Re: [Gnumed-devel] Identity numbers

[Karsten Hilbert]

On Tue, Jan 22, 2008 at 08:40:59AM -0800, James Busser wrote:

> It also reminds me the value of using a patient ID to check against 
> creating someone who already exists. These IDs will be any of the 
> following:
>
> - public health system unique identifier, if one is locally in use
> - private health insurer identifier... the patient may have more than one 
> kind of insurance, but is likely to use one in preference to the others
> - hospital or cancer treatment center chart numbers... these would seldom 
> be patients' preferred or primary identifier but without them it can be 
> very frustrating to try to track anything down for a patient who has gotten 
> care there so is sometimes worth putting into one's EMR. And maybe of use 
> in matching if these facilities send any electronic reports.
Well, just type any of those numbers into the search box and
select the patient you are after from the list. Our search
takes into account "external IDs" already.

> In the case of more than one identity number per patient, I expect there 
> will be usefulness to set some kind of preferred or default value (in the 
> case of only one identity number it would automatically be the default) and 
> that value, at least (if not all of them)  could be used when protecting 
> against duplicate patient entry.
It's a matter of educating front-desk staff. If it is
suspected the patient has been here before they should try
to find the previous record. Now, that won't help when
neither staff nor patient have any memory of attendance.

> The key should be a combination of ID_Type and Value, because two different 
> ID systems might end up having the same value.
>
> I am not sure whether we should allow user over-ride to permit a duplicate 
> ID number.
I think so.

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