On Thu, 3 Jul 2003, Tim Churches wrote:

> Andrew Ho wrote:
> > Since OIO does not include any medical terminology (or "data
> > model" as Tim Churches puts it)
>
> A terminology is not a data model.

Tim,

  You are right - but an EMR's data model must include a medical
terminology (supported by the EMR).

  For example, we can see the TORCH social history "data model" in

<socialhistory>
  <updated_time>2000/01/01</updated_time>
  <tobacco_use></tobacco_use>
  <drug_use></drug_use>
  <employment></employment>
  <education></education>
  <social_comments></social_comments>
  <homelife></homelife>
  <alcohol_use></alcohol_use>
  <updated_by></updated_by>
  <sexual_activity></sexual_activity>
</socialhistory>

(from example given by Tim Cook at
http://www.openparadigms.com/intranet/torch/emr/EMR-1052359285/export_record)

So, a terminology is not a data model but the data model often includes
the supported terminology.

> And OIO does (I think) have a data model, in fact many, user-definable
> data models,

You are correct. OIO has a data model but the data model does not include
medical / domain-specific / application-specific terminology. OIO is
similar to OpenEHR/GEHR design in this regard. The "terminology" layer is
built through forms (= OpenEHR archetypes).

So technically, OIO and OpenEHR per se are not EMR system. OIO can be used
as an "infrastructure" to build EMR's but OIO is not an EMR!

> which results in an underlying physical data model like this (using an
> E-R style diagram, where --> means has zero or more):
>
> Patient
>    -->FormA
>    -->FormB
>    -->FormC
>    -->etc
>
> Or if you prefer a more O-O view:
>
> PatientInstances
>   -->FormInstances
>      -->Value-AttributeInstances

Right.

> The latter view doesn't show that there are various Value-Attribute
> subclasses available, but UML diagrams in ASCII are a bit challenging to
> compose. Of course, I have only superficial knowledge of OIO, so I might
> be completely wrong.

You are completely right. There is an additional Attribute-type level but
it is a small detail.

Best regards,

Andrew
---
Andrew P. Ho, M.D.
OIO: Open Infrastructure for Outcomes
www.TxOutcome.Org

Reply via email to