Alvin Marcelo wrote:
> 
> Thanks Brian. That was a concise but clear presentation of Circare.
> I would like to bother you for more by asking you to give me (or us) a
> step by step approach on how to build a
> Circare system in let's say, a small primary care practice. It may be on
> your website, but it would be nice to hear it from you. Or you can refer
> us directly to the webpage.

The circare web page is at:

http://www.minoru-development.com/circare

The installation instructions are in a file called INSTALL in the source
code directory.

Circare is very much a work in progress (very little real functionality
to date) and is not easy to install at this time.  I would be happy to
provide e-mail assistance to anyone brave enough to install it in it's
present state.

There are plans to write an implementation guide that would help
installation.

The first version is targetted at health care networks, not at
individual physician practices.  The components would be very useful for
a physician practice (PIDS is very widely applicable), but they would
need to be packaged specifically for this application.
> 
> A few questions too:
> 
> 1. You are using the CORBAmed model. There have been concerns ( in
> past posts) raised about its capability to completely represent medical
> information (unlike the abstract GEHR and HL7). How do you respond to
> this?

CorbaMed is incomplete.

Taking a abstract model, such as the HL7 one or others, and elaborating
specific interfaces is a large but extremely valuable goal.  I'm sure
that the CorbaMed group would welcome proposals from the open source
community for new CorbaMed interfaces.

These interfaces do not *need* to be standardized to be useful for open
source development.

> 
> 2.  I am concerned with the ff statement (maybe because of
> lack of in-depth knowledge of CORBA technology):
> 
> > medical information.  It's architecture is based on the CorbaMed PIDS
> > standard.  The PIDS standard is primarily for patient identification but
> > it allows access to of an arbitrary number of attributes of arbitrary
> > types.
> 
> Is arbitrariness a virtue or a curse? I had thought that the less
> arbitrariness there is, the better. Am I missing the context of
> "arbitrary" here?

A standard needs to be flexible to be widely applicable.  The fact of
the matter is that I'm using this flexibility to make PIDS do something
for which it was not primarily intended.

The actual implementation is not "arbitrary".  Circare uses the
flexibility in the standard to define it's own specific attribute
types.  Implementing a system with too much flexibility is a waste of
effort.  A standard without enough flexibility, is not widely
applicable.

Total arbitraryness is bad.  Corba, along with XML, and SQL allow the
definition of "arbitrary" things (objects, documents, tables) and
provide meta information (UML, doc types, data dictionary) to be able to
work with them.

> 
> Please help.
> 
> alvin
> 
>

Reply via email to