AID 201501 Agenda
This is the agenda of the AID HL7 'user group' for the WGM in Santionio (Jan.2015)
Contents
Sunday Q3/Q4 (hosting FHIR)
- This session offers those that have implemented FHIR an opportunity to share thgeir approach and best practices with other implementers.
- Most of the presentations will only be added to the agenda during the actual connectathon
- Add your session title here
Monday Q3
- Administrative
- Approval of the minutes of the Amsterdam out-of-cycle meeting held in November 2014, available at AID 201411 Agenda.
Tuesday Q6 (hosted by EST)
- Tooling demo, 7-9 pm
- See agenda of EST for details.
Wednesday Q4
- Review/update Tool Evaluation Criteria
- Implementation Packages
- According to the PSS created last September the project should be finalized before the end of January, with as its deliverable the definition of the content of the HL7v2 as well as the CDA implementation package.
- Status update:
- Whitepaper review: List of AID Whitepapers
Minutes
Minutes can be found here
AID Wed Q4 We did not meet quorum. Present were Peter Hendler and Michael Van der Zel.
We reviewed the “Tool Evaluation Criteria” and feel it is still valid.
We discussed three White Papers.
The one on “Implementation of IVS and TS” is still valid but has decreased relevance as more and more effort is put into FHIR and less into RIM. We should leave it.
The paper on Green CDA is likewise valid but less relevant as we don’t see much activity going to Green CDA and all the activity seems to be targeted in how to express CDA as FHIR and visa versa.
The one on “Software Implementation of CDA” is still relevant and valid but could use an update to include ART DECOR.
The previous group RIMBAA spent a lot of effort exploring ways to persist RIM. It seems that today, a new effort could be initiated to document the various approaches being taken to persist FHIR. This was also mentioned at our previous quarter by Andy.
The new effort to express FHIR in RDF is another area ripe for attention. In fact, there is a natural merging of these two concerns if we were to look at RDF triple stores as a way to persist FHIR.
We then looked at an example OWL ontology that was created by an OO modeler who was not trained in “open world” or OWL. It was very interesting because the errors were almost predicable if one were to use Protege as if it were Enterprise Architect. For example, there was a Class “Organization” that has a “child member” called “Location”. This makes perfect sense in the OO world as Organizations do have Locations. But in the Ontology world what this was actually saying to the reasoner is this: If you see an instance of Location, then you must conclude that it is a sub type of Organization. Clearly not the intent. This points out the dangers of adopting RDF to HL7 without very specific training of modelers in the differences between OO and OWL/RDF. This might also be a subject for AID to focus on.
Finally we looked at the CIMI Reference Model and it is evolving to include Entities in Roles that Participate in Acts. And Acts that are related to other Acts. In some cases the names of these things are changed but the structure is clearly there.
