OHDSI Home | Forums | Wiki | Github

Diabetes: condtion or measurement domain?

I am mapping ICD9 to standard concepts. I have a few diabetes terms that map non-standard SNOMED terms as a condition. For example 250.03 maps to concept_id 44821787 with domain_id = ‘Condition’. However 44821787 is not a standard concept. Next step is to find the standard concept for 44821787 in the relationship table. 44821787 maps to
40484648. However 40484648 has domain = Measurement in the concept table. Is there a bug in the table or a bug in my logic?
250.03 (44821787) is Diabetes mellitus without mention of complication, type I [juvenile type], uncontrolled
40484648 is Type 1 diabetes mellitus uncontrolled

What version of the vocabulary are you using?

When I look up ‘250.03’ on the public version of ATLAS here (
http://www.ohdsi.org/web/atlas/#/concept/44821787), I see it mapped to two
standard concepts, and not mapped to any non-standard concepts.

On your atlas link, the details page shows 250.03 has a concept_id of 44821787 and a domain of Condition. I am calling that a “non-standard concept mapping” as it lists clearly as non-standard. (Perhaps its not a mapping as much as an assignment?). From there the related concepts are 201254 and 40484648 . Of the two, 40484648 is a closer match by name and is listed as domain = Measurement.

The concept of 44821787 is the concept for the ICD9 code itself, and that
code is definitely non-standard, since ICD9 is not a standard vocabulary.
That code maps to two standard concepts in SNOMED, CONCEPT_ID (201254,
40484648), both of which are in the domain of ‘Condition’.

Why is atlas showing 40484648 as domain = Measurement?
http://www.ohdsi.org/web/atlas/#/concept/40484648

Ah, I see now, yes, that’s a good question. @Christian_Reich, can you
share with us the legacy of how the standard SNOMED concept of ‘Type 1
diabetes mellitus uncontrolled’ got classified within the Measurement
domain?

That’s a bug, and a new one. I’ll put it in.

Well, it really is a SNOMED “feature”. 40484648 is a descendant of 443238 “Diabetic - poor control”, and that is a descendant of 40480457 “Evaluation finding”. This is wrong. Diagnoses shouldn’t be listed there. I will report it to SNOMED, and in the meantime put it into Condition by force.

Thanks for the catch.

While you are in there; same issue for 40482801 Type II diabetes mellitus uncontrolled

Thanks, @Lori. Will do.

t