[FOSS_health] summary - next step

[David Forslund]

Lee Seldon wrote:
> Thanks for the useful feedback. Here is the next version of a very very
> short summary:
> What type of interoperability do we want / can we achieve?
>   - FOSS should probably aim for 'functional' interoperability, but
> might have to start with a 'basic' one.
> Can we agree on a message format?
Picking a message format already then picks a particular communication 
technology.
>   - communications should preferably be based on XML, not on HL7 v2.x.
>
> What next?
> There are several options. It would be very useful if somebody (with
> time and/or knowledge) could write up a comparison, or point us to an
> existing one. Something like (and we can improve on this)
COAS/CORBAmed/HDTF are the same thing, although there are some outstanding
RFPs in the OMG HDTF today that will "upgrade" COAS to RLUS.
>
>                  How it     | Features | How to     | Standards   | 
> Advantages | Disad-
>                      works  |               | implement | compliance 
> |                    | vantages
> ---------------------------------------------------------------------
> CCR
> HL7 CDA
> HL7 CCD
> HL7 v3 msg
> MIRTH
> CORBAmed
> HSSP
> HDTF
> COAS     distributed objects    | flexible data |  Any language, any 
> database | It is a standard|  OSS implementation available | low adoption
> www.tolven.org
> other suggestions
> ??
>
> (I don't know if everybody watching this list has HTML email displays. 
> That reminds me, is anybody from openMRS watching? They should be 
> playing a role.)
>
> It would be good if we all understood all of the possibilities before
> trying to agree on one. There are some which I do not yet understand.
>
> Licence fees might come under Standards Compliance or Disadvantages. 
> E.g., I bought a CCR licence (copy of the xsd file), but I don't 
> remember if I am allowed to distribute it further. Ditto for SNOMED-CT 
> (which I am only allowed to use for research). On general principles I 
> personally would try to avoid anything which is not free or not 
> available in developing countries. (That leaves vocabularies ICPC2, 
> LOINC, UMLS ...)
>
> It seems to me that everybody will have to make a mapping at some point,
> unless your data structure already fits a standard, or unless all 
> recipients have the same data structure.
> Lee
> _______________________________________________
> FOSS_health mailing list
> FOSS_health at oshca.org
> http://mailman.oshca.org/mailman/listinfo.cgi/foss_health
>
>