Re: [Gnumed-devel] on meaningfully signing off reviewed items

[Karsten Hilbert]

On Sun, Jan 29, 2006 at 10:04:27AM -0800, Jim Busser wrote:

> Ian had previously pointed out the need for the original requestor of 
> information to sign it off.
We can do that. All we need is a way to identify who
requested the information. In the case of documents the
logic is this:

- default to the primary doctor of the patient
- allow that to be overridden to name another staff member

This is then stored in doc_obj.fk_intended_reviewer.

> Which meant that even if it was actioned 
> in their absence (by a co-worker) it remains for the original 
> requestor to sign.
Yep, any number of people can, and in some cases several
should, sign a given piece of information.

> Over in the lab schema / workflow we had contemplated 
> reviewed_by_clinician though I can't recall if we stored *which* 
> clinician - guess that could be inferred from the identity of the 
> user who updated that value.
Well, that's how things were :-)   The whole point of the
exercise is Ian's request that "reviewed" status be separate
from the actual reviewed information. IOW anything we
discuss here applies to test results just as much as to
docs. IOW for test results we also have meaningful values
for each of

reviewed (by any)
reviewed_by_you (current_user, that is)
reviewed_by_intended_reviewer

In case of test results (lab values) the intended reviewer
might possibly be deduced from who entered the lab request
(modified_by). Else it could be the primary doc, too. Or
overridden to be someone else.

> So the related thoughts are:
> 
> - anything that needs signing must be identifiable as whether signed 
> by the *requesting* clinician
yep, reviewed_by_intended_reviewer, IOW:

 "... where fk_reviewer = fk_intended_reviewer ..."

> - it is also important to clinical care to identify that, in the 
> possible temporary absence of the requesting clinician, the result 
> has been acknowledged (signed) as reviewed by *some* clinician
yep, .reviewed, IOW:

 "... where exists (select 1 from reviewed where pk_doc=XY) ..."

> - the concept of a requesting clinician will not apply to *all* items 
> received into GNUmed
Hehe, yes, that's why we don't put too much emphasis on
"requestor" rather than on "intended reviewer" ;-)

> Therefore, the field serving the reviewed_by_requesting_clinician 
> function will have to be allowed to be null for some items.
Sure, it's just a virtual field in a view. If review by the
intended reviewer hasn't been documented to have happened it
evaluates to "False".

> - if a doctor departs from a practice, it is possible they could keep 
> their access in order to be able to sign stuff that comes back in 
> after their last day of work. But at some point they will presumably 
> stop logging in, or may have died. So here Ian, would some other 
> doctor "take" responsibility for these and sign them "as" the 
> requesting clinician?
Well, of course, care needs to continue and someone will
have to take over responsibility. At which point
"intended_reviewer" may have to be updated in the database,
which is only prudent.

BTW, we don't allow any staff to be deleted from dem.staff
as anything is still linking to that staff entry such as to
not allow any dangling links to appear.

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