openEHR Concept Ids

Implementers (archive)
1

Hi All,

I'd like feedback on what other implementers think of the fact that we
have duplicate concept ids in the openEHR terminology.

For example:
In the version lifecycle state group deleted is 523
In the audit change type deleted is 523

Now one can certainly argue that 'deleted' is the same no matter what
group it is in. But in archetypes the concept id will appear as just
[openehr::523]

Having duplicate concept ids unnecessarily complicates the application
code because before you look up the code from the archetype you have to
figure out which group you are concerned with.

There are plenty of numbers in our number system to make group+concept
unique by concept id. I think we should modify the openEHR terminology
to have unique concept Ids.

OR......create another group called generic actions (or something
similar) and put these duplicates in their own group so we still end up
with unique ids. Though I don't think this latter suggestion is a good
one; for many reasons.

Your thoughts?

Cheers,
Tim

2

Hi Tim

I would argue that terminology should be seen in context, in this case a
name value pair. These are statements about the state or change of an
archetype and I do not see why you can't get the context as well?

Maybe I do not understand exactly what you mean.

Cheers, Sam

3

I agree with you Tim, I don't know why this code is being used in two
places - that should never be the case. The two meanings of 'deleted'
are not identical here anyway. I would think we should fix this.

- thomas

Tim Cook wrote:

4

Hi Tim

I would argue that terminology should be seen in context, in this case a
name value pair. These are statements about the state or change of an
archetype and I do not see why you can't get the context as well?

Maybe I do not understand exactly what you mean.

Hi Sam,

Thanks.

I believe you have done a very good job of helping me make my case.
Since they are in two different contexts, then they are two different
concepts. This should yield two different Concept IDs.

Cheers,
--Tim