Stream: Orders and Observation WG
Topic: Remaining Observation tracker to be Block Voted
Eric Haas (Jul 03 2018 at 15:55):
## Comment Submitters ##
• @Andrew Torres
•Casey Thompson
•Claude Nanjo
•Grahame Grieve
•Hans Buitendijk
•Mark Kramer
•Robert McClure
## Line Items ##
1. GF#16517 Haven%27t+clearly+sorted+out+patient%2C+subject%2C+and+focus+in+Observation (Mark Kramer) In Person
2. GF#17394 Change+%22magic+value%22+in+vital+sign+for+head+circumference+and+for+oxygen+saturation+to+methodless+concepts (Robert McClure) In Person
3. GF#15681 add+choice+of+duration+to+Observation.effective%5Bx%5D (Grahame Grieve) Not Persuasive
4. GF#16136 Observation.focus+-+should+this+be+a+modifier%3F (Claude Nanjo) In Person Not Persuasive
5. GF#16137 Observation.bodySite+-+why+only+a+code%3F (Claude Nanjo) In Person Not Persuasive
6. GF#16142 Observation+patterns+3+and+4+should+be+Condition (Claude Nanjo) In Person Not Persuasive
7. GF#16717 2018-May+Core+Norm+Observation+%2329 (Casey Thompson) Not Persuasive
8. GF#16721 2018-May+Core+Norm+Observation+%2330 (Casey Thompson) Not Persuasive
9. GF#16315 Observation+Status+-+Final+-+Clarification (Hans Buitendijk) Not Persuasive with Mod
10. GF#16371 Normative+observation+requires+a+non+normative+profile+to+be+implemented (Andrew Torres) Not Persuasive with Mod
11. GF#16134 Observation.basedOn+-+please+explain+further (Claude Nanjo) In Person Persuasive with Mod
12. GF#16138 Observation.specimen+-+allow+code (Claude Nanjo) In Person Persuasive with Mod
13. GF#16140 Observation.hasMember+-+QuestionnaireResponse (Claude Nanjo) In Person Persuasive with Mod
Eric Haas (Jul 03 2018 at 23:37):
@Lloyd McKenzie why pull GF#16136
I followed your instructions and rationale. The observation still stands even if it was a monkey! (This is a perfect test case for why isMod is too hard to apply evenly.) What if the patient was a monkey instead of a human does that make subject an isMod? Also there is no isModReason beside the tautological "it modifiers the interpretation of the results" so is non starter for me.
thanks for reviewing btw :-)
Lloyd McKenzie (Jul 03 2018 at 23:45):
It's not about "modifying the interpretation". It's about the potential to cause the interpretation to be false. If an observation of heart rate, blood pressure, blood type, etc. is made where the Observation.focus is "fetus" and the Observation.focus is ignored, the reader of the Observation will have the incorrect interpretation that they're looking at the heart rate, blood pressure or blood type of the mother. The fact that ignoring the element can lead to an incorrect interpretation is what makes the element a modifier.
Lloyd McKenzie (Jul 03 2018 at 23:46):
If you ignore the Observation.code, the Observation.subject, the Observation.value, etc., you'll have less information, but you won't have an incorrect interpretation of what occurred.
Eric Haas (Jul 03 2018 at 23:48):
1) what about the monkey analogy? I think that subject is just as important as focus based on the above rationale.
2) what it the non tautological ismodreason?
Lloyd McKenzie (Jul 04 2018 at 00:05):
1. What is the monkey analogy?
2. This element changes the overrides the default interpretation that the focus of the observation is the Observation.subject
Eric Haas (Jul 04 2018 at 15:33):
What if subject is another species or an environmental sample. The interpretation the observation is modified. I don't see how focus can be an isMod and subject is not an isMod.
