[Gnumed-devel] Re: Managing users: restricting access within GNUmed

[James Busser]

I would be interested in comments, either privately or on-list, on the 
suitability of ideas in my last post, namely

> Patients might be accorded a choice among privacy "directives", in 
> which their confidential issues (and any enclosed episodes and rows), 
> until the patient would alter their directive, can be visible
> - 1 = only ever to the authoring clinician
> - 2 = only ever to the authoring clinician PLUS the patient's most 
> responsible physician (MRP) within the praxis, if this individual 
> should be different
> - 3 = to the authoring clinician, PLUS any clinician who would attest 
> a serious concern for the patient
*** intending "in the absence of being able to contact,
      or communicate coherently with, the patient
> - 4 = to any clinician in the praxis

Two additional thoughts:

1) is_confidential WHERE directive is NULL could support a scenario where –  in 
the absence of a patient stipulation, but in a clinician's precautionary 
judgement – it would be prudent to designate something confidential, in order 
to better assist its default exclusion from automated outgoing reports

2) perhaps the operational (practical) difference between

   clinical data *coded* "confidential" (which any clinician can see)

and 

   clinical data *not* coded "confidential" (which any clinician can see)

lies more outside the praxis where --- as with additional thought (1) --- the 
state of an item being confidential could assist some default exclusion from 
scripted abstracting