Stream: implementers
Topic: codes in Vital Signs
David Hay (May 24 2016 at 19:56):
The profile for Vital signs specifies 2708-6 as the code for oxygen saturation as it's methodless. What's the most appropriate behaviour if you have a value with the code 2710-2 or 59408-5 ?
Eric Haas (May 24 2016 at 21:25):
So this needs to be explained better but I'm having a hard time finding the right words. The code you need to use is 2708-6 but you can put in other codes as (such as 2710-2 or 59408-5 ) as translations. We chose the methodless as most generic so all the translation should map from more specific to less specific.. BTW the O2Sat code will change to the pulse ox code. just haven't done that quite yet.
David Hay (May 24 2016 at 22:38):
so CC doesn't really fomalize a translation (AFAIK) - it simply allow multiple codings within a cc. so if I have a 2710-2 in my data store, I can certainly include that as a coding, but should I also include a 2708-6 with the same value? (ie 2 codings) marking the 2710-2 as 'userSelected'?
Erich Schulz (May 24 2016 at 23:02):
so if user choosers a code from system A, wouldn't the only motivation for including a mapped code from system B be that A was non-standard?
Grahame Grieve (May 25 2016 at 21:26):
erich there is many other motivations
Grahame Grieve (May 25 2016 at 21:28):
David it allows multiple representations of the same concept. 2710-2 and 2708-6 are the same concept (in this case). I don't think you'd mark either of them as 'user selected' - unless the user specifically selected it
Erich Schulz (May 25 2016 at 22:59):
i'm curious about the other motivations - I guess A maybe standard but there is an expectation that receiving systems may not understand the standard.
Eric Haas (May 26 2016 at 00:04):
you 'formalize" the translations using concept map which would reference the value set(s).
Erich Schulz (May 26 2016 at 00:40):
Indeed @Eric Haas which makes the motivation for including multiple codings even more interesting...
Grahame Grieve (May 26 2016 at 01:17):
it's not a binary thing. In the standard, we have minimal requirements but they may understand other representations better
Rob Hausam (May 26 2016 at 19:12):
point of clarification - even aside from the deprecation of 2710-2, I don't think it's true that "2710-2 and 2708-6 are the same concept":
2710-2 Deprecated Oxygen saturation in Capillary blood by Oximetry
Class/Type = HEMODYN.MOLEC/Clinical
2708-6 Oxygen saturation in Arterial blood
Class/Type = CHEM/Lab
it's not only the method (or lack of one in 2708-6) that is different, but also the system (specimen) and class/type
and the LOINC-suggested replacement concept for 2710-2 is also different:
59408-5 Oxygen saturation in Arterial blood by Pulse oximetry
Class/Type = PULM/Clinical
Grahame Grieve (May 26 2016 at 19:29):
well, it's a matter of perspective. Clearly, from the perspective of the terminology (LOINC) these are not the same thing - that's why they've defined different codes for them. But from the point of this particular context - that is, representing the test that was done - both codes are representations of the same concept : this is the test that we performed
Rob Hausam (May 26 2016 at 22:41):
I agree that it's a matter of perspective - as far as the equipment and the procedure used is concerned, it almost certainly doesn't matter whether the specimen is capillary vs. arterial blood - but LOINC does define separate test codes for both, and that's the way that they're typically ordered by clinicians, too (at least in most systems that I've seen/used)
so should we draw any particular conclusions from these facts?
Grahame Grieve (May 26 2016 at 22:43):
well, the main conclusion is nothing new, but maybe we could state it more crisply: what concepts are appropriate to use in a set of codings is very context sensitive. but we already do say:
The concept may be coded multiple times in different code systems (or even multiple times in the same code systems, where multiple forms are possible, such as with SNOMED CT). The different codings may have slightly different granularity due to the differences in the definitions of the underlying codes
Grahame Grieve (May 26 2016 at 22:44):
.. different granularity... is the key here
Rob Hausam (May 26 2016 at 22:44):
right
Grahame Grieve (May 26 2016 at 22:46):
in other cases, mostly you order with a very coarse granularity. in this case, you might even order with a coarse granularity ("O2SAT") but you'd order from different sources. Actually you'd order at an even higher level of granularity (ABGs, or 'do vitals')
Rob Hausam (May 26 2016 at 22:48):
yes - LOINC has always tended to struggle with the ordering level of granularity, although that's improved to some extent
Erich Schulz (May 26 2016 at 22:50):
they are very different pulse oximetry and ABG SO2...
Erich Schulz (May 26 2016 at 22:51):
a pulse SO2 is taken one way and measure effectively the SO2 in the pulsing blood...
Grahame Grieve (May 26 2016 at 22:52):
right. but 98% has the same meaning in both, as does 60%
Erich Schulz (May 26 2016 at 22:52):
all kinds of samples get put into an "ABG" machine (ie venous, cord venous, cord arterial)
Grahame Grieve (May 26 2016 at 22:52):
some numbers in the middle it might make quite a difference
Erich Schulz (May 26 2016 at 22:53):
venous and cord blood samples mean completely different things
Rob Hausam (May 26 2016 at 22:53):
yes - no doubt about that
but is one obviously and necessarily possible as a "vital sign", and the other isn't?
and is pulse oximetry always the only method that can ever be used for an "O2 SAT" vital sign?
Erich Schulz (May 26 2016 at 22:53):
ie reference to the sample becomes critical
Erich Schulz (May 26 2016 at 22:54):
AFAIK @Rob Hausam - the qualifier maybe location
Erich Schulz (May 26 2016 at 22:54):
(ear, finger, toe)
Erich Schulz (May 26 2016 at 22:54):
in 1:10000 patients that maybe material...
