahame Grieve: Arb should specify a statement of value - what is substantive change at the highest level. Then create a task force to generate rules for a specific specification set.
    • John Koisch: Each specification could provide it's own substantive change rules.

Grahame's Definition: Substantive change is any change in a specification that causes changes to conformant implementations, except in as much the implementation leverages technical publication artifacts directly.

• • Replace Services with Service (change plural to singular).

MOTION: Change the S in SAIF to singular. Unanimous approval

• • SAIF transition from ALPHA to BETA+
    • Calvin Beebe's presentation on application of governance.
    • Suggested that we pick a probblem, apply SAIF end-to-end, and refine:
      • Governance
      • Implementation Guide
      • Methodology
    • Charlie Mead will discuss experience at NCI with TSC - around four/five points:

1. Value propositon
2. Have a clearly defined scope to apply SAIF
3. Come to terms with defining artifacts and meta-data
4. Governance must recognize that you cant govern everything - look at the life cycle, and the risk of not governing certain aspects.
5. There must be a notion of change management recognition.

• • • We need an architecture informed by SAIF
    • What are the working elements of SAIF that apply regardless of messaging, documents, and service metaphors and determine how to govern "that"

• • Set direction for the week.
  • Arb Membership Criteria Review
  • SAIF Glossary Review

Sunday Q2

• Sunday Q2
  • Call to order
  • Approval of Agenda for quarter
  • Behavioural Framework (John Koisch)
    • Summing up discussions - status report
• History

1. Started with HL7 dynamic model
2. Services Oriented Dynamic Model
3. ArB SAEAF conformance stack
4. The Behavioral Framework - adoption of requirements for messaging and services

• Work Streams

1. OO Dynamic Model - Analysis model
2. MnM & ArB Discussions
3. InM + SOA + OO Discussions - Gone but the actors remain
4. Implementation Guides (From NCI, etc)
5. Services for RIMBAA - fringes of BF discussion on how BF enforms RIMBAA
6. Patient Registration - Detailed analysis - when patient reg was broken into services, it was not one-for-one as expected.

• Comments and Reconciliation
  • Three types of comments

1. Content-driven (Definitions and terminology)

2. HL7 Strategic (Does the BF buy HL7 things in terms of SAIF, or outside of SAIF)

3. HL7 Engineering "People do not need visibility of the BF ...they need the imputs to MIF"

• • Realizations
    • BF isw coherent, but that stands in the way of acceptance
      • It talks about traceability for deployed interfaces

It binds RM-ODP to create a vocabulary and grammar for discussing behavioural models It is generic with respect to tooling and not necessarily HL7 Specific It shifts the focus of HL7 givens from information focused to specification focused Doesnt include concrete instantiation(OO not working on it)

• • Huge Question
    • Does HL7 care about services
      • Does HL7 care bout workflow resolution in the standard?

Jane Curry: Workflow is important - but is it bottom-up, or top-down? You can from the bottom identify the states, and the precedence, but you cannot identify the authority - which is non-standardizable. governance problem.

Ron Parker: If we can get people to adopt this behavior, can we change the toolsets.

Jane Curry: We have a long-transaction. Admit a patient for an encounter is a huge payload.

John Koisch: HL7 Does care, but the model wont support it.

John Koisch: Is HL7 prepared to think in terms of explicit, but extensible, contracts or are we coing to have focus solely with the information components?"

Grahame Grieve: HL7 is uniquely focused - most others provide working standards, while we often provide frameworks. At what point is HL7 a horizontal provider of standards, is it a wise thing for us to split into two levels?

Charlie Mead: HL7 implementing an HL7 SAIF IG as the authoritative model will inhibit industry adoption. Ken's letter to the board recognized that we need the canonical model.

Ron Parker: I am just now getting to the point to where I can talk to my clinical community about doing SOA. They have to get organized around the concepts. HL7 has not thought yet about what it means to create a framework for describing workflow. We put out a workflow and the vendors told us to stop - and it evolved to discussions on how the workflow really happened.

John Koisch: I dont think we can determine all of the pieces ahead of time. I cna show workflow specs from decomposition into services. The intent of what Ken was saying is that we need both - top-down, bottom-up. People are balking because it is not worked out. A specification for referral will have implicit boundaries, which you may not agree. Unless you have an energized community, throwing out the use cases, we will not go ahead. Where is the energy? HL7 focused on information is great. The framework is more valuable to us than V3 messages.

