Re: [Gnumed-devel] Aggregating health issues on screen. - Contextural Info

[Karsten Hilbert]

> > I fully agree with what you say here, Richard. The only thing
> > is that we try to be logically consistent and make it possible
> > for the doctor to link distinct parts of the encounter
> > (encountlets, IOW) to their *respective* problem instead of
> > simply linking the entire encounter to a "best-fitting"
> > problem.
> 
> I'm not sure I fully understand what you are implementing, however I'll make 
> these comments.
Simply put: we are allowing the user to link data at a finer
grained level than "the whole consultation". This in no way
destroys the connection between the parts of said
consultation. Not even between those being linked to different
problems. If one so wishes one can display by encounter and
not worry about which part was connected to which problem but
rather just display all the linked problems as a list.

Nevertheless it is mandatory for clean data to allow
separation of the input per problem.

> Clinically information becomes meaningless when snipped out of its context.
It is not snipped out of context. Also this is entirely a
matter of display. What we are doing is *in*creasing the
contextual information by allowing to specify which parts of
the consultation relate to which problem. This neither says
how that data is later displayed. Also, it may not easily be
possible to cleanly discern which problem a specific data item
belongs to (in which case the user will just decide on
*something*) or the user may not even care (in which case she
will just "dump" all data onto a catch-all generic "problem").

> Most encounters are only going to be half a dozen lines,( or half a page or a 
> page at the most more rarely).
I agree - but what does that tell us ?

> Think about this. It is unusal in general practice to have to look back in 
> clinical notes.
I take it you are referring to "during consultation".
Otherwise it goes contrary to what I do at each and every
night shift - sifting through notes possibly dating back to 1992
or so (which is when that particular surgery came into
existence) trying to extract data related to a given work
injury the insurance company inquires about. I so wish it
was all linked up properly, tell you that.

> Yet when faced witha  difficult case, where the symptoms have started to 
> present on a few occasions over a period of time, and the patient history is 
> often a little vague, or changing with every presentation,  When you look 
> backwards through a set of patient notes - your eye takes in the context of 
> the key words, and what information is associated with them. If you cut out 
> this context, and link display to only part of an encounter, then you may 
> miss the clues that will cause the 'penny to drop' and help you make the 
> diagnosis. 
Sure! If needed just look at the notes in a whole-consultation
mode. Where's the problem ? It's a frontend presentation
issue.

I am sure our backend supports any presentation mode you'd
need (after all that's my/our goal !!). If not I'd sure
like to hear about it !

Karsten
-- 
GPG key ID E4071346 @ wwwkeys.pgp.net
E167 67FD A291 2BEA 73BD  4537 78B9 A9F9 E407 1346