Dear All,
From a technical POV, you're close to describing case-based reasoning,
which is a well known AI technique.
The problem tends to come with matching approximate cases: If your
database contains women aged 55 -65 with breast cancer, and you see
someone aged 66 (or 54) then can we match a case. Such things might be
solved on an individual basis, but this gets difficult to scale up and
handle combinatorial issues (age and disease type and treatment, etc.)
Try googling "case-based reasoning medicine and you'll get a lot of
useful links.
Matt
Rakesh Biswas wrote:
> Thanks Cleo,
>
> Yes I have and it is nearer to creating an experiential informational
> database than other sites. Other sites where evidence based queries are
> expressed in terms of natural language and work toward popularising
> evidence based practice are http://www.curbside.md/
>
> The problem with SERMO is that at present it may not be possible for a
> busy clinician to look up all the experiential information his
> colleagues are generating particularly when s/he is at point of care
> (and the time for point of care varies with how busy a clinician is). I
> hope someone from sermo is listening.
>
> To address this problem we are creating a web 2.0 platform that can
> address point of care experiential queries by matching them with similar
> entries into the database by other health care professionals and even
> patients, heath information specialists/medical librarians etc thus
> moving closer to creating a user driven collaborative network centred
> health care for an individual patient (rather than the
> individual clinician dependant approach that has dominated medicine for
> centuries).
>
> rakesh
>
>
> On 3/21/08, *Pappas, Cleo* <[log in to unmask] <mailto:[log in to unmask]>>
> wrote:
>
> Have any of you looked at www.sermo.com <http://www.sermo.com/>?
>
>
> Cleo Pappas, MLIS, AHIP
> Assistant Information Services Librarian & Assistant Professor
> University of Illinois at Chicago (M/C 763)
> Library of the Health Sciences
> 1750 West Polk Avenue
> Chicago, Illinois 60612
> [log in to unmask] <mailto:[log in to unmask]>
> 312.996.2759
>
> On Thu, March 20, 2008 7:10 am, Rakesh Biswas wrote:
> * Thanks Dimitri,
> *
> * I find your typology of evidence based practice very interesting as it
> * appears to have emerged out of clinical practice.
> *
> * Unfortunately we are yet to achieve a collaboration of clinical
> * practitioners (who may not have time or interest to sit through the
> * analysis
> * involved in systematic reviews) and ivory tower evidence based
> * practitioners
> * ( who are more into synthesizing systematic reviews etc but
> understandably
> * may not have time for actual clinical practice).
> *
> * The question is how do we bring about a collaborative network that can
> * accomodate these typologies? Like a self critical clinician regularly
> * opening up his/her experiential information for suggestions from
> evidence
> * based practitioners and even intelligent patients across the world
> * (without
> * losing his or her patient's privacy) would not only facilitate
> evidence
> * based practice but also augment clinical learning (as the day to day
> * experiential information that a clinician gains is otherwise lost
> to the
> * world but could actually be incorporated in his her clinical
> e-portfolio).
> * This would document experiential data (that is mostly otherwise lost
> * regularly) which in turn could be appropriately structured with
> current
> * best
> * evidence (this is something like evidence based case reports but
> we need
> * them in larger quantities, ideally for each and every patient that
> we come
> * across and not just the 'interesting cases'). Also it may not be
> * restricted
> * to clinicians as pathologists, pharmacologists, microbiologists
> and all
> * other paraclinical do get involved with patients as individuals
> and may
> * not
> * just restrict themselves to larger population based studies.
> *
> * In the end I guess it boils down to the old debate of the
> differences and
> * challenges in caring for individuals vs populations but if we can
> get out
> * of
> * that there is a way where we could achieve a collaborative merger
> between
> * caring for individuals and populations where we could start
> recording our
> * day to day individual patient clinical experiential/clinical problem
> * solving data and mash them with evidence based/population based
> data. IT
> * has
> * the power to do that, this is not only record individual
> experiences but
> * also structure them with evidence based empirical data. We are
> involved in
> * one such project and would be glad to share details individually.
> *
> * rakesh
> *
> *
> *
> *
> * On 3/19/08, Dimitri Spyridonidis <[log in to unmask]
> <mailto:[log in to unmask]>> wrote:
> *>
> *> dear all
> *>
> *> Apologies for cross posting
> *>
> *> I am trying to build up a model of how could evidence based
> *> practice(EBP)
> *> be
> *> delivered and have identified these 5 different typologies
> *>
> *> 1) clinicians who are self critical on their practices and using
> regular
> *> audits of their practices
> *> 2)clinicians who are using their own personal experience and skills
> *> interrogating of published research using thehierarchy of evidence
> *> 3) Evidence is socially constructed -Professional consensus is
> essential
> *> for
> *> the adoption of valid scientific evidence
> *> 4) use of national agreed guidelines
> *> 5) Problem Solving Model: The existence of a problem pulls knowledge
> *> into
> *> practice.The need to generate a solution or choose among
> alternatives is
> *> the
> *> impetus to facilitate EBP
> *>
> *> I would welcome any comments on this model and would really
> appreciate
> *> if
> *> you do think there could be additional models to facilitate evidence
> *> based
> *> practice
> *>
> *>
> *>
> *>
> *>
> *>
> *
>
>
>
--
http://acl.icnet.uk/~mw
http://adhominem.blogsome.com/
+44 (0)7834 899570
|