Veterinary Terminology Services Lab (VTSL)
Jeff Wilcke
Veterinary Terminology Services Laboratory at VMRCVM (VTSL) has been
working with veterinary terminology standards for 16 years. Our
organization includes 5 veterinarians with a cumulative veterinary
practice experience in excess of 30 years (including a small animal
practice co-owner (Dr. Julie Green), university internal medicine
practice (Dr. Jeff Wilcke), small animal practice (Dr. Suzanne
Santamaria), mixed animal practice (Dr. Gareth Moore), as well as
small animal practice and clinical pathology (Dr. Kurt Zimmerman)).
We have developed or are developing terminologies for the US
Department of Agriculture, the International Species Information
System (ISIS), the US Food and Drug Administration, the US Geologic
Survey, the Department of Defense, the American Animal Hospital
Association and the American Association of Equine Practitioners. As
far as we know, we are the only organization that has studied,
developed and delivered veterinary terminology for practical
purposes.
Our organizational philosophy… We work every day in an effort to
bring FUNCTIONAL terminologies (not just lists of terms) to veterinary
medicine at a significant cost savings (when compared to human
terminologies). Like all veterinarians, we are attempting to
leverage the best science that is available to medicine in order to
provide value to our clients. We are attempting to provide you all
with access to technologies built on the investment of hundreds of
millions of dollars by governments in the United States and Great
Britain (and now 14 other nations). We are attempting to leverage the
terminology investment that veterinary medicine has already made over
the past 40 years. We believe that veterinary medicine cannot afford
and should not attempt to build truly functional terminologies on its
own and we strongly believe that veterinary medicine should avoid
creating its own version of the “Tower of Babel” represented by more
than 100 terminologies that make up the Unified Medical Language
System (UMLS) (http://www.nlm.nih.gov/research/umls/knowledge_sources/
metathesaurus/release/license_agreement_appendix.html)
Standards in use today… Almost 20 years ago AVMA, its Informatics
Committee and its Standards subcommittee adopted HL7, LOINC and SNOMED
as the appropriate information standards for veterinary medicine .
(http://www.avma.org/onlnews/javma/jun02/s020601o.asp) Application of
these standards in human and veterinary medicine to date would seem to
have validated that choice. The US Government has adopted HL7, LOINC
and SNOMED for various federal programs. FDA is working on electronic
pharmaceutical labels. Standards employed include HL7 (message
architecture), LOINC (label sections) and SNOMED. USDA and AAVLD
have adopted HL7, SNOMED and LOINC based standards for diagnostic
laboratories. These same standards are actively been used in the
National Animal Health Laboratories Network (NAHLN). Small animal
medicine risks becoming an (albeit large) anomaly in the world of
veterinary information standards if it diverges from this trend.
On choosing terminologies… We believe that neither terminologies nor
messaging standards can be evaluated by staring at the content or
code. Standard terminologies are like battle plans; they appear to be
perfect as long as they don’t actually get put into action. Once real
users encounter standard terminologies, the requests for corrections
and additions are immediate. Further, it is simply naive to think you
can just “pick one (list) and be done.” Modern terminologies like
SNOMED have evolved from a need to adapt, change and grow while all
the while supporting “legacy applications” through sophisticated
history tracking mechanisms. All standards are “living documents.”
As medicine and medical practice move forward, the content of
standards must keep up with the “rate of mutation” of the profession.
To press a potentially strained metaphor, the mutation rate for
message and information model standards to the mutation rate for
terminology is like comparing the mutation rate of mammals to that of
bacteria. Terminologies should not be chosen for their existing
content, they should be chosen for their ability to gracefully evolve
and to keep pace with change in the profession they support.
How we think SNOMED-CT can and should work in veterinary medicine…
SNOMED-CT is a reference terminology that is open for use in any
country which is currently a member of the managing organization,
International Health Terminology Standards Development Organization
(IHTSDO). (It is not owned by NLM) This member-country access allows
use of core SNOMED. We believe that it is the terminology standard
that should be used in veterinary medicine. No veterinary
organization, in fact no medical organization, needs to use or should
be expected to cope with all of SNOMED-CT. We believe that SNOMED
can and should be subset for specific purposes , defined by those who
will benefit from its use. We believe that a collection of these
subsets can eventually provide the profession with a leaner more
functional “view” of SNOMED-CT. To that end, we have obtained the
authority from IHTSDO to extend SNOMED core for the purposes of
developing the content needed for veterinary systems. We have
developed and are continuing to develop what we’re calling a
“Veterinary Adaptation of SNOMED CT to include this extension content
and the subsets we’ve created for our clients. For the record, we
think SNOMED has appropriate content for: species, breeds, medical
findings & diagnoses, procedures, infectious agents (perhaps a few
other areas). Other terminologies are more suitable for other
purposes (e.g., LOINC and laboratory orders).
