Re: [Gnumed-devel] Better-supporting staff and other non-patient persons

[Sebastian Hilbert]

On Montag 15 September 2008, James Busser wrote:
> On 15-Sep-08, at 12:24 PM, Rogerio Luz wrote:
> > So we could make the DB create patients (at least) with only Name
> > LastName ? This seems reasonable to me ... even better if we can
> > flag this patient as "not demografically complete status"
>
> The last thing I would want to do -- especially with him on holiday
> -- would be to upset Karsten with objection to his carefully
> considered views about intactness of information.
>
> In the creation of people who are *patients*, it does very much make
> sense to have to input the date of birth for clinical reasons,
> because you want to take the age into account in your decisions...
> here the age has biologic - clinical importance, even if it would be
> "off" (inaccurate) by a day... say in the case of a patient who was
> born in a rural area and it is only known by her family that she was
> born approximately February of 1922.
>
> The second use of age is as a validity check to disambiguate
> otherwise-similar individuals or to further provide proof that the
> individual is who they claim to be. The limitations here are that
> they have no value unless linkable to the external world by virtue of
> the person having given that same information to some other source of
> data, or being registered under that date of birth with some
> authority (whether or not is was the person's true birthday when this
> might not have been known). So this is a purpose totally separable
> from the clinical one.
>
> This is why I would say:
>
> - for persons who would be created in GNUmed *without* any plan of
> them getting medical care, do not require (as for staff and other
> persons) the date of birth. Enable it -- sure -- but do not require
> it, because to do so only places unnecessary constraints.
>
> - for persons who would be patients, I can accept the constraint that
> a date of birth is mandatory because is it *usually* available at the
> point of creation. It is not *always* immediately available, which is
> a problem that I and some others might face, but it could be overcome
> in two ways:
>
> 1) input a ridiculous dob like 1800.01.01 and trust that it will be
> obvious that it will need to be changed at the first suitable
> opportunity
>
I would vote for 'unknown' instead. No false data should be entered.

-- 
Sebastian Hilbert 
Leipzig / Germany
[www.gnumed.de]  -> PGP welcome, HTML ->/dev/null