ENVO and OMRSE with respect to health care facilities

[ddooley]

Is OMRSE attracted to the idea of having built environment references in ENVO that OMRSE organization and role related axioms could work off of? I see opportune modelling in OMRSE regarding architectural structure and underlying facilities with respect to organizations that administer them and use them for service provision. But would OMRSE prefer to use ENVO terms for the facilities themselves? I ask because I was poised to add about 60 hospital / clinic / healthcare facility and component terms to ENVO, and then realized OMRSE had a handful of them.

If there are reservations to migrating the terms to ENVO, what would they be? The motivation to have these terms in ENVO would be that such facilities have many aspects to them - building maintenance, costing, engineering, environmental impact, waste disposal, etc. that go beyond a strictly medical purview. Or a compelling argument to shift into OMRSE?

If it seems like a good harmonization effort, we can dive into what you would need out of term definitions sitting in ENVO.

healthcare facility
  acute care facility
  clinical patient assessment facility
     hospital unit
        anaesthetics unit
        breast screening unit
        chaplaincy unit
        coronary care unit
        diagnostic imaging unit
        discharge unit
        emergency unit
           emergency unit room
        gastroenterology unit
        general surgery unit
        geriatric care unit
           geriatric intensive care unit
        gynaecology unit
        haematology unit
        intensive care unit
           surgical intensive care unit
           traumatic intensive care unit
           intensive care unit room
        maternity unit
        mobile intensive care unit
        micobiology unit
        neonatal unit
           neonatal intensive care unit
        nephrology unit
        neurology unit
        nutrition and dietetics unit
        occupational therapy unit
        oncology unit
        ophthalmology unit
        orthopaedics unit
        pain management unit
        pediatric unit
           pediatric intensive care unit
        pharmacy unit
        physiotherapy unit
        post-anesthesia care unit
        high dependency unit
        psychiatric unit
           psychiatric intensive care unit
        radiotherapy unit
        rheumatology unit
        urology unit
     patient room
     tertiary referral hospital
  clinic
     medical clinic
        chiropractic clinic
        dental clinic
        family practice clinic
        massage therapy clinic
        ophthalmology clinic
        optometry clinic
        pain clinic
        physiotherapy clinic
        walk-in clinic
  pharmacy
  long-term care facility
     assisted living facility
     group home facility

The issue was triggered when I saw #153 re. OBI "hospital" and OMRSE "hospital facility". The list is motivated by sampling location description or transmission context with respect to covid19 and other infectious diseases.

Thanks for input,

Damion

[kaiiam]

@ddooley regardless of where the terms go it's definitely better not to have a term like hospital be in all of OBI, OMRSE and ENVO. I think it would be good to clarify this so were all together upholding OBO principal 5 (not duplicating across ontologies).

At the moment my understanding is that ENVO’s scope is all environments natural and antropogenic, as well as many materials natural or human produced. Perhaps this is fine but it does seem like quite a lot of ground to cover, especially with @ddooley's health-care related environments. Are these better suited elsewhere? Perhaps some ENVO, OBI and OMRSE folks have thoughts on this? @pbuttigieg, @cmungall, @jamesaoverton, @zhengj2007, @matentzn, @hoganwr @dillerm?

@ddooley has a large COVID-related term set he's trying to get into various OBO ontologies so I think clarifying this will help with his efforts.

[hoganwr]

I think it would be worth discussing, probably on a call with relevant stakeholders. Some principles that underlie this work: 1. The term 'hospital' is used ambiguously in natural language to refer to a building, an organization, and perhaps more (a licensure from the state to operate a particular physical facility as a hospital, the function of being a hospital, etc.) 2. Physical subunits of buildings change function all the time. For example, when I was a medical resident, the medical ICU moved from Presbyterian Hospital to Montefiore Hospital (after an organizational merger of Montefiore into UPMC, so now both facilities were owned and managed by one organization). Patients were transported from the Presby ED across two "bridges", one from Presby to Eye and Ear Hospital, and then across another bridge that was L shaped, from E&EH to Montefiore. 3. Thus the pain clinic, the pharmacy, the dental clinic all might stay organizationally identical while moving from one physical place to another, and that move might even cross entire buildings blocks apart. 4. Buildings that serve healthcare functions sometimes even end up ceasing to serve any healthcare functions at all (e.g., apartment buildings: https://www.physicianleaders.org/news/so-what-are-your-plans-for-that-old-hospital-building) , and vice versa (e.g., a physician office acquiring a space in a strip mall and converting it to healthcare use). My experience is that although healthcare organizations do not carefully distinguish these things in their data, their infection control personnel frequently lament said lack. Bill

