At 02:31 PM 11/17/99 -0500, Gunther Schadow wrote:
>Wayne Wilson wrote:
>
> > > Response from Sam some weeks ago:
> > > >
> > >> GEHR will easily accept all HL7 messages (the GEHR
> > > > model is more
> > > > comprehensive than the RIM). Applications working with the
> > > > Ocean GEHR kernel
> > > > will be able to store all HL7 messages 'anatomically'.
> > > >
> > And previously Dave Forslund wrote:
> > > It became clear to all that there is enormous enormous synergy
> > > between those efforts, and that the MOF and could generate XML
> > > through the XMI standards. In addition, it is clear that the PRA
> > > could be express quite nicely through the COAS service.
>
>and here Wayne carries on:
> > I will not claim to be an expert in the modeling areas of either HL7 or
> > GEHR, but I can't help but feel that ability of COAS to represent HL7
> > PRA is directly due to the influence of the GEHR model upon design of
> > COAS. I know this to be case as I was there during the design
> > discussion between the parties. I also can't help but believe that XML
> > via XMI and a proper meta architecture ala the MOF allows one to do
> > anything.
>
>Again: slowly, slowly! This sentence is full of beliefs that seem to
>not hold a closer look: COAS can not represent the HL7 PRA because
>COAS is observation structure oriented whereas PRA is document oriented.
I don't believe this is true. COAS can represent essentially any XML
document and
allow for retrieval of parts of it, much like the XML servers that are on
the market today.
>They have quite little in common -- of course you can bridge between
>any two architectures, but that doesn't count. The second belief is
>based on Sam's statement above that "GEHR can easily accept all HL7
>messages since GEHR is more comprehensive than the HL7 RIM." What?
>First of all, existing HL7 messages (v2) are not based on the RIM,
>second GEHR can not sensibly get into an HL7 order transaction, can't
>do HL7 based prescriptions, can't do clinical trials, can't do
>master file transactions, etc. Then there is this statement that
>GEHR is more "comprehensive" than the HL7 RIM -- is not true either.
>Again, where does GEHR allow you to place orders, medications? How
>are care plans and guidelines represented in GEHR? What about
>longitudinal problem lists in GEHR?
>
>I can see the similarity of GEHR and COAS, but that's about all
>that makes sense to me in this comparison/synergy discussion. The
>problem with both COAS and GEHR is that they only deal with
>observations but not with the big picture of health care, which
>kind of should include interventions, problem lists, data dictionary
>management, ADT, etc.
Data dictionary management is handled largely by TQS. Certainly more
capability is required than COAS currently represents and this is being
worked on.
I still believe there is great synergy in work going on between HL7 and
CORBAmed, because
HL7 is dealing primarily (but certainly not exclusively) with the
information model of RM-ODP
and CORBAmed is dealing with the Computational Model of RM-ODP. Both are
needed (and more)
for a comprehensive solution in healthcare (and other domains).
Dave
>The other point about GEHR is that it isn't maintained any more,
>is it? The current version is 1.0 as of June 30 1995.
>
>regards
>-Gunther
>
>--
>Gunther_Schadow-------------------------------http://aurora.rg.iupui.edu
>Regenstrief Institute for Health Care
>1050 Wishard Blvd., Indianapolis IN 46202, Phone: (317) 630 7960
>[EMAIL PROTECTED]#include <usual/disclaimer>
