Re: [Gnumed-devel] how and when to select encounter type

[Richard Terry]

I think we should have a default user-defined encounter type set.

I'd like to make an interesting point in regard to the suggestion to prompt 
'on the way out' for the type of consultation - which at first glance may 
seem quite silly, and that is of accuracy of patient identification.

One of the commonest mistakes I make with my medical records is the following.

I've seen the patient and entered details in the notes. I get called out of my 
room to do something - or have patients in 2-3 rooms on a busy day. I return 
to my room, call in the next patient, chat, take history, forget that the 
last patients notes are still on the computer, write a script etc, print it 
out. This mistake is aways picked up by the chemist because the patient has 
to be identified. However, I'd be emabarrassed to say how often I do it. I'm 
sure I'm not alone, although, of course none of you bright people would do 
it. From talking to our chemists, it is not an uncommon scenario.

Hence I feel strongly that we should have the following.

On loading +/- exiting the patient record one should have the type of screen I 
attatched to my medical records program (see png attatched). One could 
include in this the type of consultation as per this thread.

Also I'd opt for a locked modal window which pops up into place after a 
pre-defined periord of inactivity which needs you to confirm patient identify 
to continue working.

Sounds extreme, but I don't think it is.

Regards

richard



On Tue, 10 Feb 2004 12:10 pm, Karsten Hilbert wrote:
> Hi all,
>
> while putting the finishing touches on the first version of
> client remote control (see tentative roadmap item #4 at
>  http://mail.gnu.org/archive/html/gnumed-devel/2003-06/msg00045.html
> ) I stumbled upon the question of when to start an encounter.
>
> Hardliners might argue that any time I do anything with a
> patient's record or file or chart is an encounter (of suitable
> type, not always with the patient involved). Given this
> assumption we'd have to ask for the type of encounter anytime
> we select a patient. A default encounter type (per patient)
> would be remembered.
>
> Another approach would be to delay asking for the type of
> encounter until something get's written to the patient's
> clinical (?) record. The question arises whether that's
> sufficient and whether we'd want to record "administrative"
> encounters as such as well (say, the patient just comes in to
> change her address details). This approach would allow to
> avoid creating an encounter when just looking up some data,
> say, revisiting a stored referral report for memory jogging.
>
> Now, what do you say, which way shall we go ? Or shall we make
> that configurable (easy, technically) ?
>
> Karsten

patient_identify_10Feb04.png
Description: PNG image