20100225 arb minutes
Architecture review Board Meeting Minutes
February 25, 2010
Tony Julian, Jane Curry, Ron Parker, Wendell Ocasio, Abdul Malik Shakir, Andy Bond, Jane Curry, Cecil Lynch Guests: Russ Hamm, Tony Weida, Cliff Thomas, Karen Smith Architecture review Board Meeting Minutes
February 25, 2010
- Call to order
- Approval of agenda
- Approval of the minutes
- Review of Security Ontology project
- TSC discussion
- Review of SAIF
- Review of Task force Statement
- Review of GF
- Review of Decision Making Practices?
- Action Items
- Other business and planning for next meeting
Call to order
The meeting was called to order at 6:05pm U.S. Eastern with Ron Parker as Chair and Tony Julian as scribe..
Approval of agenda
Approval of the minutes
MMS to approve minutes February 18, 2010 Minutes(Jane/Tony) (7-0-0)
Review of Security Ontology project
Ron parker: Have shipped the correct document.
Cecil: My comments were not changed based on the updated statement. I understand the need for having formal ontology to organize information. I am struggling with how you intend to use clinical and biomedical ontology to hook security and Privacy. #3 success criteria: When mapping to a clinical ontology has not been defined here. I have only had partial responses on how 4b can be achieved, what value it brings, and how to do by November 2010.
Tony Weida: Members of my consulting group are participants. I have not developed proposal, schedule, or details. Relationship between security/privacy and clinical. Control of the disposition of medical information based on subject matter, and regulations about substance abuse and STD information. Needs to give providers control over the revelation of same. Dispositions about collection, access, disclosure, and the security used. makes sense to use SNOMED to provide context for governance makes sense.
Tony Weida: It is not a non-trival excercise. We are exploring the indentification of sensitive areas of SNOMED. There are chalenges to using, because of the different levels of sensitivity. There are items that might indicate presence. One thing is to have terms you can use to talk about likely levels of concern - high level concepts, but the next step is enumerating the various concepts that would be indicative of the subject matter under concern.
Cecil Lynch: Yes, it is not a trivial matter. I have already done this . . . for HIV issues you have to look across tons of orgamisms, not only HIV but parasitic organisms, brain diseases, and does not include all of the medications associated, so you can do clinical decision support.
Tony Weida: This proposal is not proposing to provide those definitions. Just where appropriate being able to tie in to areas of concern in a general way.
Ron Parker: Our concern is timeline, complexity, and participation, as well as presumptions of the final product. ArB primary concern should be looking for a way for ANY HL7 participant to see a framework for linking, defining the appropriate method. That is one of the things we are most interesed in, the most useful artifact.
Cecil Lynch: When you look at this there is no clear roadmap on how you are going to do this. E.G. what is a recoginized ontology? You have information viewpoint, where you tie the governance viewpoint, as well as the engineering how to make it work. The Scope statemnt should reflect your success criteria. You cannot do that by November 2010. Author has not recognized the lifecyle defined in SAEAF.
Tony Weida: I would not have a problem with defering the ontology mapping.
Ron Parker: Tony and Cecil, this is good work and is really important, will be a piece of heavy lifting. Our primary interest is defining the scope, understanding that it will be a learning curve for all of us. I dont want to be too constraining.
Cecil Lynch: I am concerned about the resources being applied, I understand where they are trying to go, but the resources needed need to be mapped out. You cannot apply resources if you do not have clear objectives. You will have difficulty engaging the appropriate groups. The biggest issues I have are clear objectives - need, objectives, success criteria do not align. I was hoping Mike would be here. People who have proposed are not clear.
Wendell Ocasio: How does it work with the SAIF?
Ron Parker: TSC says ArB has ability to comment on the project, we are not more vital than say the Foundation and Technology Steering Division - it is up to FTSD and TSC to approve. As a group we need to sayw what would better define the project.