…

On Mon, Mar 15, 2021 at 6:17 AM Kai Blumberg ***@***.***> wrote: @ddooley <https://github.com/ddooley> regardless of where the terms go it's definitely better not to have a term like hospital be in all of OBI, OMRSE and ENVO. I think it would be good to clarify this so were all together upholding OBO principal 5 <http://www.obofoundry.org/principles/fp-005-delineated-content.html> (not duplicating across ontologies). At the moment my understanding is that ENVO’s scope is all environments natural and antropogenic, as well as many materials natural or human produced. Perhaps this is fine but it does seem like quite a lot of ground to cover, especially with @ddooley <https://github.com/ddooley>'s health-care related environments. Are these better suited elsewhere? Perhaps some ENVO, OBI and OMRSE folks have thoughts on this? @pbuttigieg <https://github.com/pbuttigieg>, @cmungall <https://github.com/cmungall>, @jamesaoverton <https://github.com/jamesaoverton>, @zhengj2007 <https://github.com/zhengj2007>, @matentzn <https://github.com/matentzn>, @hoganwr <https://github.com/hoganwr> @dillerm <https://github.com/dillerm>? @ddooley <https://github.com/ddooley> has a large COVID-related term set he's trying to get into various OBO ontologies so I think clarifying this will help with his efforts. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#157 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAJR55WRAXZLWAUUXR7AH23TDXNCBANCNFSM4YP4SP4A> .

[ddooley]

@hoganwr indeed I noticed OMRSE's use of "... facility" - which I like perhaps more than just "unit" as it implies physical built environment rather than organizational/operational /management semantics. I am only after naming the physical built stuff assigned to a functional role / capability - to provide physical context to infectious disease biosamples or processes at some time t. I leave the organizational /management/ semantics in the good hands of OMRSE!

[wdduncan]

Given the important distinctions brought out by @hoganwr regarding the functions/roles that the healthcare facilities have, I do not think it would be good for EnvO to take ownership of things like hospitals and ICU units. However, EnvO does have class named hospital [ENVO_00002173], but envo:hospital does not have not have any health related roles/functions associated with it, and it shouldn't (IMHO) as that would be too far out of scope for EnvO. It might even be best for EnvO to cede envo:hospital to OMRSE.

That being said, it might be best for OMRSE to import EnvO's higher level classes, such as building and building part. This may allow OMRSE to logically define as envo:building and (.....).

There also appears to be some duplication between envo:human construction and omrse:architectural structure that we should straighten out.

[pbuttigieg]

Hi all

@wdduncan I quite like your suggestion, but it seems more natural to me that the (physical) building classes are stored in ENVO, and then imported to OMRSE where they can be embellished with roles that work for OMRSE's scope.

The "units" however, feel like they fit in OMRSE well, as they are not (necessarily) bound to rooms or floors (as @hoganwr notes in his points 2 and 3 here)

Ultimately, OBO really needs a dedicated ontology for building types and/or architectural elements that can be imported as a module. Similar to how FOODON branched off of ENVO.

@hoganwr

1. The term 'hospital' is used ambiguously in natural language to refer to
   a building, an organization, and perhaps more (a licensure from the state
   to operate a particular physical facility as a hospital, the function of
   being a hospital, etc.)

