DSM-5 archetypes

Does anyone know of any archetypes that model DSM-5 (Diagnostics and statistics manual)?
Ideally a cluster for the additional details slot of EVALUATION.problem/diagnosis.
@sebastian.iancu maybe?

I could not find anything searching “DSM” on ckm, or the 4 national CKMs or google.

I’m not all that familiar with DSM but from my understanding most of it should fit neatly into core problem/diagnosis name.
Can you give us an example of where we would need something more than the name of the ‘diagnosis’?

1 Like

Our custom (dutch) ADL2 archetype: openEHR-EHR-CLUSTER.dsm_classification.v0.0.1.adls (6.3 KB)
DSM classificatie - Nedap Archeditor.pdf (93.0 KB)

DSM-5 (and perhaps also DSM-4) is a classification used in mental care - see also a human readable version at DSM-5 - Wikipedia

We currently have a diagnosis/DSM-5 archetype in use by our clients, I estimate that is around 80% overlapping your archetype. The ICD-10 codes we use it as terminology and bindings. From my point of view that GAF-score should be a separate archetype (but we did not implemented that split yet). My wish/goal (for NL at least) is to have this concept replaced by something compatible with a Problem ZIB (problem/diagnosis archetype) - this needs to be worked out.


My first impressions of DSM is that it can comprise clinical information from various axes, often data we would record independently, such as a Problem/Diagnosis or a new OBSERVATION for the GAF score.

Perhaps we need a ‘meta EVALUATION’ archetype that explicitly links to each of the archetypes that contain data for each of the 5 axes. We could use an explicit DV_EHR_URI data type to represent each required element, allowing one or more links to other archetypes that carry the data for each Axis, as required.

For example, one or more Problem/diagnoses (current EVAL) for each of Axes 1,2 & 3 (+/- current CLUSTER P/D qualifier to carry ‘Main’ status); and another for Axis 5 that links a new OBSERVATION for each GAF assessment/measurement against Current/Start/Max/??End as events. Not sure about Axis 4 for now.

I need to do more investigation and thinking. Until then, am I heading in the right direction?

1 Like

Happy this is being picked up by the senior modellers:)

A few thoughts:
DSM is a classification (originally for research) that’s (unhappily) used as a diagnosis coding (like ICD). So the match with problem/diagnosis is controversial.

I agree that in the openEHR world it makes sense to record every individual classification as a separate (problem) EVALUATION, but I believe psychiatrists will see a DSM classification as a single entity. (Do we have any psychiatrists in our community?)

Our current approach was to record a clinical description of the diagnosis in a problem/diagnosis evaluation and add repeating clusters per DSM classification. The current cluster is just a list of the fields we need for this, not too much thinking has gone into the design, so very much open to suggestions or a participating in a thorough approach.

Hi Sebastian thanks for weighing in. I like your goal and I probably agree on GAF. Would be nice to collaborate. Would you be willing to also share your archetypes?

I was about to say that we have a GAF archetype, but it turns out that I misremembered. We do have a Norwegian-only archetype for “GAF-S”, though, which is a Norwegian variation where symptoms and functioning are split into two separate scales. A new international GAF archetype could well use this as a blueprint, though: https://arketyper.no/ckm/archetypes/1078.36.1334

1 Like

Your proposal/thoughts with meta Evaluation sounds good to me, it is also very close to what I had in mind, as in NL we are also “encouraged” to adopt HCIM/ZIBs. In my opinion DSM-5 should use or refer to Problem, which in turn is quite close to openEHR problem/diagnosis archetype.


@joostholslag, i will first need to checkout the last version from our customers. It is anyways not that different than yours, and not using ZIBs yet.

1 Like