Thank you for your thoughtful and informative comment. Short reply:
1) I know, I am a member of HL7, but getting all that in would have
changed the focus of the article and leaving it out completely is
unsatisfactory.
2) Depending upon whose definition of 'protocol' you are using. In any
case I have changed 'interoperability protocol' to 'messaging
standard'. I've also changed the UI paragraph slightly to emphasize
that I'm referring to the user interface, not the data layer.
Again, thank you very much for your insightful interest and comment.
-- Ignacio Valdes, MD, MS
-- Editor: Linux Medical News
-- http://www.linuxmednews.com
On Mon, 12 Dec 2005 12:59:41 -0700
"Kevin M. Coonan, M.D." <[EMAIL PROTECTED]> wrote:
Ignacio,
Thanks for taking the time to put this comment together. I most
certainly agree that the utility of commercial monolithic/integrated
EHRs is
yet to be published. You likely would agree that most of the
documentation
user interfaces are kludges that slows physicians down, CPOE systems
are
poorly thought out and over all design does not reflect the
cognitive model
of the provider or the workflow of the clinical environment. I am
sure that
if we had a standard "back end" or workbench (think Eclipse) then
the
functionality we need could be accomplished by small applications
that do a
few things very well, and could be swapped with another if the first
wasn't
what the user liked (sounds like 'nix...hmmmm....maybe onto
something).
Informatics is a maturing field. Given the complexity of the
domain
(how many banks have data dictionaries like the UMLS?) and the
substantial
underfunding industry wide the early (1960s even) promises of a
panacea seem
naïve. Hype from vendors is, well, hype from vendors. Consider the
source.
Press releases from a hospitals (or other) PR firms are obviously
not a very
realistic portrayal of reality. If you believe those, I have a new
ACEi (or
'statin or SSRI) I would like to sell you. Good things come to
those who
wait!
However, I believe you have made a few comments that are quite in
error. First, HL7 isn't an "interoperability protocol"--it is an
all
volunteer organization of people (of which anyone is free to join).
HL7
develops standards (just like W3C does) for messaging between
disparate
systems at the application (not user) level. The messaging standard
(version 2.x) is possibly what you are referring to, is widely used
(for
example so and EHR can communicate w/ a laboratory system) and
provides a
syntactic framework, but is far from ideal. The emerging, newer,
version of
the messaging standard (version 3) is much more specific, much more
detailed. In addition to the specification, there will be
conformance
criteria to assure adherence. While the interoperability protocol
is not
specified by HL7, most use conventional approaches (e.g. HTTP, SSL,
etc.)
In addition, version 3 is designed to provide semantic
interoperability. Local variations and options are not permitted
(although
some variation between countries is permitted). The degree of
specification
and standardization is tremendously different between the two
versions.
Beyond the messaging standard, HL7 provides a standard for exchange
of
documents (CDA), single log-in to multiple applications (CCOW), and
creation
of medical logic modules (Arden syntax). In addition, HL7 is
creating
functional models for various systems, again with conformance
criteria.
The comment about users needing to learn new interfaces between
systems is complete nonsense. The HL7 specifications are all at the
application level, and users would have no knowledge or awareness of
which
(if any) messaging protocols they are using. This is like
suggestion that
visitors to a web site would need to adapt to a new XML Schema each
time
they went to a different web site. The browser needs to grok the
underlying
representation, but if users who see XML (or HTML) code show up on
their
screen there is an error somewhere. Users may have to learn the
applications they use, but this is hardly a novel observation, nor
restricted to the medical domain. Given the complexity of what we
do, it is
much acutely felt.
As for the RHIOs, if they are making up standards they are doing it
wrong. HHS (via the NCVHS) has specified the standards that should
be used.
If there are needs for specialized regional messages, that is fine,
but it
would make no sense to create, validate and test a new wheel. If
they have
a health care system interoperability need that isn't met by the
available
standards they need to tell someone so that it can be addressed
properly. I
doubt, however, this is a significant occurrence. Calling RHIOs a
failure
is premature, given that most exist only in MOUs at this time and
those who
have been functioning (some up to two years!) seem to be doing some
good. I
would be interested if you can provide some examples of the problems
you
cite.
I would encourage you to visit HL7's web site and take a look at
the
RIM and version 3 specifications. Similarly, if you can provide
some
constructive critique of CDA R2, I am sure it would be most welcome.
I believe if you look into VistA you will find that in spite of
it's
warts, it was a transforming technology. Similarly, other large
system
(mostly home built at places like Columbia, MGH, Brigham and
Women's, LDS
Hospital) have shown considerable benefits. There is a dearth of
similar
publications about commercial systems, so extrapolating these
results to the
vendor supplied applications would be unwise. The concept of a
single
source, all-in-one system that you can just adapt to your
institution is
probably part of the current problem. This is one area in
particular that
the readers of these lists can help inform others. 'nix success is
largely
due to the lack of such a design (unlike some other OS) where
various
components can be selected, combined, experimented with and
discarded if
useless. The EMR community needs to move towards the notion of a
distribution of components, many of which may be shared between
vendors. If
everyone speaks the same language (such as HL7 v3) on the back end,
there
will be much less risk in adopting EMRs, and a much higher chance of
individual success.
Thanks for you continued efforts in this community.
Kevin
Disclaimer of possible conflict of interest: I belong to HL7 and
use SuSE
as well as that "other OS." Sorry, some of us need a 12-step
program or
something. I am also a former user of CHCS, which is a derivative
of VistA
the DoD uses (which, 10 years ago, was still better than what my
commercial
system can do today).
___________________________________________________________
Kevin M. Coonan, MD
Adjunct Assistant Professor, Division of Emergency Medicine
NLM Fellow, Department of Medical Informatics
University of Utah School of Medicine
[EMAIL PROTECTED]
-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED]
On Behalf Of Ignacio Valdes
Sent: Monday, December 12, 2005 11:25 AM
To: [email protected];
[email protected];
[EMAIL PROTECTED]; [email protected];
[EMAIL PROTECTED]; 'Tia Abner'
Subject: [os-wg] Editorial: RHIO's and the Illusion of Health IT
Success
"Does it bother anyone that for years, Health Information Technology
(IT) successes implied by the news and even in casual conversation
may
largely be an illusion? Does it bother anyone that Regional Health
Information Organization (RHIO)'s might be failing at a very high
rate? It
is important to ask the question given the United States rich
history of
failure and two notable successes with large scale Health IT."
Read the full article at
http://www.linuxmednews.com/1134404398/index_html
-- Ignacio Valdes, MD, MS
-- Editor: Linux Medical News
-- http://www.linuxmednews.com
P.S.: Please link the article and the website to your page if you
find it
useful.
_______________________________________________
os-wg mailing list
[EMAIL PROTECTED]
http://mailman.amia.org/mailman/listinfo/os-wg
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click
_______________________________________________
Hardhats-members mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/hardhats-members