Cecil Lynch: Role of ArB should be to look at the outcomes,think about the deliverables, and how they fit HL7's needs and requirements. This delivers an ontology, how do we incorporate into HL7, is there RIM binding? Our job is architectural integrity - we have something new that we cannot fit into the architecture. I have tried to talk to someone to give me insght into the deliverable fit into our architecture.
Ron Parker: What is our go-forward? We need more dialog.
Tony Weida: It is clear that Mike needs to be part of the conversation, and how he see the artifacts from the project tying into the RIM and other artifacts. I would not be looking for this to hook into the RIM, but to inform work that occurs later on. My Perspective.
Cecil Lynch: We dont have a place to put this standard as a DSTU, or to evaluate it. Should this be in HL7? Should it be DSTU or informative document? If DSTU where does it fit in? You are leading the way down a path we(HL7) should be going. As the first project you need to meet a higher bar.
Abdul-Malik Shakir: We think that the goals are worthwhile. Are the timeframes realistic, do we have the right people? I would be comfortable with approving with the understanding that there would be a checkpoint after the elaboration before it is constructed, or goes to ballot.
Jane Curry: If the success includes how to apply in othere environments, as well as applying standards policy - when you speak of clinical decision support, and not the other artifacts, there would be a mis-match of intent. It should move forward, but we need to work together to discover impact on other HL7 artifacts. We dont know what the form of the normative document would look like.
Cecil Lynch: How do we apply ECCF?
Ron Parker: Need to have ArB member involved - need to be helpful in shaping it. We are supportive, Cecil is excited about it. Tony Weida if you could put on the agenda of the next meeting. Cecil you need to hear the group work their way through it.
Tony Weida: I will pass to Mike.
Ron Parker: We have a strong concern, and TSC is interested in ArB decision. Thaks Tony, and Thanks Cecil for you work
Cecil Lynch: We (HL7) have to do that.
- TSC comments.
- Issue of protocol on the precidence of review came up.
Ron Parker: We had a sense of hurry-up, but the TSC call discused that it had not gone to FTSD yet. It was not expected on the TSC agenda. What was not a problem was that we had material: at any point ArB could pick up and run with it. but should wait until FTSD has commented. We will be more careful about when things appear to the TSC.
Ron Parker: Not vote is necessary.
Review of Task force Statement
Review of GF
- Managing between Jurisdictions
Jane Curry: What is coherence, and from whose perspective? The ArB thinks that SAIF will allow HL7 to produce coherent standards. The implementers by implementing standards produced by HL7 will have implemented consistent interoperability across specification.
Jane Curry: The problem is multiple jursidictions. We already understand levels of profiles, which leads to specifications that meet local needs, and still apply to global standards. You can choose incompatible derivative product - when you have a regulator (legislative, insurance, etc) who is not aware of what others have to do. Developers will get contradictory instructions - different interactions, vocabularies, policies. There needs to be sufficient discussion outside the HL7 and Implementation community to define the overarching regulatory environment. Some of the feedback on the last version of the GF, there was a risk of not being able to achieve interoperability because of mis-alignment of the external environment.
Jane Curry: Should I bring out in the framework? It is implied in the ECCF - I could make it more implicit in the governance section.
Cliff Thompson: It is required. Policies live within healthcare as well as outside, there is always a conflict. You cannot always derive back to a single set of governance instructions. We should formalize the traceability, and document the compromises.
Jane Curry: Reinforces my point of view - it is essential, or we are ducking the problem. Implementors make an assumption that it is easy, when the conversation does not occur at the necessary regulatory level.
Ron Parker: Implementation guide should discuss. Should inform the service contract specification.
Jane Curry: RM-ODP makes assumptions within a closed paradigm.
Ron Parker: We have precidence in service specifications. We need to work through the artifacts, and this dimension needs to be visited.
Jane Curry: I have tried to do so - you can decide on the next round.
Review of SAIF
Ron Parker: Line numbered documents are available.
Next call = Review the spreadsheet with the required reviewers. Once we have agreed on the list, we will start the official process. Will take up on next call, with a target to release for review on Friday.
Other business and planning for next meeting
- Wiki review
- Review the spreadsheet with the required reviewers
- Task force Statement