Grahame Grieve (May 26 2016 at 22:54):
right. but I don't know why anyone would bother with VBG 02Sat. And with blood gases, arterial/venous is so significant, I always saw the specimen type pre-coordinated in the test
Rob Hausam (May 26 2016 at 22:55):
that's why the new codes that LOINC and we chose are for arterial, rather than capillary blood
Erich Schulz (May 26 2016 at 22:55):
well then there is "mixed venous O2Sat" (ie pulmonary artery)
Erich Schulz (May 26 2016 at 22:56):
it analogous with BP measured manually, by automatic cuff and via a arterial line...
Erich Schulz (May 26 2016 at 22:57):
ultimately you maybe measureing the same physiological parameter but the interpretation (and artifeacts and qualifiers) can be quite different
Erich Schulz (May 26 2016 at 22:58):
(mixed venous was trendy with the "surviving sepsis" campaign but I think its never really taken off)
Grahame Grieve (May 26 2016 at 22:58):
good
Rob Hausam (May 26 2016 at 22:59):
exactly - but so far, as far as I know, no one has proposed to consider "mixed venous O2Sat" as a "vital sign"
but if you could measure it non-invasively somehow, maybe someone would propose it (I think it will stick with arterial, but who knows for sure)
Erich Schulz (May 26 2016 at 23:00):
cant see it being popular anytime soon (MVSO2 that is)
Rob Hausam (May 26 2016 at 23:00):
I agree - can't imagine it ever would
Rob Hausam (May 26 2016 at 23:01):
but I think what we don't want to do is lock ourselves into particular methods - if we can avoid it
Erich Schulz (May 26 2016 at 23:01):
isn't 59408-5 the one you are looking for @Rob Hausam ?
Rob Hausam (May 26 2016 at 23:02):
that one specifies the method as pulse oximetry
Erich Schulz (May 26 2016 at 23:02):
well that is how its done...
Erich Schulz (May 26 2016 at 23:02):
if your looking at a vital sign panel that seems to work
Rob Hausam (May 26 2016 at 23:04):
that is how it is done - today (and tomorrow as far as we can see at the moment)
but why should we mandate that?
there could be a new method to do that next month or next year
Erich Schulz (May 26 2016 at 23:05):
erm...
Erich Schulz (May 26 2016 at 23:05):
that's a pretty big "could"
Erich Schulz (May 26 2016 at 23:06):
the point here is to differentiate it from ABG measures
Erich Schulz (May 26 2016 at 23:07):
i do think 59408-5 is what you are looking for...
Rob Hausam (May 26 2016 at 23:07):
well, that's the discussion
Erich Schulz (May 26 2016 at 23:07):
without bleeding a patient I really cant see another method
Rob Hausam (May 26 2016 at 23:08):
of course you can't - until someone comes up with one
what about oxygen diffused from the skin surface? - maybe someone could measure that?
Erich Schulz (May 26 2016 at 23:09):
they do that new borns...
Erich Schulz (May 26 2016 at 23:09):
its hopeless
Rob Hausam (May 26 2016 at 23:09):
there you go - that's not pulse oximetry
Erich Schulz (May 26 2016 at 23:10):
and the interpretation would be completely different
Erich Schulz (May 26 2016 at 23:10):
exactly...
Erich Schulz (May 26 2016 at 23:10):
relax be happy :-)
Erich Schulz (May 26 2016 at 23:10):
59408-5
Erich Schulz (May 26 2016 at 23:10):
(oops enter twice)
Rob Hausam (May 26 2016 at 23:11):
if you say so ... ;)
Rob Hausam (May 26 2016 at 23:11):
thought maybe you were trying to emphasize your point!
Erich Schulz (May 26 2016 at 23:11):
heh
Erich Schulz (May 26 2016 at 23:12):
(still getting used to zulip)
Robert McClure (May 27 2016 at 15:52):
Not sure what the conclusion was here - but @Grahame Grieve you've made some comments that are dangerously incorrect in this thread where you assert that "both codes are representations of the same concept : this is the test that we performed" as in: "well, it's a matter of perspective. Clearly, from the perspective of the terminology (LOINC) these are not the same thing - that's why they've defined different codes for them. But from the point of this particular context - that is, representing the test that was done - both codes are representations of the same concept : this is the test that we performed". By saying that "both code are representations of the same concept" you are egregiously misusing the word "concept". What you mean to say is "the meaning converyed by the entire resource is not dependent on the detailed meaning of the specific test used." Correct?
Grahame Grieve (May 27 2016 at 21:29):
can't agree with the words you've put in my mouth there, sorry. But I suspect probably don't disagree about the intent, just the language.
Mark Jones (Jun 11 2021 at 18:18):
http://hl7.org/fhir/R4/observation-vitalsigns.html lists the LOINC as previously 59408-5 but now 2708-6. However, the US Core doc at https://www.hl7.org/fhir/us/core/StructureDefinition-us-core-pulse-oximetry.html still indicates 59408-5 and Inferno tests for that too. Are US Core and Inferno purposely trailing the main FHIR spec?
Eric Haas (Jun 11 2021 at 18:49):
No US Core is based on FHIR R4. Check the versions
Eric Haas (Jun 11 2021 at 18:51):
Also read the implementer notes in the us core profile. Because if profile inheritance you supply both
Yunwei Wang (Jun 12 2021 at 01:37):
2708-6 is Oxygen saturation in Arterial blood
59408-5 is Oxygen saturation in Arterial blood by Pulse oximetry
Last updated: Apr 12 2022 at 19:14 UTC
