xml-dev - Re: Reification, was Re: Ontologies

[Jonathan Borden <jborden@mediaone.net>]

Bob Kline wrote:
> On Wed, 20 Dec 2000, Jonathan Borden wrote:
>
> >     In any case one of the activities of the ASTM XML Healthcare
> > committee will be to issue an XML DTD for the codification of
> > medical guidelines in which many of the raised issues are addressed
> > (for example, the authority issuing the guideline, links to
> > evidence/articles supporting the guideline, the time period in which
> > the guideline applies, etc. etc.).
>
> What ironic timing for your example, coming as it does in the wake of
> yesterday's article in the New England Journal of Medicine reporting
> that doctors are consistently ignoring the warnings sent to them about
> conditions under which a given medication should not be prescribed. :->}
>
    Despite your surprise, the medical community is aware of errors and
there are many people actively interested in ways to reduce errors. Yet the
problems are not simple, but technology while greatly increasing the
complexity of the system (and hence potential for errors), can also help to
reduce errors.

    The 2001 edition of the Physician's Desk Reference (for medications) is
3506 pages of very fine print. It is updated yearly. Imagine if you were to
need to look up each language token in a 3506 page specification before you
typed it into a terminal. How much work would you get done? Now realize that
many empoyers of physicians alot 15 minutes for a visit.  So either the
practice of medicine gets alot more efficient or more errors get made or the
country decides to spend alot more money on healthcare. Something gives.

    Without some of the mechanisms we are discussing it is not possible to
integrate various warnings and other guidelines issued by various
organizations into accessable references AND provide for the absolute
necessity of local variations in practice.

    In my own practice, particularly as we are involved in especially risky
procedures, we have a rigorous quality assourance program staffed by a team
of Ph.D. trained physicists. Using techniques similar to those associated
with quality control on airplane control systems we are able to reduce our
error rate to acceptable levels (i.e. we have had no uncorrected errors
since this program started). But this is costly.

Jonathan Borden (M.D.)
The Open Healthcare Group
http://www.openhealth.org