OHDSI Home | Forums | Wiki | Github

Observation room - CMS place of service - is it inpatient, outpatient or ER

OMOP Vocabulary in the visit domain have three high level concepts

The standard vocabulary - CMS place of service - rolls up to these top level concepts.

I observed today that the conceptId ‘Observation Room’ 581385 maps up to ‘Outpatient Hospital’ and then to ‘Outpatient Visit’.

This seems counter intuitive to me - because when I used to work in the ER the patients Observation Room was usually a physical place in the Emergency room, where a person would be observed for several hours. Patients commonly stay overnight in observation room (i.e. they would sleep in the hospital/ER). Common scenarios of - patients with chest pain and waiting for three troponin blood draws to be below threshold, or patients with MI but the hospital does not have a bed yet to take them. etc.

In addition, guidance for coding for Observation room here Article - Billing and Coding: Acute Care: Inpatient, Observation and Treatment Room Services (A52985) seems to also suggest that it should not be used for elective visits eg…

Per CR 6492 (and the manual): “General standing orders for observation services following all outpatient surgery are not recognized. Hospitals should not report as observation care, services that are part of another Part B service, such as postoperative monitoring during a standard recovery period (e.g., 4-6 hours), which should be billed as recovery room services.” Observation services should not be ordered by the physician for future, elective outpatient surgeries.

So i wanted to hear others opinion on this topic:

  • can patients be in observation room without being in ER?
  • should observation room be considered outpatient?

@Christian_Reich you and I worked on this about 5 years ago correct?

it turns out that this conceptId is not CMS place of service vocabulary, even though the OMOP vocabulary has labeled in CMS place of service - as seen here

i.e. its vocabularyId should be different from CMS place of service (maybe its an OMOP vocabulary)

In Colorado, our data does not contain CMS Place of Service codes. However, we do have patient classification coding which is related to billing for a particular visit. One of the patient classification codes is “observation”. Our data contain observation encounters in the ER and other care sites. However, most observation visits are in the ER.

Per Medicare.gov, “Observation services are hospital outpatient services you get while your doctor decides whether to admit you as an inpatient or discharge you. You can get observation services in the emergency department or another area of the hospital”. I think observation visits should be the child of an outpatient visit as a default since the person is not admitted as an inpatient to the hospital. However, if the person is in observation for more than 24 hours, our CDM v5.4 conventions for visits clearly define an inpatient visit as “Person visiting hospital, at a Care Site, in bed, for duration of more than one day, with physicians and other Providers permanently available to deliver service around the clock”. Once a person stays more than one day, it’s obvious they need the care an inpatient visit provides.

Then - It sounds like based on the pattern of use of the code for Observation room -it represents a high acuity state with intense services (blood draw, frequent monitoring) that may or may not lead to inpatient hospital. It’s definitely not a Doctor’s office/clinic visit. It is most likely part of emergency room visit. It may lead to an inpatient visit.

In phenotyping/cohort definitions we tend to use visit concepts as a surrogate for high acuity care. E.g. we use a definition like 1) 'person’s with myocardial infraction during inpatient or emergency room visit ’ as a cohort definition with better performance characteristics compared to the more simpler 2) ‘person’s with myocardial infarction during any visit’.

It sounds to me, in EHR data definition 1) would have sensitivity errors by missing persons with ‘myocardial infarction evaluated in an observation room visit’.

Well, you might want to rethink using the Emergency Room visit_concept_id for a surrogate to high acuity care. How many persons who go to the emergency room (ER) for care really need to be seen in the ER? Most broken bones, upper respiratory and intestinal symptoms, skin issues, and aches and pains can easily be treated in the outpatient setting. However, people still visit the ER for these complaints. For true ER or high acuity care, you should further define your idea of interest and not solely rely on concept_id = 9203.

I disagree. The standard of care for an acute myocardial infarction (AMI) , thrombolysis and reperfusion, prevents the condition from being treated in an “observation room visit”. The drugs and procedures used to treat a myocardial infarction require the person to have ER and inpatient care. I will argue you have improved your sensitivity by excluding persons who didn’t have a true AMI because it’s not something that is treated in the outpatient/observation room setting.

The observation room visit is a little different than all the other categories. The person needs a higher level of care than the traditional doctor’s office or clinic provides. Yet, they don’t need ER level of care. However, they aren’t quite ill enough to justify an inpatient visit quite yet. The providers need some more data points (will this medicine provide relief, do the lab tests suggest worsening disease, etc.) to make a decision. We should leave the inpatient visit definition as it is. If a person needs care for more than one day, then the person is an inpatient regardless of what the hospital labels it.

I think we are all agreeing - ‘observation room’ has higher care intensity compared to a physician office visit, and we will see care in this setting that are not done in physician office.

This is my argument

  1. the vocabulary id is shown to CMS Place of service, while there is no such concept in CMS place of service. So that’s wrong - this is an OMOP (extension) vocabulary that we created i.e. needs to be fixed.
  2. Observation room concept is not in US claims data, I have seen it in EMR data sources. In US hospitals (i.e. EMR data), the term ‘Observation room’ - based on my understanding - refers to a place in between Emergency room and Inpatient stay. If this is true, then Observation room should be treated as a special case similar to ER-IP concept. In vocabulary hierarchy it maps to ‘outpatient visit’ via ‘outpatient hospital’ and belonging to the same hierarchy as ‘Office visits’. I am arguing that this hierarchy is wrong and Observation room - should be in the same hierarchy as ‘Emergency room Visit’.

Correct. Somebody made a mistake. Will be fixed.

I think we agree it is not an outpatient visit. It could be its own thing, or it could be a child of an ER. I’d go for the latter. Should we?

1 Like