Tim Churches wrote: > [EMAIL PROTECTED] wrote: > >>Hi Geoff, I accept what you say about the GP world needing motivation. > > > The motivation will come when there are really good decision support > available - not just catching prescribing errors, but tools which genuinely > help with differential diagnosis, investigation ordering and care planning > and monitoring - but such tools can only ever be as good and as detailed as > their input data, which is why SNOMED-CT and the like are so important. Of > course, we have a chicken-and-egg situation (or since this is an informatics > list, I should say a bootstrapping problem): clinicians won't enter detailed > SCT-encoded data until there are reasons to do so i.e. good decsision support > tools, and the investment and R&D to create those tools won't happen unless > clinicians collect detailed, coded data. > > Which is why Jon's projects to make SNOMED-CT encoding from free text and > other forms of natural or semi-natural language are so important. Ahh, you, > say, but that presumes that GPs will type all this stuff in... but see below.
A good example of this is the way in which people are currently using the history items in MD. You can currently use a DOCLE diagnosis (picklist) *or* use a freetext uncoded diagnosis. A number of GP's (thankfully a small number) down here use history items as a way of writing what they feel are more useful referral letters eg Not OSTEOPOROSIS (coded) But OSTEOPOROSIS on Dexa (-2.7) (uncoded) Usage follows purpose ... As there are not perceived to be other uses for the History items (!) then the whole point is lost. Michael Tooth GP Hobart _______________________________________________ Gpcg_talk mailing list [email protected] http://ozdocit.org/cgi-bin/mailman/listinfo/gpcg_talk
