openEHR-clinical Digest, Vol 56, Issue 3

Helo Pablo e Heather,

Recently happened the same with me: I’m validating some data in obstetric’s archetypes and some informations that were present in some archetypes that I already had downloaded before, were not present in CKM currently. I found out they were “withdrawn”.

This fact took our local researchers to ask why they were superseded since that some informations that were present are important and should be maintained in the currently version.

Exemple: in the archetype openEHR-EHR-EVALUATION.pregnancy.v1, that I had in my computer, talks about the membrane status during the labor, in cases if the rupture was spontaneous or artificial.

In fact, we realize that we do not have the option of membrane stay intact all the time in childbirth (which is what we want in the most natural process possible).

However, in the available archetype with the same information, the archetype openEHR-EHR-EVALUATION.pregnancy_summary.v0, doesn’t have anymore the information about the state of the membranes.

It is true that the current archetype brings better and clearer information than the old but I think some important information should not simply be deleted.

Anyway, it is only observations that Heather has already explained me and, once all our validation considerations have been completed, I will submit those contributions to CKM reviewers.

This is only the register, because this doubt and observation may be happening to other researchers as well.

Hi Danielle,

I suspect that the archetype you were using is this one, or similar – http://www.openehr.org/ckm/#showArchetype_1013.1.177_2

The current archetype is still the same artefact but it has evolved and had a name change in early 2016, plus reverted to v0 when our versioning rules changed for draft archetypes. The audit trail demonstrates all of the changes, but I can understand your frustration - http://www.openehr.org/ckm/#showArchetype_1013.1.177_REVISIONHISTORY.

Draft archetypes are only early candidates and so will very likely change during the review process – this is now why they are flagged with both a lifecycle state of ‘draft’ but also v0. Regard these as unstable and effectively a risk to use and implementation – they could change dramatically from when you download them to when they get published as scope, detail etc become refined through the peer review rounds.

Changes made to the draft archetypes are usually not made arbitrarily but because they have been requested by clinicians, including removal of the data elements you have identified! If you look at the audit trail you will see that the version after the archetype linked above was forked from the Australian CKM after there was a number of review rounds with clinicians – this Australian archetype was a major improvement on the previous international archetype. What is now missing is the details about the ROM – these would have been removed for a reason. If we view the review rounds in the Australian CKM we should be able to see that decision-making process – it should be transparent.

While the archetype remains unpublished, you have the opportunity to further improve it with requirements that you have identified. The archetype is still draft and there is still potential for change – please add these as a change request if you would like, but consider that if they have been removed previously, then perhaps the data elements need to be recorded in another archetype and are out of scope for this one. From memory, these data points were felt to be part of the recording of the labour details rather than the higher level overall pregnancy summary.

The intent is that the end result will be a family of archetypes that are used to record details of pregnancy, birth and postnatal period, each with specific scope and aiming to minimise gaps, minimise overlaps. It is not a simple process. Clearly the international archetype has not yet been agreed by clinicians as ready for publication/fit for use.

It is a tension that we have to manage – the need to evolve the archetypes vs understanding that people are using unstable archetypes. Your experience is not unique – it is not the first nor the last.

Unfortunately there is no perfect solution. So the compromise we have working at present is that the drafts will likely change, and usually for very good reasons – so user beware! But once published we will protect the archetype versioning very tightly, understanding that the downstream impact of subsequent change on modellers and implementers is significant and should be avoided where present.

So please engage with the process and participate actively – adopt the archetype so that you can be included in the review process when it is initiated, get notifications about any changes so that you can make decisions about how to manage those changes, make suggestions thought change requests, offer to act as an Editor during the review process etc.

Kind regards

Heather