Dear All,
I've been looking for some time for ways of injecting knowledge into the ecosystem so that it is available to an EHR, but also to other systems that might want to use it.
I currently think I need to create a terminology (or maybe more than one), but I've found vanishingly few open tools and little guidance on what I could use to do this, and experiment to see if it does what I need.
I'd be grateful for any advice...
Yours,
Matthew
Dear Matthew,
In the attachment a candidate terminology as PDF
It is known as SIAMM
Semantic Interpretabily Artefact Modelling method
Gerard Freriks
+31 620347088
gfrer@luna.nl
Kattensingel 20
2801 CA Gouda
the Netherlands
Dear Gerard,
Many thanks – that’s interesting as an expression of the terminology, but I’m guessing that was either generated from a machine-readable expression of the terminology, or would need such a machine-readable version to exist before it could be instantiated within a terminology server/service? I’m also interested to try that out, so that other software might be able to interrogate the resource…
Yours,
Matthew
Matthew, what is the scope of your terminology? Are the terms intended to appear in data instances? If terms are intrinsic to a set of archetypes then you could probably define the terms as constraint bindings in each archetype.
Dear Diego,
I don’t know precisely yet, but the problem space arises when, for example, a procedure is standardised (by a protocol or SOP), and thus has a set of agreed “allowed” outputs which one might need to record in the kind of way that would fit with mapping to a DV_CODED_TEXT from an archetype. When this goes finer-grained than any existing terminology (for example, when the basis is something like a locus-specific mutation database) there needs to be a way to create pick-lists etc. without embedding the logic inside an archetype.
My current thinking is that the protocol/SOP maps to 1…* archetypes, but the knowledge needs not to be embedded there because it has broader uses.
Does that make sense?
Yours,
Matthew
Dear Gerard,
Many thanks – that’s interesting as an expression of the terminology, but I’m guessing that was either generated from a machine-readable expression of the terminology, or would need such a machine-readable version to exist before it could be instantiated within a terminology server/service? I’m also interested to try that out, so that other software might be able to interrogate the resource…
Yours,
Matthew
HI Matthew
You may be interested in HL7 InfoButton - which is a standard for linking clinical systems to knowledge bases – it is a fairly simple and open REST interface – allowing you to define the sorts of search parameters that you are expecting - and retuning the knowledge in a few possible formats.
Whether this will work for you depends upon how you anticipate your knowledge being accessed - and whether the process will be triggered from within the EHR or from some other user interaction.
I have not used it but it does not seem too tied to other HL7 standards, and so may also be of use and relevance in an OpenEHR context. At the very least it will be an interesting example of folk trying to do something similar to you 
The following page looks like a good place to start — http://www.openinfobutton.org/hl7-infobutton-standard
all the best
Charlie
Dear Charlie,
That’s interesting, and not something I’d come across – many thanks!
Is anyone aware of an open, freely available terminology service suite that could be used to experiment with what should be in a terminology in order to make it useful for various purposes..?
Yours,
Matthew
Hi Matthew,
Great to hear from you. https://wiki.nci.nih.gov/display/lexevs/lexevs worked for me back in 2009 or so, it is apparently still alive. maybe worth taking a look at.
All the best
Seref