Wayne Wilson wrote:
> Boy, is this discussion bringing back old memories, Dave Forslund reminds me
> exactly how old later on and I have some inline comments.
>
> Just below, Tom Beale suggests interoperability built on engineering
> principles. Some discussion follows about the difficulty of connecting
> systems
> with different models, etc.
Um, is some of this discussion occurring on a different list, or
privately - no problem if so, but just curious because I haven't seen
all the messages which you quote below from the openhealth list on Yahoo.
> So, I agree with Tom about what experience tells us about standards and de
> facto
> momentum.
Also agree that de facto engineering standards, rather than ones nutted
out by small, select technical committees, tend to be much better.
However, de facto engineering standards only develop when one company or
group comes up with a very good implementation way ahead of everyone
else. I don't think that is the case in much of health informatics -
typically engineering solutions which might become standards take so
long to develop that competing or alternative solutions inevitably
spring up during the development period. Unless one solution is so
vastly superior on all counts to the others (which hardly ever happens
in health informatics), we end up with a plurality of engineering
solutions and national authorities or standards committees then have to
decide between them and anoint one or the other, or at least a few.
> > David Forslund wrote:
> > The vendor lock-in was the
> >biggest factor that worked against
> >the adoption of the fairly reasonable OMG specifications we worked on
> >from 96-01.
> >
> Gosh, was it that long ago? I have come to believe that alongside vendor
> lock-in, high complexity played a significant role.
I agree. Complexity + lack-of-Internet in the mid-1990s -> lack of CORBA
take-up. If you look at all the SOAP, WSDL and other current WS-*
standards, they are just as complex as CORBA (but mostly not as good) -
but the social network effects of the Internet mean that the complexity
can be conquered, even by a small start-up software company.
> I am not as sold on multi-source components as I once was. I believe that
> nearly all the value lies in process (workflow in it's generic sense) and how
> various 'natural' groupings of functions (Things that people do in a patient
> care setting) need to hand off to each other.
Absolutely. We have recently realised, with respect to public health
information systems, that workflow management is just as important as
semantic and syntactic information management, if not more so. We're
busily reading up everything we can on various workflow engines and
their underlying calculus and theoretical bases. I think the same
applies to hospital-based clinical medicine, and to primary care. In
fact, some form of workflow engine are vital to any human enterprise
which can't be done or isn't typically done by a small group who can all
talk to one-another face-to-face in order to co-ordinate their actions.
> Clearly many people 'sense' this value and find all encompassing suites
> attractive. Before Health care IT trys to build the PeopleSoft or SAP of
> clinical care I think we need to absorb the lesson of those big suites:
> Change
> in operations is constant and loosely coupled systems change faster than
> tightly
> coupled systems.
Yup.
Tim C
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/openhealth/
<*> To unsubscribe from this group, send an email to:
[EMAIL PROTECTED]
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
