A problem-oriented medical record OWL ontology

During the teleconference before last we had decided to take a 
first shot at a W3C note on the problem statement for computer-based 
patient records.  I had preferred that we develop literature with guidance 
towards solutions in this space as I think the problem statement is well 
defined (at least with regards to the functional requirements needed to 
support healthcare and life sciences).  It seems to me that best practices 
and the development of vocabularies to provide guidance to others in our 
community is one of the core tenants of our charter.

I think if there is a challenge in this regard it is the development of 
uniform core vocabularies that span 'from the bench to the bedside' in 
concert with our efforts to demonstrate bench to bedside solutions - not 
so much the problem statement in general.  By uniform vocabularies
I don't mean a patchwork usage of ontologies that 
model biomedical reality but have little to no connection to 
patient-oriented healthcare information.  I also think we can go a step 
further from conccneus that HL7 RIM (as an OWL ontology) has well-known 
ontological issues.  The availablity of well established foundational 
ontologies (both within and across domains) reduces the problem to an 
exercise in ontology engineering at most (IMHO).  It seems to me that 
concensus in this regard is long.

So, for some time I've been developing (essentially) a foundational 
ontology for patient medical records which uses the Problem-oriented 
Medical Record model as a guide for how concepts in the realm of 
healthcare information (well modelled by HL7 RIM) can be grounded in both 
a domain-agnostic foundational ontology (DOLCE) as well as foundational 
ontologies for clinical terminology (GALEN) - both of which are freely 
available.  In addition, I took the oppurtunity to attempt to resolve some 
of the well articulated ontological failings of HL7 RIM (by being explicit 
about the concept of the action of recording clinical information and the 
phenomena being described by such recordings).

I've put up a Wiki that attempts to capture the motivation and rational 
for the ontology as well as links to the vocabularies that were used. 
There is also a link to the OWL / Protege files as well.

Best case scenario, I'm hoping the interest group adopts it as a 
foundation for expressing concepts from the realm of patient medical 
records (as an alternative to using HL7 RIM OWL exports as such a 
foundation).  Otherwise, I'm hoping to at least facilitate some 
conversation on how we can contribute towards some concensus on medical 
record vocabularies through the use of OWL and foundational ontologies - 
since such a contribution falls squarely under our charter.

http://esw.w3.org/topic/HCLS/POMROntology

Chimezie Ogbuji
Lead Systems Analyst
Thoracic and Cardiovascular Surgery 
Cleveland Clinic Foundation
9500 Euclid Avenue/ W26
Cleveland, Ohio 44195 
Office: (216)444-8593
ogbujic@ccf.org

Received on Saturday, 2 December 2006 23:09:37 UTC