In repsonse to Andre these are some of the things we are doing related to
subsetting SCT.
We have completed a project with the Family Medicine Research Centre to map
ICPC-2PLUS to SCT. We were able to match about 80% of ICPC, the remainder
needed to be done by hand. Many different matching strategies were used,
the simplest being string matching and the most sophisticated using
morphological transformations of words, searching in regions of the SCT
ontology, etc. The FMRC did a manual check on all matches and showed an
error rate <5%, and has to complete the last 20% by manual matching -that
is a work in progress, and we are waiting to hear from them. We are
reporting the results of this work at the upcoming HISA conference in
August.
For those of you using ICD-10AM we are working with the NCCH to achieve a
mapping but the work is much more slow going - progress will also be
reported at HISA.
I have an Honours project running this year to build a terminological
server that addresses , among other things, how to maintain a subsytem of
SCT for particular medical sub-groups, for example we have just
implemented (in a simple way) the local lexicon/abbreviations of the RPAH
in our Text-to-SCT processor.
cheers
jon

Quoting Andre Duszynski <[EMAIL PROTECTED]>:

> Quoting Andrew Patterson <[EMAIL PROTECTED]>:
>
> > > SNOMED concepts are one thing, the other is to contextualise each
> > of
> > > these i.e. the 'concept' of Diabetes Mellitus is all that is
> > required
> > > for a diagnosis, however there are a raft of supporting clinical
> > > measures, diagnostic tests and other items (workflow, referral,
> > > interaction with allied health & specialists) required to establish
> > this
> > > diagnosis of DM.
> >
> > I agree that for common problems such as diabetes there
> > will be archetypes that constrain the snomed terms used and help
> > with workflow etc, but there is always going to be a need for a
> > generic
> > 'diagnosis' of a patient. I mean, the patient walks in, the GP takes
> > some
> > notes, makes some observations and then decides you have 'blah'..
> > The clinical system needs to be able to save a coded diagnosis for
> > this event ala
> >
> >
>
http://svn.openehr.org/knowledge/archetypes/dev/html/en/openEHR-EHR-EVALUATION.problem-diagnosis.v1.html
> >
> > So the question there is how the diagnosis code for
> > 'blah' chosen from the 300000 snomed
> > terms.. obviously we can restrict it to all the terms in snomed that
> > are 'findings'.. but that still leaves a pretty large number of
> > codes.
> >
> > Andrew
>
> Limiting the scope of SNOMED concepts for general practice use would
> intersect with that undertaken within the GP Vocabulary project to some
> degree...
>
> Is such an activity ongoing and being tackled worldwide; general
> practice, family physicians, primary care SNOMED concepts ??
>
> Andre.
> _______________________________________________
> Gpcg_talk mailing list
> [email protected]
> http://ozdocit.org/cgi-bin/mailman/listinfo/gpcg_talk
>




----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.
_______________________________________________
Gpcg_talk mailing list
[email protected]
http://ozdocit.org/cgi-bin/mailman/listinfo/gpcg_talk

Reply via email to