20101005 arb cambridge minutes
ArB Minutes Template
Meeting Information
| HL7 ArB Work Group Meeting Minutes Location: Cambridge, MA |
Date: 20101005 Time: 3:30pm U.S. EDT | |||||
| Facilitator | Jane Curry | Note taker(s) | Tony Julian | |||
| Attendee | Name | Affiliation | ||||
| . | Bond,Andy | NEHTA | ||||
| X | Curry, Jane | Health Information Strategies | ||||
| . | 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 | ||||
| . | Mead, Charlie | National Cancer Institute | ||||
| . | Nelson, Dale | II4SM | ||||
| . | Ocasio, Wendell | Agilex Technologies | ||||
| . | Parker, Ron | CA Infoway | ||||
| . | Quinn, John | Health Level Seven, Inc. | ||||
| . | Shakir, Abdul-Malik | Shakir Consulting | ||||
| . | Guests | |||||
| X | Haddorff, Richard | Project Services | ||||
| X | Koehn, Marc | Gordon point Informatics LTD | ||||
| X | Krynicky, Leah | Project Services | ||||
| X | Singureanu, Ioana | Project Services | ||||
| Quorum Requirements Met: No - no chair. | ||||||
Agenda
Agenda Topics
- Call to order
- Agenda review and approval
- Safe and Sound discussion
- Adjournment
Minutes
Minutes/Conclusions Reached:
Jane Curry assumed chair, and conducted introductions.
Rick Haddorff: Does the IG document cross use-cases?
Jane Curry: Lifecycle documents the progression of a spec through the requirements, implementation. We also have coherence across use-cases. It is usually reflected to the organization through ballot. There are more processes that need to emerge, that are now being surfaced informally. Pharmacy as well as security and privacy, notice that things are discordant - but not intentionally.
Rick Haddorff: The SAIF implementation guide will address, but looking for other groups to provide something.
Jane Curry: The governance framework should provide pattern of governance, the IG should be developed by organization, such as project services.
Marc Koehn: Ultimately every group has to ask themselves - have we accounted for SAIF in process and documentation? There is a set of open questions: forgetting the current situation, between the creation of a standard, and SAIF, there are processes/artifacts for which we need to account.
Rick Haddorff: Would you see project services developing the IG?
Marc Koehn: You would be the first candidates for drafting the process part.
Jane Curry: We are too small with the resources we have, and need to engage more deeply with others, so the IG is not the responsibility of the arb, except as oversight.
Rick Haddorff: That was the take I got from Ron Parker.
Jane Curry: Governance is who/what/when and who enforces. In V3 we enforce rules through the tools. Other is enforced by the project service lifecycle. Pulling artifacts into a specification, you have to describe the traceability and consistency. We don’t have a clear statement of shared purpose, which is needed at the beginning of any specification - not same as project scope statement.
Marc Koehn: Part of SAIF takes place of Project scope statement?
Jane Curry: Statement of shared purpose - defines the success criteria. A lot of governance works outside the automation boundary. e.g. You must have a Notification of Intent to Ballot, before the tooling will allow you to go forward.
Rick Haddorff: The point I heard about people reacting against governance, requires people know their roles, and respect them.
Jane Curry: Inspection at the end of the project is bad, quality should be validated through the process. Tools for publishing are supposed to ensure well-formed-ness, which is automatable, where content is not. There are opportunities for clarifying the roles, of work groups and their artifacts, which are not as they should be.
Marc Koehn: There will always be that grey area, where people will push the rules. We know the ballot is not the best way to catch issues, too much energy involved. The decisions should be applied before then. The project services power-point is a great start. If we can see that all the artifacts are reflected, then we can socialize.
Jane Curry: It is an opportunity for alignment between EHRS profiles and the suites - the cross references. It is currently inferred. In RIM-Speak, the artifacts, the domain analysis model, needs to be defined at that level, as well as their linkages. The explicitness of the linkages across domains is missing in the HDF.
Rick Haddorff: Feedback we got at Rio was it was too soon to talk about artifacts from SAIF. Are we doing it too fast?
Jane Curry: Difference between safe and trial.
Marc Koehn: If we demand we will get pushback, but describing the journeys for artifacts is better. Is ArB seeing more people willing to develop the items?
Jane Curry: We don’t have criteria for a domain analysis model as an information exchange structure, rather than as a information structure. If we did everything as draft, and talked to those who tried, we could mark artifacts as more mature than others. An evolutionary way - vocab is mature, changes incorporated, and mechanisms exist. Those might serve a linkage between artifacts that you do. The difference between proposed and produced. It was good draft material - I used it thinking through the problem space.
Rick Haddorff: I hear we should go forward with the whitepaper, and bring out.
Jane Curry: I would like to see it bounced off in idea state with each of the different workgroups responsible for foundation elements, so not just the ArB is exercising it - the groups who have ownership around an artifacts use and life cycle. MnM, Vocab, SOA, EHR, ICT, ITS, Security - not supposed to leave them out as after thoughts. A lot of ECCF is how to create a specification derived from the artifacts. ECCF if focused on a specification, not governance.
Marc Koehn: John, Cecil, Tony, you are comfortable with the project services?
Cecil Lynch: Don’t know enough about it.
Jane Curry: Rio presentation is core.
Rick Haddorff: Of whitepaper.
Cecil Lynch: I have to review.
Marc Koehn: Is project services going in the right direction, and are they the right people? I guess is is.
Jane Curry: They are documenting.
Marc Koehn: They are moving to faster balloting.
Rick Haddorff: We historically have documented the existing processes.
