Ingest the Nebraska lexicon into the vocabulary tables.

[mgurley]

• Parse input. Create attribute-value relationship.
• Link to sites: breast, colon
• Replace LOINC attribute codes with SNOMED when available

[dimshitc]

here's the thing:

not all the entries from pdf-protocol exist in release SNOMED-like Nebraska files.
And the first thing is to ingest what's in pdf-protocol rather then what in distribution files, right?
So there will be regular LOINC codes from the protocol with relationship to Site and values from the protocol.
for example:

Note, this concept doesnt' have SNOMED equivalent
so we'll have
85331-7 | 'Has site (Nebraska)' | 76752008 | Breast structure
85331-7 | 'Has value (Nebraska)' | 260407003 | Weak
85331-7 | 'Has value (Nebraska)' | 4470001000004101|Moderate
85331-7 | 'Has value (Nebraska)' | 260404005 | Strong

Is this the way you see it?

[dimshitc]

attaching the protocols
CAP_colon_biomarker_encoded_20170331.pdf
CAP_colorectal_encoded_20170331.pdf
CAP_Invasive_breast_cancer_encoded_20170331.pdf
CAP_breast_biomarker_encoded_20170331.pdf

[mgurley]

@dimshitc Yes, I am a little confused about how the distribution files relate to the protocol PDF. If we can't use the distribution files as the basis for the ingestion and need to scrape the PDF, then we have a problem. This was why I sent the email to Scott Campbell yesterday asking about the latest version. I thought he mentioned at AMIA that LOINC was being removed from the Nebraska Lexicon because of licensing issues. Your layout of the concept relationships look good but we need to talk to Scott about the distribution files.

[dimshitc]

Scrape the PDF is not a problem actually for now, we already copy-pasted this information manually and it can be used in a first version of Nebraska OMOP vocabulary
but later we need to get some automated way of course

[dimshitc]

So, is this still actual?
Is it related somehow to CAP eCC?
Can we talk about this on the one of the vocabulary meetings?

[meerapatelmd]

@mgurley @dimshitc I think Nebraska is key from some of our pathology mappings. What is the status of this effort? I downloaded the Nebraska files, but they weren't as intuitive as I was expecting.

[ekorchmar]

Currently the SNOMED version of Nebraska Lexicon is being under assessment for inclusion in OMOP CDM Vocabularies. It does not seem to include any crosswalks to PDF protocols.

[sratwani]

Completed as a part of the 2020 Symposium effort.