I’m relatively new to working with openEHR and I’m currently engaged in the process of creating templates that are suitable for representing the official German oBDS dataset. While existing templates from the Use Case Oncology aren’t entirely applicable, I am using them as a starting point for building my templates. I’ve encountered some challenges along the way where I’m unsure about the best course of action.
For instance, I’m developing a new CLUSTER archetype for the prostate module, which is being reused in many templates I’m planning to create. In this module, I need to record the primary and secondary grades, as well as the score result of the Gleason Score. There’s an existing CLUSTER archetype for the Gleason Score, but it’s only available in English. While I can create a SLOT(CLUSTER) within my CLUSTER to potentially insert the existing Gleason Score, the lack of a German translation poses an issue for me.
I’m seeking guidance on how to proceed in this situation. Should I consider translating the existing “Gleason Score” archetype into German and then utilize it, or are there other alternatives worth exploring? (I believe the translation shouldn’t be too complex, as this archetype comprises only 5 items.) Converting my own templates or the CLUSTER to English doesn’t seem practical, as my goal is to specifically align with the German oBDS dataset.
I truly appreciate your time and assistance in this matter.
I was the original author of Gleason archetype , and although it is now fairly old, I suspect it still stands up fairly well. On that basis, I think a German tranlsation is likely to be helpful and not ‘wasted’ by a lot of changes to the archetype, particulary if you stick to dataitems and only do minimal work on Use/ musues aspects ( fro now).
I am aware of quite a few openEHR modellers who are getting into cancer dataset modelling and so there probably will be some drive to get this one actively reviewed and published.
The only slight hesitation I have is that there might be some discussion about whether this merits being an Observation, rather than a cluster. However that should not really affect the translation per-se. I’d be interested in others’ views
Hi, and welcome to this wonderful community, Herwig
As far as I know, any user can translate an archetype. Click on the “Details” button and then “Translate archetype”. Next, click on the button “Checkout Archetype”. Now you have a branch you can work with. Chose the language you want to translate to, and go ahead.
Note that you need an editor to commit the archetype back to the trunk, or else it will stay in your branch.
Also note that Gleason Score Clinical Knowledge Manager, still is in Draft status, and never has been on review. It might be changed, possibly substantially, ahead of or during the review process before becoming in status Published. It is not safe to use any archetype without it being Published.
It is about to be used also in Norway, and @lekkim has a plan to move it forward towards publishing. It would be smart to finish the translation to German and Norwegian first.
Regarding the oBDS datasets, for a direct look at the current schemas, you can visit: basisdatensatz.de/xml/.
Additionally, you can access the details about the Gleason Score (with some limitations, as this field in the XML actually comprises three data entries: Primary Grade, Secondary Grade, and Score Result) here.
@lekkim - has there been any discussion about Gleason as an Observation rather than as a Cluster? My preference is still with Cluster but I can imagine other arguments.
Working on it. Theory for now is to keep as CLUSTER as this seems to be grading of one biopsy sample, and will be SLOT’ed in Lab/Histopath result OBSERVATION. Another biopsy might be scored differently.
One of the challenges we are facing is the need to handle cancer reporting that often works at a higher level e.g there is a field for ‘staging’ that is agnostic of any specific TNM or even other staging like FIGO, with similar issues around local invasion, tumour margins and nodal spread.
Welsh colleagues have suggested possibly using a generic ‘Staging’ Cluster to capture that high-level perspective / valuesets etc and drop the detailed TNM/ other staging down into a cluster slot . They also have a way of generating TNM per-cancer specific valuesets from the cancer type which they are willing to share.
Is this worth a bit of a remote ‘show and tell’ as I know others like Catalonia, Sweden are working in this territory?
Perhaps a show and tell might be warranted - but we’d need to get help with the explanation for the clinical bit.
As @varntzen noted, we do have a use case where the need to represent a ‘percentage of positive total biopsy cores graded at Gleason grade group 4 or above’ level - and for that use case keeping this as a cluster makes sense.
