Re: access to review audits (was Re: [Gnumed-devel] encounter edit before final save)

[Karsten Hilbert]

On Sat, Aug 16, 2008 at 10:16:06AM -0700, James Busser wrote:

>>>> Jim, do you think we should invalidate a test results review
>>>> when said result is modified ? (this would, in turn, notify
>>>> the responsible clinician courtesy of the then-lacking review)
>
> I agree that if the person who had *last* reviewed a result is  
> *different* from the individual who is currently modifying a result,

Oh, well, *that* surely IS checked before actively sending
a notification :-)

Also, if a change is written to the database but no actual
semantic change is being made (the review stays the same) no
notification is sent.

> the 
> prior results review should be invalidated and the fact that patient 
> results lack a review does need to be signaled somehow.
If the review is invalidated the result again lacks a review
and thus automatically falls under the "signal lacking
review" criterion.

> Whatever previous review had been made as to abnormality, importance,  
> and comment may remain valid despite the modification (e.g. Kalium /  
> potassium level report of 5.9 corrected to 5.8 may justify preservation 
> of the prior review). Therefore it might reduce the pain of there were an 
> option for whoever is modifying the result to simply re-confirm, or at 
> their option alter/update, the last review.

The widget for updating a measurement affords immediate
re-review by whoever updates the result. If the updated uses
that no notification will be sent (unless the new review
changes the old review).

> I wonder if we can avoid an unnecessary notification if whoever would  
> modify a result would also, in the same session, mark it as reviewed? I 
> cannot remember whether the appearance of items in an inbox is the  
> result of the writing of "notification records" records (which would  
> need to be deleted after the review was done), or if the notice is not in 
> fact a "record" at all, just a dynamic indication of the result of a 
> query,
The latter.

> which could have the ability to immediately refresh and would go 
> unseen
It would, yes.

> The intended reviewer, if different than the provider who last reviewed, 
> or if there is no valid review on account of failure of the person who 
> modified the result to take responsibility, would still detect an 
> unreviewed result, right?
Yep, but only if the assessment changed.

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