Retrieve and Parse the Encounter Code ValueSet from VSAC

[chrisroederucdenver]

To help with mapping for visit_concept_id #72 getting values for visit_concept_id, it would be helpful to get the valueset specified in CCDA.

https://vsac.nlm.nih.gov/valueset/2.16.840.1.113883.3.88.12.80.32/expansion/Latest

This involves not just login and permissions from VSAC, but also uploading to the Foundry.

[richard1933]

vsac_EncounterTypeCode.json

[chrisroederucdenver]

Thanks @richard1933 ! I just sent it as an attachment to Matt Owens who will upload it. That turns out to be the easy way. You were right.

I asked him to put it into CCDA/transforms/mapping-reference-files. FYI @stephanieshong @tannerzhang

[chrisroederucdenver]

OK, uploaded here: https://foundry.cladplatform.org/workspace/data-integration/dataset/preview/ri.foundry.main.dataset.6a83f3d6-4cd6-413b-9db3-9a792118e69f/master

It will need extracting from the dataset and more significantly parsing into a table.

[chrisroederucdenver]

@AdamLeeIT The way we did it with the other file was to download it and e-mail the file to him (at his request). Also, could you CC me on Slacks or e-mails to Matt?

[AdamLeeIT]

@tannerzhang @chrisroederucdenver This ticket is now complete. I converted the VASC ValueSet JSON file uploaded by Palantir into a tabular format dataset.

The dataset is named: vsac_EncounterTypeCode It is located here: /All of Us-cdb223/Identified: HIN - HIE/CCDA/transform/vsac_EncounterTypeCode

The dataset has the following columns: [id, system, version, code, display]

oid: This column contains the OID of the value set, extracted from the root id element of the JSON. It is the same value for every row and provides a consistent reference to the source of the data. Its value is always 2.16.840.1.113883.3.88.12.80.32 for this dataset. system: Specifies the system or namespace for the coding scheme (e.g., http://www.ama-assn.org/go/cpt). This was as sourced from the JSON file and is always http://www.ama-assn.org/go/cpt for this dataset. version: Indicates the version as sources from the JSON. It is always 2025 for this dataset code: The unique CPT code within the system that identifies a specific encounter type or procedure. These codes are often numeric or alphanumeric. display: A human-readable description of the code, explaining the type of encounter or service it represents (e.g., "Observation Care Discharge Services").

The Python for generating this dataset is located In CCDA-Tools WorkSpace. CCDA-Tools WorkSpace vsac_encountertypecode.ipynb

[richard1933]

mapped 186 out of 203 CPT4 code to OMOP concepts, dataset is located at https://foundry.cladplatform.org/workspace/data-integration/dataset/preview/ri.foundry.main.dataset.7c182b87-2c2b-4e41-8c9c-1f93e7b463d8/master

[chrisroederucdenver]

@AdamLeeIT The link you provided above is to an instance of a VM that isn't running. I get an error when using that link.

Instead send links to code in code repositories instead like this: https://foundry.cladplatform.org/workspace/data-integration/code/repos/ri.stemma.main.repository.42f809ff-135a-45ce-bc46-683c2ca00426/contents/refs%2Fheads%2Fmaster/vsac_encountertypecode.ipynb

[chrisroederucdenver]

@AdamLeeIT in the vsac_encountertypecode.ipynb, when you write Dataset.get("ccda_encounter_code_value_set") Does that reference a dataset transformation (wormhole) or use some other mechanism. I'm trying to locate, from the code, where the json file is? I know, because I worked with Matt to upload it that it is in /All of Us-cdb223/Identified: HIN - HIE/sharedResources/CCDA Encounter Code value set, but how would anyone reading your code know that?

[chrisroederucdenver]

@richard1933 this looks like a start at mapping the vsac set above to standard concepts? Am I right??

What we ultimately need is a mapping to concepts for use in the visit_occurrence.visit_concept_id, which I believe is the Visit domain. @mbangudi had expressed interest in that mapping. Please include her. There's a ticket for it #72

Monique, I'm talking about a query in the workbook Richard linked above: https://foundry.cladplatform.org/workspace/vector/view/ri.vector.main.workbook.03b65dcf-84b4-4d25-ac7e-579b548667f3?table=728a58fb-7751-4cc3-b106-f3c6688760d5&branch=master

[chrisroederucdenver]

@AdamLeeIT also for vocabulary things like this I think putting the output in the mapping-reference-files under transforms.