Re: [Gnumed-devel] EMR tree display of allergy

[Karsten Hilbert]

I believe we should document our findings in this thread in the Wiki. It should 
be a chapter
in the manual.

> To your point about responsibility (litigibility):
> 
> Until the point where the question gets asked, the reasons for  
> "unknown" seem rather moot.
I agree. That is modelled as NULL value.

> Inputting "unknown because the record was  
> only just created and the patient has not yet been seen in the  
> praxis" feels to me to have no value.
Agree. Same as above. NULL value.

> So while I grant (accept) that if you are supplied with information  
> about a patient, say on transfer of care, then you could potentially  
> know their allergy information without having asked. Yet at the point  
> of transferring it into GNUmed you maybe ought to confirm, any way,  
> by direct inquiry from the patient. So "unasked" does remain, to my  
> mind, the "state" that cannot be left this way.
I agree.

> Once a clinician asks, it is still possible for it to be unknown or  
> at least uncertain.
I agree.

> The patient may
> 
> 1) confirm what already exists in other form, for example in transfer  
> or personal documents
> 2) provide you with the information from memory
Which would then result in allergies being documented.

> So we are not yet fully through the list of possibilities and yet  
> have already reached a branch point. 

4) indicate that according to current knowledge there is no allergy

> Suppose a patient discloses that  
> they are allergic to nuts, but you judge somehow that they are  
> holding back something else from you. Do you enter the nut allergy  
> *and* put "Undisclosed" if the program logic would even allow this?
...
> 3) indicate that there exists something about their information that  
> they are not prepared to (fully) share
OK, that is hard. I would then add an allergy named "undisclosed trigger" :-)

> Next, even if no nut allergy would have existed, removing from issue  
> the program logic, do you put "Undisclosed" just on suspicion, or do  
> you at least get affirmation from the patient that there is at least  
> partial undisclosure
I would try, yes.

> Once you share your view of the above, could you then share how you  
> would handle the situation where the patient does recall pretty  
> confidently that they did indeed, in the past, have a problem with  
> one of their previous medications, but simply cannot remember the  
> medication, but they do or don't clearly recall the adverse reaction?
I would enter an allergy - after all very likely there IS one - with an 
unspecific name like
trigger = "some previous medication".

BTW, I am fine with symbols in the CAVE field. We may later have to reconsider 
them to make them
non-ambigous against "other" therapeutic precautions (pacemaker, on warfarin, 
etc).

Jim, when you prepare the post-Japan summary mail may I ask for a split between 
"0.3.2-bugs"
and "other" ? 

Thanks,
Karsten
-- 
Ist Ihr Browser Vista-kompatibel? Jetzt die neuesten 
Browser-Versionen downloaden: http://www.gmx.net/de/go/browser