OHDSI Home | Forums | Wiki | Github

Vaccine Mapping - Current State and QA

Implementers
#1

Good evening -
I have been downloading the OMOP standardized vocabularies for the purpose of reloading our source_to_standard table, and it seems to work pretty well. The new vocabulary set had all the COVID-related concepts we were missing. Thank you all for the hard work!

However, I did notice one quirk, the new dataset is missing, for certain vaccines, some of the CPT4 > standard concept mapping that was included in my old dataset.

SELECT * FROM concept_relationship
WHERE (concept_id_1 IN (44816520, 42628616, 42628541, 2213492)) AND relationship_id = 'Maps to
old dataset: 4 rows
new dataset: 0 rows

I did run get a UMLS key and run the Java utility to map the CPT4 codes, which ran without incident.

The front end of Athena has the non-standard > standard mapping for the above concept IDs, which makes me nervous that the problem is my implementation, not a change in business rules > eg https://athena.ohdsi.org/search-terms/terms/2213492

I understand that there’s a lot of discussion around improving the mapping of exactly these codes (that’s not super penetrable for me as a data engineer, not a medical vocabulary expert) Vaccine concept mapping improvement

Really what I’m asking, is the change in the concept_relationship table expected behavior, or is it my implementation of the concept_relationship table?

Thank you for your attention and for everything that you do.

#2

If CPT4 codes are standard concepts (and have no maps)
https://www.ohdsi.org/web/wiki/doku.php?id=documentation:vocabulary:cpt4
why do these four codes in the CONCEPT table have standard_concept IS NULL?

SELECT * FROM concept
WHERE concept_id IN (44816520, 42628616, 42628541, 2213492)

Rob

#3

This may help to explain. Looks like vocab group wanted to standardize on CVX for vaccines.

SELECT c1.concept_code, c1.concept_name, c2.concept_name, c2.concept_id
FROM concept c1
JOIN concept_relationship ON concept_id_1 = c1.concept_id
JOIN concept c2 on c2.concept_id = concept_id_2
WHERE c1.concept_id IN (44816520, 42628616, 42628541, 2213492)
AND relationship_id = ‘Maps to’
AND c2.standard_concept=‘S’;

concept_code concept_name concept_name concept_id
90750 Zoster (shingles) vaccine (HZV), recombinant, subunit, adjuvanted, for intramuscular use zoster vaccine recombinant 706103
90686 Influenza virus vaccine, quadrivalent (IIV4), split virus, preservative free, 0.5 mL dosage, for intramuscular use Influenza, injectable, quadrivalent, preservative free 40213146
90682 Influenza virus vaccine, quadrivalent (RIV4), derived from recombinant DNA, hemagglutinin (HA) protein only, preservative and antibiotic free, for intramuscular use Seasonal, quadrivalent, recombinant, injectable influenza vaccine, preservative free 40213152
90746 Hepatitis B vaccine (HepB), adult dosage, 3 dose schedule, for intramuscular use hepatitis B vaccine, adult dosage 40213306
1 Like
#4

Thanks for your response @DTorok - I guess my question is, I don’t have Maps To CVX row in my concept_relationship table. You seem to have these rows. (Your query above returns 0 rows for me.) this would seem to indicate a faulty implementation of the standardized vocabularies, would you agree? Should I try the download from Athena / java utility / \COPY process again?

#5

Yes I would download the vocabulary again. Make sure to check the box for CVX since it is not included by default.

1 Like
#6

Ah there it is @DTorok thanks for your time - have a nice week. - Rob

#7

@DTorok worked like a charm

t