OHDSI Home | Forums | Wiki | Github

Mapping "Bronchoalveolar Lavage" specimen

Hey all,

We are ETL-ing our study on Ventilator-associated Pneumonia to OMOP-CDM.

Under ‘sample type’ in our microbiological identification and antimicrobial susceptibility testing forms, we collect - “Bronchoalveolar Lavage (BAL)” and “Mini-BAL”, as separate options.

Technically, the distinction between “BAL” and “Mini-BAL” is that they are two separate procedures -

  • BAL = Bronchoalveolar lavage [37204340] / Bronchoscopic bronchoalveolar lavage [37203792]
  • Mini-BAL = Blind bronchoalveolar lavage [37203791]

And, specimen collected through either of these procedures can be categorized as “Bronchoalveolar Lavage Fluid Sample”, which exists in Athena as [4121345] and maps to the SPECIMEN domain in OMOP-CDM.

The challenge we are facing is to distinguish “Bronchoalveolar Lavage Fluid Sample” specimen collected using BAL v/s Mini-BAL.

As we understand, SPECIMEN domain in OMOP-CDM does not allow the use of [value_as_concept_id] or [value_as_string]. Otherwise, we would have tried to map these two data items by using the concept ids for the two procedures in the [value_as_concept_id] field. As, also shown in the screenshot. But, this obviously is not permitted as per the OMOP-CDM SPECIMEN domains rules.

So, I was wondering if any one else has faced this challenge? And, if you could please advise us on how best to tackle this? Perhaps, it makes sense to use the [anatomic_site_concept_id]?

Many thanks in advance for your help.


1 Like


It shouldn’t matter for the specimen. You want to make the distinction at the procedure level and have it recorded at the same day as the specimen.

However, this is not as straightforward either. I am not 100% sure how SNOMED considers these procedures:

Somebody who knows this should look at these procedures.

1 Like