HCPCS K1034 Maps to two concepts, one in the type concept domain

[clairblacketer]

Describe the problem in content or desired feature
The HCPCS code K1034 maps to two concepts. The first concept is in the measurement domain representing a measurement for covid and the second concept is in the type concept domain representing a type concept for patient self-test. I don't understand why this is post-coordinated, other concepts for SARS-Cov-2 testing are not post-coordinated. The other issue is mapping to a type concept. Type concepts are meant to indicate the provenance of a record. In this case, it is reporting that a person provisioned a self-test, but the record itself is not the result of a patient self-test. A type concept is not meant to be used as a domain concept id (measurement_concept_id, procedure_concept_id, etc.) The only place to put this is the observation table. I think the larger issue though is that type concepts are really related to provenance, not to represent the data of the record itself.

How to find it
https://athena.ohdsi.org/search-terms/terms/801258

Expected adjustments
Please update the mapping such that a type concept is no longer used or pre-coordinate this mapping so that the HCPCS code maps to one standard concept.

[Alexdavv]

Hi @clairblacketer

type concept domain representing a type concept for patient self-test. I don't understand why this is post-coordinated

This is what we introduced with the COVID conventions. To address some features that people didn't even think about before COVID, we started mapping concepts to Visit, Type and Status. The 3rd we later rollbacked and replaced with COVID diagnostic criteria.

other concepts for SARS-Cov-2 testing are not post-coordinated

I think they all are except in many cases we don't know the details.

Type concepts are meant to indicate the provenance of a record. In this case, it is reporting that a person provisioned a self-test, but the record itself is not the result of a patient self-test.

I think it's both self-provisioned and self-tested where we introduced the latter as the hierarchical descendant of self-reported which was the eligible source of information.

A type concept is not meant to be used as a domain concept id (measurement_concept_id, procedure_concept_id, etc.) The only place to put this is the observation table.

This was not the idea. The idea is exactly to direct the type concept to the _type_concept_id_field.

Overall, it's pretty much the same "usual ETL scripts can't do that" problem that we faced in HCPCS/CPT4 visits mapped to the Visit Domain concepts. Want to rollback this thing also? ☺️

[MelaniePhilofsky]

Ah, the intricacies of the US billing codes. So much fun :)

This is an interesting HCPCS code. The code refers to a COVID-19 home test kit. These were distributed by the US government at no cost to persons living in the US for a few years. I am going to go a completely different direction with this and argue these are Devices. It isn't any different than a glucometer. Just because someone has one doesn't mean they used it to do a test / measurement.

HCPCS codes are "billing" codes. In the US, providers (very loose term) use these codes to bill insurance companies for services and products rendered. This particular code describes a product / device given to a person. It is not a measurement.

And the type_concept_id for this record would be the provenance of the record (claim record, EHR billing record, etc.). A person isn't going to self-report HCPCS code K1034. Patients self-report drugs they are currently taking, past medical history, etc. If they self-report a positive Covid-19 test result, it will not be coded as HCPCS code K1034. It'll be a note, survey, flowsheet, or other check box like data type. Not a billing code.