The mapping doesn’t specify the Observation code , which is mandatory / essential element . Without it mapping race info to Observation barely makes any sense
PID race-to-observation
Which code you use depends on what country you’re in. In the UK, they’d use SNOMED. In the U.S. and much of the rest of the world, you’d use LOINC. Some places might use other code systems. Because v2 is used internationally, we can’t dictate which code is used. (The FHIR spec only dictates Observation.codes for a very limited set of vital signs, and even that has created considerable angst.)
I’ve tried to use the below Loinc codes as Observation.code .
None of them pass validation ( FHIR R4), it fails with
Unable to validate code http://loinc.org#46463-6 - Code can not be found in CodeSystem for ‘http://loinc.org#46463-6’ .
| LOINC_NUM | COMPONENT |
|---|---|
| 46463-6 | Race or ethnicity |
| 59362-4 | Race or ethnicity |
| 64606-7 | Race - 2 character code |
| 64987-1 | Race or ethnicity |
| 67288-1 | Race or ethnicity |
| 67755-9 | Some other race alone |
| 76362-3 | Race or ethnicity panel |
| 76663-4 | Race or ethnicity |
| 86800-0 | Coded race and ethnicity - recommended CDA R1 sections |
Are you seeing any messages about problems accessing the terminology server when you run the validator?
no, no messages about access. moreover if I use another code 10018-0 ( it has nothing to do with race/ethnic info) it passes the validation
All of the above listed LOINC codes (I don’t know what the image in the last line is supposed to be) are present and respond appropriately with the $lookup operation on the tx.fhir.org FHIR build terminology server, so the FHIR Validator (and IG Publisher) should be able to validate them. Many of the descriptions aren’t the same as the ones listed above, and in some contexts that could generate warnings or errors, but the codes themselves should all be valid.