Thank you, @mcordell, for bringing up a complicated but very valid question.
It was myself, on behalf of the Norwegian NRC, who originally requested the lateral promotion of the Australian concepts regarding cervical HPV screening. We received a request from clinicians at a regional hospital for concepts representing cervical screening tests of different kinds. The clinicians are dependent on precise and correct concepts to ensure the proper billing, statistics and treatment route for the patients in question.
I want to add my support especially to the points you’ve made about OE vs. EP - the editorial advice for evaluation procedures has essentially been “don’t make or use them” for several years now, but they aren’t practically nor in principle the exact same as observable entities, in my opinion. We have several use cases, both in modeling new concepts and also representing procedures in EHRs, that aren’t really supported by the observable entity hierarchy design and logical definition.
We’ve felt the need to create several evaluation procedures in our national extension where we’ve deemed them necessary and appropriate. We still feel that the guidance for how end users (or even edition/extension managers) should practically use OEs rather than EPs is lacking. The current guidance thoroughly covers lab implementation, but we struggle with the needs of f.ex. the nursing community in describing nursing procedures.
Additionally, a lot of the evaluation procedures are, as you point out, complex and yet necessary in clinical practice. In a perfect world we could express all the components of how a clinician interprets information, but that would probably involve a combinatory explosion that most likely would do more harm than good. At the same time, we need to express the procedures in a way that is understandable and useful. I’d rather have primitive or minimally modeled concepts than nothing at all in the cases where “just using an observable entity” isn’t really an option.
Furthermore, I agree with the reflections regarding screening procedures. The contextual differences between a screening procedure and a diagnostic procedure are internationally aligned as far as I’ve been able to find, and the context of why a procedure/test was performed will impact the follow-up as well as inform registries and national programs in different ways.
In maintaining a terminology of this size, rules need to be stable and mostly quite strict. I completely understand and empathize with the need for logical models and adherence to principles and rules, and the burden that is already on the authoring team’s shoulders connected to this work. However, I hope that as both the people working with SNOMED CT and the terminology itself grows, we don’t immediately dismiss a discussion just because we’ve had it before.
The world around us is changing, the clinical needs change at a breakneck speed, and I hope we can ensure that our main focus is always to meet those needs to the best of our abilities and within the ontological principles deemed valid and appropriate at any given time.