I agree with Dqavid's points.
The world, unfortunately, is not perfect. Understanding how the ISO data
types standard came into being might be useful in understanding why it is
as it is. After more than 5 years in trying to get a g;obal standard for
data type, a group, lead by Graham Grieve, was able to put together a
combination of documents from ISO, CEN, and HL7. The result was an
overwhelming and certainly more than we need or will probably ever use data
types.
I would be interested in knowing why this standard is considered to be
complex. Is it a result on the large number of data types? Is is a result
of the complexety of some of the data types?
I would be interested in knowing how anyone would change the simplier data
types. It seems to me that most of these "primitive" data types are
exactly wahat we have been using for a long time.
I would suggest that NCI - and others - simply identify what subset of the
datatypes they propose to use and move ahead. On the o0ther hand, that
migtht happen naturally out of the full set. If its\'s too complex or not
useful, it will not be used.
I agree that everyone haveing their own set does not solve the problem. If
any set meets my needs, I don't care what else is in the package.
W. Ed Hammond, Ph.D.
Director, Duke Center for Health Informatics
Thomas Beale
<thomas.beale at oce
aninformatics.com To
> For openEHR technical discussions
Sent by: <openehr-technical at openehr.org>
openehr-technical cc
-bounces at openehr.
org Subject
Re: ISO 21090 data types too
complex?
11/06/2010 11:59
AM
Please respond to
For openEHR
technical
discussions
<openehr-technica
l at openehr.org>
On 06/11/2010 15:29, David wrote:
Thomas,
I've been retired for some time now but have kept a quiet watch over
things that have been happening in the world of heath data
standardisation. I stay quiet because I do not wish to intrude into
arguments that people in the real world are having.
However, Thomas, I think that what these "Guidelines" from NCI CBIIT
suggest is exactly what most IT people working on large health IT
projects (or any type of IT project for that matter) think - ITS
EASIER AND CHEAPER TO DO IT OUR WAY (rather than trying to use
standards) IN ORDER TO GET OUR INTERNAL SYSTEMS UP AND RUNNING. This
is a narrow and somewhat selfish approach (some would say justifiable
in $ terms) but it achieves internal agendas. I have had some
experience with this sort of thinking having worked for the
Australian Institute of Health and Welfare and its providers.
I think that these "Guidelines" include any untested standard which,
unfortunately, includes openEHR. It's called throwing the baby out
with the bath water.
I agree: the same response might be possible with another standard. My
question is about the complexity level of the standard in question. In
other words, is it possible to imagine that the NCI's response to a
different kind of data types standard might have been something like:
the standard appears to be concise and lightweight
it will work well with a) other standards that we use b) our existing
modelling frameworks
it will cost something to implement but not too much
our ability to verify the standard from available implementations
makes it low-risk
our ability to re-use available open source implementation in various
languages makes it easy to guarantee its being deployed uniformly
across the enterprise
it may in fact save us work in the long run, by providing an internal
standard toward which to migrate data types that we already use.
If it is not impossible to imagine the above response, then we have to ask:
why are the SDOs not instead making concise, lightweight, validated and
implemented standards?
If the NCI CBIIT go for an "orthodox" solution rather than a
"standards" oriented solution, then you are still no closer to
implementing yours or anyone else's standards! And what are the
chances that, even if they went with a "standards" approach, they
would choose to go with any "standard" that is somewhat untested in
such a large system?
exactly my point ;-)
From what I've seen, it seems to me that openEHR would work well on
most levels of health requirements, however, if at least one of the
big guys (IT, government, health orgs etc) doesn't get behind it
because of their vested interests or inertia, you are going to be
pushing it up hill. I think that a relationship with IHTSDO could be
an important one as long as you don't get subsumed into their agenda
rather than pursuing your own.
well, let's say, a significantly augmented agenda needs to be considered...
- thomas
_______________________________________________
openEHR-technical mailing list
openEHR-technical at openehr.org
http://lists.chime.ucl.ac.uk/mailman/listinfo/openehr-technical