On intellectual property and codes… The intellectual property
associated with terminology is very difficult to track and to
comprehend. It is an anathema to many, that terminology management
organizations (like AMA, CAP, etc.) were claiming intellectual
property rights to medical terms in the first place. We maintain that
the intellectual property is NOT held in the terms themselves (you
can’t keep me from saying influenza) but rather in curating the list
(which terms are selected for use). In the case of AAHA, we believe
that AAHA’s intellectual property is in the list “definition” NOT
actually in the terms themselves. The situation gets more
complicated as the AAHA list has been created as a subset of SNOMED.
The IHTSDO claim rights to the intellectual property of the concept
codes and the terms. A cautionary note: During the 1990s, squabbles
over intellectual property were the greatest single impediment to the
development of effective terminologies. As for the codes, we view
them as no more and no less than database keys that allow for
effective and rapid indexing, history tracking and the most
appropriate means of supporting the use of multiple synonyms that
describe single concepts.
In regards to cost and support… Standards require administrative
support of editorial and financial concerns and support for physical
maintenance and distribution of the standards themselves. VetXML
presumably has the means to develop and distribute its message
standards. AAHA may have the ability to support an editorial board to
supervise curating the terminology content for small animal
practice. Neither has the capacity to curate and distribute the
physical terminology or maintain terminology history and control
versioning. Neither has expressed an interest in supporting a broader
scope that would include ALL of veterinary medicine (equine, food
animal, lab animal, public health…). VTSL has, over the past 15
years, built up an infrastructure for terminology maintenance that
includes editing, subsetting and (webservices based) terminology
distribution capabilities. But, we have thus far been unsuccessful in
finding a permanent, or at least long-term source of funding that
would allow us to provide terminology across the board and so we
currently work for one project at a time where funding permits. It
is very important to remember that this isn’t a one shot deal. There
WILL be maintenance and change that is necessary to the success of the
terminology as well as messaging standard. A broad based, well-
funded, well-manned group will be required for this to be ultimately
effective.
As concerns AAHA’s diagnostic code project… AAHA coordinated the
intellectual effort of veterinarians in creating a list of terms that
are appropriate for application in capturing information about small
animal patient findings and diagnoses. AAHA’s original intent was to
maintain its own terminology and a map to SNOMED. We convinced them
that this doubled their maintenance burden (maintain the terminology
and maintain the map). As a consequence, what has emerged (the
content of the AAHA “XML file”) is a subset of SNOMED concepts and
descriptions (terms). From our perspective, our project with AAHA is
not “hidden” or “proprietary”; it’s just “in progress”. While AAHA
can (whether or not they do is not within our control) provide you
with a copy of the XML distribution file (the primary deliverable of
our recent contract with them), we (at VTSL) do not believe that
viewing it is an effective way to evaluate the work that has been
done. Most of you are familiar with the VTSL browser at
http://terminology.vetmed.vt.edu/. It presents a sort of tree view
of SNOMED-CT core. At present, the AAHA terminology captures some
large portion of the leaves of the “small animal medicine” tree. It
does not, however, have most of the supporting branches.
Additionally, the upper hierarchy in SNOMED was modified ½ way through
our project so we must go back and re-evaluate. We apologize for the
inevitable delay this produces in any ability to publicly display the
terms on our website (in a manner we feel is useful to assessment of
its overall quality) but we are working on it.
What we think we can contribute… If a broad-based group can be
established (regardless of who leads the effort (AAHA, IHE-based
working group…), what messaging standards (HL7,VetXML…) are selected,
etc.) we would be willing and likely enthusiastic about working with
the group to evaluate messaging standards, explore ways of
implementing terminology within the messaging standards, and being
available to support the terminology needs of the group.
- Jeff R. Wilcke, DVM, MS, DACVCP
Metcalf Professor of Veterinary Medical Informatics, VMRCVM, Virginia
Tech
Director, VTSL
Member, IHTSDO Content Committee
Jim Young
I just learned more from your message than all the time that I had try
and research this issue on my own!
Thank you
Jim