Biobank data into OMOP CDM

[CarlosLopezERPSS]

Hello, our hospital already has the EHR data transformed into OMOP CDM, and now we are in the process of enriching our OMOP CDM with new source data. The hospital has an associated biobank with information on available specimens (from patients that are already included in our OMOP), but we haven’t found any official guidelines on how to map this information into OMOP.

We figured we could use the concept_id 46235133 (Biobank specimens are stored and available for research) in the Observation table to point to the specimen sample. Then, we thought of using concept_id 3018323 (Volume of Specimen) in the Measurement table to capture the quantity of specimen left.

Is there an official way to represent this information? Has anyone already included biobank data in OMOP so we could follow the same approach?

Thanks in advance for any insights or guidance!

Carlos

[MelaniePhilofsky]

Hello @CarlosLopezERPSS,

We have this information in our OMOP CDM. How you store your data depends on your use case. For us, the biobank data is for internal use only. And the logic for who is included in the different biobank groups of persons is quite complex. So, we have created cohorts in the OMOP CDM Cohort table and utilized the Cohort Definition table to identify the different groups (aka cohorts) of persons for each.

Since the records in the Cohort table were created outside of Atlas, they are not viewable within Atlas. But that's ok for us because we don't use Atlas. Also, since this isn't an OHDSI rule, requirement or convention, we won't be able to collaborate with other OMOP sites. But that's also ok because we aren't sharing this information with other institutions.

What's your use case? Why do you need these data in OMOP? How will they be used?

[CarlosLopezERPSS]

Hi @MelaniePhilofsky , thanks a lot for your reply and for sharing your approach.

Our goal is to include biobank information in our OMOP instance so we can easily identify which patients have specimens stored and available for future studies. This would allow us, for example, to:

• Identify a cohort of patients who already underwent a certain test, and if some patients in the cohort did not have the test performed but do have a specimen in the biobank, we could perform the test on the specimen.
• Incorporate the results of these new tests back into our OMOP CDM, enriching the dataset for subsequent research.
• Enable projects where biobank specimens are directly used as part of the study design.

In short, the idea is that by having the linkage between patients and available specimens represented in OMOP, we can both facilitate study feasibility assessments and seamlessly integrate future biobank-derived results into our OMOP CDM.

Thanks again for your insights!

Carlos

[MelaniePhilofsky]

Hello @CarlosLopezERPSS ,

Sounds like this will be for your institution only, correct? You don't mention an OHDSI network use case.

There isn't a convention on how to represent these data. OHDSI conventions are use case driven. So, when a researcher/study needs certain variables to enable network research, the Themis WG will help guide them. Since this is for internal research, the approach you suggests sounds reasonable to me. I will also direct you to use the *_event_id and *_event_field_concept_id fields to connect the Measurement and Observation records.