Wendell Occasio: There are some things in the MIF that are canonical - we need to spell these out.

Jane Curry: We have 'new renames' in the dynamic model in the MIF. There is some additional concrete-ness that is missing.

John Koisch: The context that MIF puts the dynamic model:

• MIF-ixed version of BF
  • Theis requires that we look at the BF as pieces
    • Behavioual modeling peices, would become contract framework
    • Specification Pieces
    • Vocabulary/Grammar Pieces
• Continue th OO DAM effrorst to support reconcilation
• More implementations and implementation guides
• We need to start balloting, enve in draft state, SAIF specifications
  • There is a lot of material that cn be reconstitute to be SAIF compliant
  • Not doing this is holding back a number of good standards and approaches to problems

Jane Curry: explained history of the current messaging interactions. The OO model is broken because it is dependent on long transactions.

Ron Parker: ArB is in a critical role in terms of solving this in terms of things we say and actions we take. People are beginning to get it - but it will require the patience of JOB before we see it happening. People are beginning to understand - need for discrete services, organizing to solve the problem. The absence is the way to go do it. What do i need to be able to do? What does patient admin need to get to solve the workflow problems?

John Koisch: The pushback is we dont have tooling - it is industry standard, not HL7 standard.

Jane Curry: If we could show that we would push it forward.

Steve Hufnagel: EHR-FM has dealt with this - the should/shall/may. Profiles allow the constaint of the should/shall/may. Orchestration of services provides workflow. Use the ECCF to specify the profile.

Jane Curry: Does the EHRS-FM come into play ?

Charlie Mead: Yes

Steve Hufnagel: Yes, at MHS, not yet to HL7.

Jane Curry: That is the key - the functional profile identifies the expectation.

Ron Parker: Functions available.

Grahame Grieve: "Not doing this is holding back a number of good standards and approaches to problems" we have a group of people, but no funding to do so.

Ron Parker: IG may speak to the 'just do it'.

John Koisch: At NCI we are doing it - not at the rigor required for ballot. There is a resourcing issue. Tuesday night I will show what we are doing at the NCI. To get it to a level that responds to the questions that come out of HL7 is hard. Funding the MIF-ized version of BF is finite work, but has strategic pieces that will have to be adjudicated.

Jane Curry: There is a governance problem.

Ron Parker: Lets see the outcome of TSC and Co-chairs meeting. The Alpha projects took on a very not-ripe set of things, which informed us. They did get benefits.

• • • • • You cannot talk about behavior unless you are willing to talk about decomposing information
        • What is being standardized WILL change; There WILL be more STANDARDIZED PIECES
      • I Do not believe that we can determine all of the pieces that make SAIF or the BF work ahead of time - this is a proactive

Initiate Reponse to TSC on "Substantive change

Grahame Grieve: Arb should specify a statement of value - what is substantive change at the highest level. Then create a task force to generate rules for a specific specification set.

John Koisch: Each specification could provide it's own substantive change rules.

Grahame's Definition: Substantive change is any change in a specification that causes changes to conformant implementations, except in as much the implementation leverages technical publication artifacts directly.

Charlie Mead: Four Steps:

1. Develop during this meeting a candidate value statement on scope of substantive change to include backward compatibility, versioning, and sunsetting (ArB Project)
2. Circulate to co-chairs to ask for participation
3. The task force will take this on to report back to TSC
4. TSC will operationalize it (TSC project)

Motion: Charlie to take to TSC. (Wendell/Grahame) Unanimous

Glossary Review

No one will admit to having had time to review the glossary.

Wendell Ocasio: Many of the terms in the glossary are not defined in the way they are used.

Jane Curry: We define terms in context, and dont apply qualifiers. Quarter adjourned at 12:30pm U.S. EDT.

Sunday Q3

• Call to order
• Approval of Agenda for quarter
• Information Framework and IFIG (Cecil Lynch)
  • Must have a data model
  • Must have data-types
  • Must bind vocabulary

• Language for "substantive change"

(We discussed this for 40 minutes) Charlie will take the following to the TSC:

Substantive change is any change in a specification that causes changes to conformant implementations.

• SAIF Implementation Tutorial Preview
• Adjourned 3:30pm U.S. EDT.

Tony Julian 19:26, 3 October 2010 (UTC)

Meeting Outcomes

Actions (Include Owner, Action Item, and due date) .

Next Meeting/Preliminary Agenda Items .

Tony Julian 03:17, 3 October 2010 (UTC)