Incongruence with "patient transfer" procedures

[funaesthesia (Clemens Rieder, M.D.)]

Hi everyone !

I want to map transfer procedures where patients are transferred between hospitals or dedicated wards within the same hospital.

A quick glance at “patient transfer” concepts of SNOMED showed that all concepts belong to the concept_class of Procedure and the majority belongs to the domain of Observation whereas only 5 are mapped to the Procedure domain.

I think mapping an active process (i.e. “Taking a patient, putting them on the stretcher, transferring them to another physical location”) should belong to the Procedure domain and not the Observation domain.

Checking the hierarchy shows that the domain switches from Procedure to Observation within the first step down the ladder.

What are your thoughts on that?

[MPhilofsky (Melanie Philofsky)]

Check out the Visit Detail table, “The VISIT_DETAIL table is an optional table used to represents details of each record in the parent VISIT_OCCURRENCE table. A good example of this would be the movement between units in a hospital during an inpatient stay or claim lines associated with a one insurance claim.”

Also, the Procedure table should contain “This table contains records of activities or processes ordered by, or carried out by, a healthcare provider on the patient with a diagnostic or therapeutic purpose.”

[funaesthesia (Clemens Rieder, M.D.)]

Thank you for your reply!

I use the VISIT_DETAIL for in-house transfers already.
Regarding your second paragraph one could argue that a transfer to - let’s say - a cardiac ICU in a different hospital could be a therapeutic purpose. Or do you think differently?

[MPhilofsky (Melanie Philofsky)]

@funaesthesia

one could argue that a transfer to - let’s say - a cardiac ICU in a different hospital could be a therapeutic purpose. Or do you think differently?

I’m going to disagree.

However, using your logic, going on vacation is a therapeutic purpose

[funaesthesia (Clemens Rieder, M.D.)]

@MPhilofsky

However, using your logic, going on vacation is a therapeutic purpose

It sure is!

So how would you treat any inter-hospital transfers with the OMOP CDM? Should I ignore the fact that they are members of the Observation domain and continue mapping to the concepts mentioned above?

[MPhilofsky (Melanie Philofsky)]

You can continue to use the Visit Detail table. The FK to the Care Site table will distinguish the two.

[rosie (Rosie S)]

Hi @MPhilofsky,

I’m also keen to use visit_detail for movements within the same hospital/facility. When looking in Athena, standard concepts in the Visit Domain don’t appear to be at the “ward” level. They look more like different hospitals/facilities. E.g. What concepts could I use for general/medical, surgical, or rehabilitation wards to populate the visit_detail_concept_id, admitted_from_concept_id, discharged_to_concept_id fields? Thanks

[MPhilofsky (Melanie Philofsky)]

@rosie

Standard concept_ids with domain_id = ‘Visit’ are the only allowed concepts to be used in the visit_detail_concept_id, admitted_from_concept_id, and discharged_to_concept_id fields. And these data are not at the ward level. The OHDSI community has yet to find a vocabulary with this level of granularity. Do you know of one?

At the University of Colorado we only populate the visit_detail_concept_id for Intensive Care Unit visits since we have a use case to identify these visits. The rest of our data is populated with concept_id = 0. And that’s ok because we still have the Visit_Detail.care_site_id field which is a FK to the Care Site table which has the care_site_name which is what we use to identify other care sites of interests. It’s not an ideal solution, Care Site is not utilized in Atlas, but it works for our analysts who have direct access to our OMOP CDM.