Re: [Gnumed-devel] Gnumed's design suitable for complex encounter requirements

[Karsten Hilbert]

On Wed, Dec 23, 2009 at 09:55:11PM -0800, Jim Busser wrote:

> … He and researchers digitally recorded 167 consultation
> by 15 GPs... they were then able to compare the interaction
> between doctors and what was being entered into four
> different EMR software systems. [T]hey found often
> “there’s tons of navigation for very little use.”

This is one area where we can try to improve and where
non-coders can help.

> … Also, consultations rarely happened in a structured,
> ordered manner like the EMR’s recording system. Patients
> would jump from topic to topic or symptom to symptom.

We all knew this. This is why I keep thinking of a widget
where I first enter "facts" into a non-structured text area.
Within that area I can "pre-sort" by placing "facts" on
individual lines. Later on I can use a tabular/grid-like
view to move facts between problems.

This isn't quite thought out which is why I haven't
attempted it yet.

> … In half a dozen of the consultations they recorded,
> the person accompanying the patient also wanted medical
> attention.

...

> … Interruptions were just a normal part of a
> consultation. All but two of the GPs were interrupted and
> six were interrupted more than once. There were 13 phone
> calls. “This is the reality of primary-care practice and
> we should be designing systems that take account of this,”
> he said.

This seems obvious to me.

> Worse, from the point of view of a researcher who wants to
> extract data, many of the consultations didn’t involve any
> imputing [sic] of easily harvested coded data. Many
> consultations resulted in nothing more than text entry.

I dare say that's true.

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