AID 201510 Meeting in Atlanta
Jump to navigation Jump to search
This is the agenda of the AID HL7 'user group' for the WGM in Atlanta (October 2015)
- Project and issue review
Sunday Q3/Q4 (hosting FHIR)
- Rene Spronk, and 40 others (40 in Q3, 24 in Q4)
- Rene: as per our DMP I'm asking for your permission to record and publish video of the various presentations. No objections were raised.
- Simple JSON Parser Mobile App (Jeff Ting, Systems Made Simple, US)
- I wrote a mobile app for the first time, and one that can parse JSON to any level of depth to reach FHIR elements of interest. I will discuss the parsing design, as well as lessons learned on user interface layouts for mobile.
- Comment by James Agnew: could also use HAPI, it has an Android library.
- General FHIR Server Design Review (Brian Postlethwaite, Healthconnex, AU)
- Covers different types and approaches for implementing a FHIR server (Generic FHIR Server, Publish/Sync, Façade, Cache)
- Aims to raise discussion about these various approaches
- Comment by Ewout: may lead to situation whereby primary key (as used internal to the persistence layer) is publicly exposed as the resource.id.
- Question by Rene: excellent overview, IMHO should be used as a starter version for one of our whitepapers. does this room also feel that to be the case? Yes. Also a whitepaper on the v2-FHIR mapping issues.
- clinFHIR implementation experiences (David Hay, Orion Health, NZ)
- See http://clinfhir.com/ for the clinFHIR site.
- FHIR Enable Epic EHR using Cloverleaf (Ron Archambault, Infor, US)
- Infor developed FHIR DSTU-2 Message bundles with patient and questionnaire resources using the Epic InterConnect API
- Development of "FHIR cell for i2b2" (Kavi Wagholikar MD, MGH, US)
- This presentation will describe the technical details and learnings in the development of the fhir interface for i2b2.
- Over 100 institutions in the US have i2b2 installations for identifying research cohorts. The fhir api is aimed at providing a readonly interface to the data hosted in the i2b2 instances, allowing the institutions to run SMART apps over i2b2.
- Executes a query agains an in-memory bundle of resources.
- HL7 ORM v2.3 into FHIR REST-API (Joerg-Uwe Meyer, MT2IT, DE)
- Migration/translation issue v2-FHIR
- Creating the FHIR interface for the NHS FGM Risk Identification System (Richard Kavanagh, HSCIC, England)
- An overview of the journey of implementing FHIR into the English NHS national architecture (see http://data.developer.nhs.uk/fhir/fgm/ ). The challenges of moving away from HL7v3, dealing with a DSTU standard and the absence of tooling infrastructure. Lessons learned and the roadmap for the future.
- Challenge will be 'using FHIR at scale'.
- Adjournement at 17:00
- Peter Hendler, KP
- Dmytro Rud, Roche Diagnostics
- Jason Walonski, Mitre
- Marc Hadley, Mitre
- Brian pech, KP
- Justin Fyfe, Mohawk College
- Rene Spronk, Ringholm
- Michael van der Zel, UMCG
- Dale Nelson,
- Nancy Orvis,
- Rene: as per our DMP I'd like to request permission to record the presentation. No objections were raised.
- Approval of the minutes of the May WGM in Paris, available at AID 201505 Meeting in Paris
- MOTION to approve the minutes of the May WGM in Paris, available at AID 201505 Meeting in Paris (Peter/Michael, 2-0-5 (Y/N/Abst)).
- Use of a RIM-lite OWL triple store to calculate quality measures (Peter Hendler, KP)
- Peter has been working with Oxford University UK dept of Computer Science and KP National Quality. They have an OWL ontology based on Entities in Roles Participating in Acts (ERPA; it is not the full RIM by any means, it is their own RIM lite OWL). They have successfully calculated some Diabetes quality measures (HEDIS, Healthcare Effectiveness Data and Information Set) using OWL, RDF and SNOMED.
- Peter will discuss some of the core design considerations, as well as other experiences.
- See http://www.cs.ox.ac.uk/isg/tools/RDFox/ for details of RDFox, a highly scalable in-memory RDF triple store.
- Peter: we're using static data, reporting is based on the 'last year'. Could be done 'live' if so desired.
- Marc H: can't we do this with SPARQL ? Peter: No. SPARQL is for queries, this creates new knowledge, new triples of truth. But someone from Oxford would probably be able to explain this is a better way.