On Sat, Jun 13, 2009 at 11:48:04PM -0700, Jim Busser wrote:
>> There will be tables defining persons as various (one or more) types of
>> people.
>
> We do already have a staff table. We do already have a dem.occupation
> table and maybe here could be stored a person's occupation as doctor
> and
> type of doctor (GP vs cardiologist vs respirologist) since if we slit out
> a separate table for doctors do we then need separate tables to identify
> which individuals are physiotherapists, which are nurses etc?
No, it would become a dem.healthcare_worker table which
links to dem.identity, and dem.org, and allows to define
roles and specialities.
> What would define a person as having been a patient? Would it be
> achieved
> - by inference, meaning that after (and only after) associated clinical
> records can be shown to exist, this must have been a patient, at some
> point in time?
Currently yes.
> - or would it be achieved by an explicit decision on the part of the
> user, for example a secretary or doctor creates the person and declares
> "this person is now defined as a patient, so that even if we did not yet
> create associated clinical entries, it is only because it is a new
> referral and we did not yet see the patient, they have an appointment to
> come and see us in 2 weeks.
In 99.9% of cases there will be some morsel of information
as to why this patient is coming to see us. There will be a
clin.encounter (even if not documented) with RFE set to
"phones in for appointment to discuss worrying headache".
Even an AOE could be contrived: "appt. 21.6.09, recommended
earlier attendance if vomiting or worsening" or some such.
It is only very rarely that there will not be ANY clinical
information. Even if it is "as-yet undisclosed but worrisome
trouble with well-being".
> There is value to be able to know, in a large database of persons, which
> of them has been a patient,
agree
> in order to assist appropriate record
> selection during (for example) data exports desired to consist purely of
> patient-type persons and similarly, when doing statistics, if you had
> 5,000 patients and 300 non-patients and wanted to know what percent were
> diabetics, you want this calculated with a denominator of 5,000 patients
> not 5,300 people.
Ah, good point !
I dare say the condition of being a patient would be quite
well described by having any sort of SOAP entry in
clin.v_emr_journal. IOW whether anything (lab result,
narrative, allergy, substance intake) was *documented* about
them.
This would then NOT include encounters, episodes and health
issues as those might be administrational only.
Does that sound like a usable definition ?
Karsten
--
GPG key ID E4071346 @
wwwkeys.pgp.net
E167 67FD A291 2BEA 73BD 4537 78B9 A9F9 E407 1346
_______________________________________________