OHDSI Home | Forums | Wiki | Github

Do we have concepts for surveillance and/or vaccine effectiveness variables (Covid-19, influenza) in OMOP?

General
#1

Hi there,

Could someone help with this question - whether it would be possible to curate surveillance /VE variables for infectious diseases (e.g. influenza, Covid-19) in SNOMED CT, and then use them in OMOP CDM?

I guess there’s a few things happening here:

Does SNOMED CT already have these variables or do they need curating? If not, why not? Can this be added?
Would it be possible to map them to OMOP CDM? If not currently possible, how would one go about this process?

Apologies for being slightly vague - I’m looking for any info - literature, links etc. We are mapping to OMOP CDM, but would love to run a study related to these variables, so was wondering what is possible and what are the processes. Apologies also for not being a clinician familiar with SNOMED CT, but a data analyst.

Thank you.

#2

@AnnaLuiza:

There is no such a thing as a “variable” in the OMOP CDM. You have concepts, and they live in tables according to their domains. So, your typical infectious diseases and their outcomes are Conditions recorded in the table CONDITION_OCCURRENCE. So, you don’t have to do anything special for Covid research, just record the facts as concepts.

You may want to read up in the Book of OHDSI.

#3

Hello @AnnaLuiza,

For intents and purposes OMOP CDM is created for, it is not adapted to store inferred data about applied properties of concepts. Standardized vocabularies do not store medical knowledge, like drugs indications, contraindications, effectiveness or administration instructions. It only contains necessary properties to support possible entries from patient data, like vaccine name, ingredient or dose volume.

If you need to store additional properties of vaccine concepts, you could probably link Qualifier Value concepts to vaccine concepts using FACT_RELATIONSHIP table.

#4

Hi Christian,

Thank you for replying, I’ve been reading the book of OHDSI, but needed more context.

#5

Hi Eduard,

Thank you so much. As someone who will be mapping without any medical experience I did get mixed up in CDM concepts vs medical information, and this has really helped me to separate what would be what you called medical knowledge from potential entries in different patient data tables.

Thank you for alerting me to the way I may add additional properties to concepts, I admit I did not fully understand the FACT_RELATIONSHIP table.

I will keep re-reading the Book of OHDSI, it helps.

Thank you,

Ana Lena

t