Re: [Gnumed-devel] clinicians: coding use case survey - please respond !

[Karsten Hilbert]

On Tue, May 03, 2011 at 10:04:56PM -0300, Rogerio Luz wrote:

> > #--------------------------------------------------------
> > In a list of terms with their codes I add/modify/remove a code
> > to/from the term "Fracture of left tibia".
> >
> > I would expect that now all patients with episode "Fracture
> > of left tibia" have the updated list of codes for that term:
> >
> >        yes
> >        no
> >        comment
> 
> Didn't get this one ... you mean to say that the code of Tibia was wrong and
> someone fixed it?

Or selected a more suitable code, yes.

> This should only be permited to the dbo.

That's easily done, yes.

> And then we get a
> HUGE problem, we will have to think twice about what to do here.

I'm not sure I see the big problem:

I've got 100 patients with "Fracture of left tibia".
Depending on which way the coding system is set up in the
database either all of them or some of them have the wrong
code directly or indirectly associated with that episode.

Indirect association via episode.description = coded_term.term:

        - fixing the list of .codes for .term in coded_term will automatically
          make all patient with episode.description = .term have the updated
          list of codes

Direct linking via .fk_episode = episode.pk:

        - fixing the list of .fk_codes in lnk_code2episode will make
          *that* episode (and not even other episodes within the same
          patient !) have the fixed list of .fk_codes - NONE of the
          other patients also having the wrong list of .fk_codes will
          benefit from the fix

Again, pick your poison !

Karsten
-- 
GPG key ID E4071346 @ gpg-keyserver.de
E167 67FD A291 2BEA 73BD  4537 78B9 A9F9 E407 1346