20:24:15 From Grahame Grieve : sorry - the noise is me - I am in an airport corridor and it's loud. I will try not to say anything but may comment here 20:27:33 From Grahame Grieve : good thing that one's just an example... 20:32:44 From Grahame Grieve : that v2 table is crap. we should just reference snomed? 20:33:05 From Deborah Drake; NHS Digital UK : I agree with Graheme G 20:33:09 From Rob Hausam : yes 20:33:12 From Deborah Drake; NHS Digital UK : SNOMED CT is much better here 20:35:33 From Grahame Grieve : should we have a more limited value set for a specimen additive? 20:36:18 From Jim Case : The concept model is not yet documented in the ed guide. It is under devleopment...sorry 20:36:26 From Deborah Drake; NHS Digital UK : Possibility of international reference sets depending on context of use? 20:37:05 From Jim Case : Only a list: 20:37:07 From Jim Case : Specimen Specimen procedure Specimen source identity Specimen source morphology Specimen source topography Specimen substance 20:42:47 From Grahame Grieve : Daniel I'm interested in your last comment about structureal equivalence - could you design 2 instances of the specimen resource that say the same thing? I bet that if you could, I could capture logic in the FHIR spec that allows anyone to capture that - maybe an exercise for us later 20:45:24 From Jeremy Rogers : There *is* relevant content in SNOMED, its just not all under one roof 20:48:21 From Jeremy Rogers : <<706046003 has a lot of the obvious suspects... 20:48:38 From Grahame Grieve : I should have said "would you do that" rather than 'could you' - followed your logic, yes. Just, if you actually do this in practice, it will give us something solid to cut our teeth on. But this is homework for you and me 20:52:13 From Grahame Grieve : The comments here - are those an international standard? (they match Au ok, but is that for all of us?) 20:53:46 From Rob Hausam : I doubt that it is entirely international 20:54:09 From Grahame Grieve : if that's manufacturer specific, HL7 shouldn't really have comments like that? That's a v2 issue outright... 20:54:10 From Deborah Drake; NHS Digital UK : Yes I agree 20:54:31 From Grahame Grieve : lol 20:57:40 From Jeremy Rogers : If the specimen is obtained from an implant or other device - e.g. a catheter specimen of urine - what's value should appear in the Specimen.bodySite element? The FHIR model seems to compile in an assumption that all specimens come from a body site, and not potentially from a device. Including, in some cases, from a prosthetic device that has not yet been implanted...? 20:58:56 From Grahame Grieve : we think that prosthetics etc are body site, I think 20:59:24 From Jeremy Rogers : So that would need the terminology binding for that element to be broader than pure anatomy 20:59:54 From Jeremy Rogers : And include e.g. tubes from an external heart-lung bypass machine 21:00:24 From Jeremy Rogers : And the crud on the floor in the corner of the theatre contaminated with MRSA... 21:01:22 From Jeremy Rogers : So calling the element 'bodySite' seems rather inappropriate! 21:01:27 From Jim Case : I think we need separate out the analyte being tested from the device used to collect it, implanted or not 21:01:49 From Jim Case : Thus a urine specimen from a catheter is still a "bladder" specimen 21:01:57 From Jim Case : source that it 21:02:17 From Jim Case : I mean the source... 21:02:30 From Jim Case : it is not a catheter specimen 21:02:34 From Deborah Drake; NHS Digital UK : same with a blood bag specimen post tx reaction 21:02:44 From Jim Case : agree with Debbie 21:02:46 From Deborah Drake; NHS Digital UK : agree with you Jim 21:03:29 From Jim Case : That then is a device specimen 21:03:47 From Deborah Drake; NHS Digital UK : yes agree, need to gather some principles as we go along 21:04:03 From Jim Case : ALways commwnting on you 21:04:17 From Jim Case : this is really hard...so it needs discussion 21:05:04 From Jim Case : one of the things that you need to overcome is the particular idions used in different domains, in order to get to the "real" specimen 21:05:16 From Jim Case : idioms...fumble fingers 21:10:48 From Jeremy Rogers : In Ye Olde SNOMED Concepte Model (circa 2009), the 'valueset' for the specimenSourceIdentity (which, I think, is the SNOMED equivalent of Specimen.collection.bodySite here) was <<125676002 OR <<133928008 OR <<35359004 OR <<276339004 OR <<49062001 21:11:10 From Jeremy Rogers : Or, more human readably: 21:11:11 From Jeremy Rogers : -------------------------------------------------------------------------------------------------------------------------------------------- SCT INTENSIONAL CLUSTER DEFINITION (1.3) FOR SPECIMEN_CONTAINERS Containers for passing specimens -------------------------------------------------------------------------------------------------------------------------------------------- INCLUDE: << 125676002 Person (code and all its descendants) << 133928008 Community (code and all its descendants) << 35359004 Family (code and all its descendants) << 276339004 Environment (code and all its descendants) << 49062001 Device (code and all its descendants) EXCLUDE: -< No exclusion clauses >- -------------------------------------------------------------------------------------------------------------------------------------------- 21:11:28 From Jeremy Rogers : (ignoring the inappropriate header...)