Jane Curry: Is there anything specific we should discuss.
Rich: We have not been concentrating on it. It is on our agenda for Thursday.
Jane Curry: What is your plan to present? If we don’t say anything different, to whom, and when.
Rick Haddorff: We will start to work on it at our conference calls, then develop the plan to present it. We have been working on the project scope statement, instead of the project.
Jane Curry: Have you achieved sufficient guidance, or do you need more engagement.
Rick Haddorff: I feel comfortable about the whitepaper, but murky about the IG, and how ArB wants project services to engage in that area.
Ioana Singureanu: Since the SAIF was introduced, ECCF was updated twice. There are questions that need to be addressed.
Anthony Julian: Glossary needs to be harmonized with the glossary - there is a project by Vocab, so we should get involved.
Jane Curry: The SAIF IG will merge with the HDF.
Cecil Lynch: I don’t agree with the approach - one of the issues giving people difficulty in aligning the artifacts is the confusion over taking a DAM to a DIM, has no measure where the cutoff is. A reliable way is can I take this from a PIM to a PISM in a single level. If my PIM is serializable, it is a PIM, but if not, it is still a CIM.
Jane Curry: A quality criteria we have not defined.
Cecil Lynch: Do i have an entry point that I can serializable.
Ioana Singureanu: Conceptual model may not be traversable.
Cecil Lynch: If more than one entry point, then not serializable e.g. DMIM is not, RMIM is. Each DAM has multiple layers. There are layers to a PIM, which are used to form a PISM. We need to measure conformance, and if a PIM looks like a CIM, we cannot use it to claim conformance.
Rick Haddorff: The SAIF implementation guide will find where the artifacts fit in the stack.
Cecil Lynch: Vocab elements are bound at the RIM, PIM, and PISM. That is why serializability makes sense. That is the detail we need.
Marc Koehn: We must separate the steps. Should these guys continue, and further define the work.
Cecil Lynch: The process is there - i could not make it work yet.
Jane Curry: A RASCII chart for each work group, showing responsibilities, as well as processes, will help us. In harmonization we have missing people.
Marc Koehn: Are Cecil's issues captured, to the point where the paper can proceed?
Cecil Lynch: At the service layer, John will need to comment.
John Koisch: The problem with putting into a methodology is what are you trying to specify, where we are with the MIF, where the tool forces you into levels of granularity, you need to leave it for people to express at their granularity.
Cecil Lynch: The slide set is pretty, with little detail. The whitepaper will be good to see, when it is ready to share.
Jane Curry: Even pieces, then we can be engaged, and provide ongoing guidance, so that by the time it is published, everyone will be ready.
Rich: Need to plan the paper, steps, and audience.
Jane Curry: TSC is standing behind cross-workgroup efforts. e.g. CDA-R3 and composite orders. Need agreement by participants with the process. Arb can be engaged as we flesh out the frameworks. Supposed to be engaging with the groups.
John Koisch: It is easy to draw a PowerPoint, and think you have it, but it requires iteration to find the best.
Jane Curry: I agree. Because we are all working on less-than-mature documents that are interdependent.
Cecil Lynch: Marc's suggestion that we have working meetings face-to-face, we can get cohesion.
Jane Curry: How many people will be in Australia? We don’t have to wait until January. Have a 2 hour teleconference, and don’t wait for face-to-face.
Marc Koehn: The nuances you brought forth are necessary.
Cecil Lynch: It is granularity as well as business requirements.
Jane Curry: A statement of purpose is a scope management issue, with 17 people you have 18 ideas of the scope. You need finite requirements. By the end of October I am getting close to the CF. I did a whole new view, shared information, I did not have before.
Jane Curry: Reflect it back and watch the space on the ArB and Project Services.
Marc Koehn: Should chairs get a date/time to setup the two calls? To verify the direction?
John Koisch: Yes
Jane Curry: We will commit to coordinate two calls between now and Sydney. On Thursday we need to change our call time.
Marc Koehn: Not just joint ideas, but make sure the artifacts are properly reflected. Make sure the intent is well reflected, and make sense. Allow change in the cycle. You want to lock in the requirements/intent early, and the decision making reflects that.
John Koisch: The expression of the methodology is linear. it is easy to say this is just our expression, but the SAIF is a bag of artifacts. There is a notion that you will ballot conceptual spec, pim, pism separately. We need a way to clarify a way that it is a ballot lifecycle, not a development lifecycle.
Jane Curry: Consistent with the agile approach, not the waterfall approach.
Marc Koehn: If we dont solve that, why are we doing this. The early decisions don’t follow the lifecycle. We should be harsher. Allow flexibility where needed.
Jane Curry: Measure of success is coherence within as well as across. If our specs are perfect, but in conflect, we are in trouble. It is not dictation, but explanation of where they are not coherence.
Cecil Lynch: it is explicit in information framework- it separates the model levels.
Jane Curry: Also for behavioral?
Cecil Lynch: You have the same issue but is it harder to delineate where a stakeholders view cuts off. Keith Boone looked at the ECCF, and drew a stakeholder map, to keep in mind as you are trying to determine which master you are trying to serve
Jane Curry: Some roles can be delegated to systems, but many must be done by bodies - we loose business purpose when we drop to the implementation - whether service, messages, or documents. We are going to affect the GOM again.
Jane Curry: We are using ballot process as a communication process. We need to communicate earlier. We do not have tools yet.
- Adjourned 5:00pm U.S. EDT
Meeting Outcomes
Actions (Include Owner, Action Item, and due date)
|
Next Meeting/Preliminary Agenda Items
|
