On Thu, 20 Sep 2001, Bruce Slater, MD wrote:
...
> People would create open and proprietary software that resonated with that
> data model. Everyone benefits. Actually, perhaps it is the data model that
> should be opened and let others create create ever more wonderful
> emblishments to display and enter data into it.

Hi Bruce,
  An open data model is necessary but not sufficient. GEHR(with
archetypes) is an example (www.gehr.org). As you pointed out, adequate
applications that allow real clinicians/researchers to make use of the
data models are necessary. It remains to be seen whether proprietary
applications can make use of an open data model without deviating from
that sharable data model.

> We have started a clinical outcomes research software development model that
> is PDA centric and web friendly.

Great! When you make your model public, I'll be happy to review it. In the
meantime, you may be interested in reviewing our web-centric outcomes
research software (data model + application) that we like to port to
PDA. We started work on a JAVA ME version several months ago.

> When I say started, I mean embryonic, but
> determined. We have not written anything that could not be realigned with a
> public domain or open source data model for health care. The question I pose
> is - where is that great hope; the health care data model?

Since you asked, I invite you to take a look at Open Infrastructure for
Outcomes (OIO). The project site is http://www.TxOutcome.Org.

It is based on plug-and-play forms (=mini-schema) that can be weaved
together to build systems. It integrates data analysis and XML-mediated
metadata/data portability features.

> I need one that
> is translatable into the mundane small potatoes developer realm and not need
> a UML PhD to know how to unfold the 10 page diagram it comes on.

I agree 100%. The OIO project does not have any UML diagrams but hopefully
you will still be able to understand how it works.

> Why can't
> we have a data model that is collapsable? that is the comp sci knowledge
> bots can drill deeper into many levels of definition, but I can float
> happily on the surface of a 2 page  print out that covers all the basic
> entities that i need to write software that does what I need today and will
> support what I write tomorrow.

That's what the OIO system is all about :-). You can create an outcomes
management system TODAY using OIO software and modify/add to the system
TOMORROW when your outcome measures/data elements change. The data model
is build-as-you-go through a flexible metadata layer. Interoperations
between metadata components are supported through a common metamodel.

The "side-effect" of this kind of flexibility include portable metadata
and portable patient records.

> Does that make any sense?

Makes sense to me.

> Does such exist?

I believe so. The OIO system has been in use at Harbor-UCLA Medical Center
since March 2000 and was released under the GPL in August 2000. It is used
to track about 2300 patients at Harbor-UCLA and there are two known sites
using the system outside of Harbor-UCLA.

> <apology>
> Sorry if I am butting in, but the ideas flowing around here are too good to
> leave alone.
> </apology>

Thanks for asking... It gave me a chance to share what I have been working
on.

Best regards,

Andrew
---
Andrew P. Ho, M.D.
OIO:Open Infrastructure for Outcomes
www.TxOutcome.Org (hosting OIO Library #1)
Assistant Clinical Professor, Department of Psychiatry
Harbor-UCLA Medical Center
University of California, Los Angeles


Reply via email to