Lloyd McKenzie (Jul 04 2018 at 15:55):
It doesn't always have to falsify the interpretation, it just has to have the potential to do so. Observation.status doesn't necessarily falsify the interpretation, but it can (if the status is entered-in-error). And because that potential exists to falsify the interpretation, you can't ignore it. Similarly, because Observation.focus has the potential to falsify the meaning, we have to treat is as a modifier. On the other hand, if you ignore subject, that can't falsify the meaning, regardless of what the subject is set to.
Eric Haas (Jul 04 2018 at 16:08):
I don't think the status and focus are the same. the Observation ( measure X : value y ) is nullified if entered in error. but the Observation still stands no matter the subject or focus. How the consumer interprets or assessed the Observation is based on the subject, focus, date, method, specimen, etc. I don't think we are talking about the same thing here and I think we have come full circle to my disposition. I do think that IF we keep isMod with all its flaws I would make .component, code.coding and our soon to be unveiled extension modifiedBy all isMods because they directly mess with measure X .... but that is a major discussion for another day soon.
Lloyd McKenzie (Jul 04 2018 at 16:18):
It's not a question of whether the Observation "stands". It's a question of whether someone who ignored the element could come to a false understanding of what happened. With status, they could come to a false understanding that the Observation had happened at all, when in fact it hadn't. With focus, they could come to a false understanding that the characteristic measured was of the patient, when in fact it was a fetus
Lloyd McKenzie (Jul 04 2018 at 16:19):
If we were going to mark component or code as modifiers, we'd have to show that ignoring them could lead to a false interpretation - what's your scenario for that?
Eric Haas (Jul 04 2018 at 20:03):
if you don't know the subject then that dread disease you cultured could be for a human patient, and envelope or an animal or your reference culture. So it is equally true that "someone who ignored the element could come to a false understanding of what happened" this is true for dates, methods, specimens too... So I think that is not the best way to frame the criteria.
Nullifying the results with status is a clear cut case. Measure X with result Y is no longer true in a given context.
This focus is modifier reason is not clear cut at all. I still assert is as critical to the interpretation as the subject and other elements because you ignore them at your own peril. But I am not suggesting we pile a bunch of isMods to the resource.
Lloyd McKenzie (Jul 04 2018 at 21:29):
There's no false understanding about unstated information. You know that the lab result found a dread disease and when. You don't know for whom. There's no false interpretation that can be made from the information provided.
Lloyd McKenzie (Jul 04 2018 at 21:31):
"Jane smith has blood type B+" is a false interpretation if the actual observation is "Jane Smith's fetus has blood type B+". But "Someone/something has blood type B+" is still an accurate interpretation if subject gets ignored.
Grahame Grieve (Jul 04 2018 at 23:24):
so, summarising a long thread from FHIR-I on this here:
Grahame Grieve (Jul 04 2018 at 23:25):
An element is a modifier if it cannot be safely ignored because its value, or it's meaning if missing, may cause the interpretation of the containing element or one of its descendants to no longer conform to the stated definition for the element
Grahame Grieve (Jul 04 2018 at 23:26):
by this definition, Observation.focus is not a modifier. However, we are concerned that focus is a surprising thing for implementers, and believe that OO have not been clear enough about this.
Grahame Grieve (Jul 04 2018 at 23:27):
in fact, I as Product Director, ask OO to be clearer about the fact that observations may be made on a non-subject focus
Grahame Grieve (Jul 04 2018 at 23:29):
OO could do so by one or more of the following:
- changing the definitions around so that focus becomes a modifier by the technical criteria above
- adding observation examples for the various cases where focus is used
- pointing out how focus works, with some examples of use, in the scope, with links to full observation examples
- adding a note to the safety page about not being caught out by focus
Grahame Grieve (Jul 04 2018 at 23:30):
- improving the definitions of both subject and focus to be clearer (e.g. currently, the signficance of focus is implicit in the definition of subject, and only explicit in a few words in the comments)
Last updated: Apr 12 2022 at 19:14 UTC
