[Openehr-13606] Some comments to the first proposal

ISO 13606 (archive)
1

Dear archetype and standards friends,

I very much welcome the initiative by Thomas Beale to collect views on a possible direction of such a revision.

It will be very important to seek co-operation and eventually consensus between in particular those stakeholders who have already implemented the present EN/ISO 13606 series.

Even if it is right that not all 13606 implementations will be plug and play compatible, this standard with profiles nationally (or in other contexts) has been very important and it will be important to cater for the already made investments while fixing obvious deficiencies.

I would very much appreciate if there could be real harmonisation between the openEHR extract model and the new formal standard nd as pointed out this may require changes in both specs.

While it is fine to have a preliminary discussion with this list, it will be necessary to soon turn to the official standards bodies, members of ISO and CEN (I think it should be run as a parallel CEN-ISO project to end up with completely aligned versions). Since it takes considerable time to formally initiate project in ISO we should prepare a proposal for a work item quite soon. This could of course be directed through the WGs but it will be important to have a country to fully back this up and provide a project leader. Can we count on UK and Dipak Kalra again?

With regard to the Datatype issue which always have been a difficult one, we should recall that even ISO 13606:2008 refers normatively to CEN/TS 14796:2004, Health Informatics — Data Types even if it is subset thereof. It is hardly correct to assume that it was ISO 21090 which was far from approved at this time. I would like to see a harmonisation here between openEHR and ISO 13606 datatypes as more important than aligning with other HL7 specs.

The updating of part 2 to ADL 1.5 will hardly be difficult.

May I also remind us about the existence of the other three parts. I think all of them should be reconsidered,

Kind regards

Gunnar

2

Gunnar,

thanks for this. We will be relying on people with long involvement and experience with process to help us get things happening in the right way. I have taken the liberty of cc:ing Dipak and also Stephen Kay, since I don’t think they are subscribed, but will hopefully be interested in your UK suggestion. (List subscription here).

On the subject of data types: I think that the 21090 question has been overtaken by events in HL7 - by the FHIR data types - part of HL7’s ‘fresh look’. These are in the early stages, but probably should not suffer from the problems of 21090/HL7 data types(subtractive profiling, over-complexity), and include some openEHR semantics. They are not currently expressed in UML, but I see no reason why they should not be. Alternatively, do we want to try and sell ISO the openEHR data types? I suspect that would be hard work… anyway - this is one of the issues to be debated.

As mentioned in my other post: the wiki page is here - I think the subscribers of this list should feel free to add to it their comments directly (maybe with initials, so we know whose bit is what). We can clean it up as we go along.

  • thomas