This is an excellent point, and I think it offers us the way forward: ENVO can limit itself to the physical buildings and building parts, and the subtleties of role-defined sub-(fiat)-parts of thes be covered in OMRSE

Happy to have a call about this too

[ddooley]

For my needs I would like to rename the above "X unit" to "X unit facility" to make clear the physical environment of the unit is being referenced.
I can send out a doodle poll to the people involved about a good telecon time in next few weeks (one in which invitees can get others who are interested to take the poll too).

[pbuttigieg]

Thanks @ddooley

I'm sensing something like:

• ENVO would contain terms for buildings that have been deliberately built as hospitals, and rooms therein built with specific purposes in mind
• OMRSE would have a broader notion of "hospital" which can be role based and realisable in any process where any building is used for hospital-like activities. The same thinking applies to parts of those buildings.
• Combining the above, this is then a way for applications to import any building type from ENVO (barns, schools, houses) and assign them the OMRSE hospital roles/functions as needed, to capture a wide variety of situations.

Units and facilities are linguistically fuzzy, so we'll need to sort out our approach consistently

[mbrochhausen]

Hi, just adding from the perspective of the CAFE project, we will also need hospital organizations (and their parts). I suggest those are in OMRSE, too. Best, Mathias

…

On Wed, Mar 17, 2021 at 1:33 PM Pier Luigi Buttigieg < ***@***.***> wrote: Thanks @ddooley <https://github.com/ddooley> I'm sensing something like: - ENVO would contain terms for buildings that have been deliberately built as hospitals, and rooms therein *built* with specific purposes in mind - OMRSE would have a broader notion of "hospital" which can be role based and realisable in any process where any building is used for hospital-like activities. The same thinking applies to parts of those buildings. - Combining the above, this is then a way for applications to import any building type from ENVO (barns, schools, houses) and assign them the OMRSE hospital roles/functions as needed, to capture a wide variety of situations. Units and facilities are linguistically fuzzy, so we'll need to sort out our approach consistently — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#157 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACF6DLWMHU3E5KCL5BN2D4LTEDYZFANCNFSM4YP4SP4A> .

[pbuttigieg]

just adding from the perspective of the CAFE project, we will also need
hospital organizations (and their parts). I suggest those are in OMRSE, too.

Yes, the organisations are socially grounded. OBI's organization class would be the root, I suppose, but this should be abstracted out of OBI (perhaps moved to COB, if not already planned)

[mbrochhausen]

I totally agree that the OBI class is the root, but if I am not mistaken that is imported by OMRSE. Best, Mathias

…

On Wed, Mar 17, 2021 at 1:54 PM Pier Luigi Buttigieg < ***@***.***> wrote: just adding from the perspective of the CAFE project, we will also need hospital organizations (and their parts). I suggest those are in OMRSE, too. Yes, the organisations are socially grounded. OBI's organization <http://purl.obolibrary.org/obo/OBI_0000245> class would be the root, I suppose, but this should be abstracted out of OBI (perhaps moved to COB, if not already planned) — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#157 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACF6DLRMS3XWINNL3KSFDVDTED3FTANCNFSM4YP4SP4A> .

[cmungall]

This makes sense to me

…

On Wed, Mar 17, 2021, 11:33 Pier Luigi Buttigieg ***@***.***> wrote: Thanks @ddooley <https://github.com/ddooley> I'm sensing something like: - ENVO would contain terms for buildings that have been deliberately built as hospitals, and rooms therein *built* with specific purposes in mind - OMRSE would have a broader notion of "hospital" which can be role based and realisable in any process where any building is used for hospital-like activities. The same thinking applies to parts of those buildings. - Combining the above, this is then a way for applications to import any building type from ENVO (barns, schools, houses) and assign them the OMRSE hospital roles/functions as needed, to capture a wide variety of situations. Units and facilities are linguistically fuzzy, so we'll need to sort out our approach consistently — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#157 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAAMMOODE54IG7OJ6YUGEETTEDYZFANCNFSM4YP4SP4A> .

[wdduncan]

If ENVO is going to host the classes, I think we need to be careful about how to handle poly-hierarchies. For example, a building on a medical campus may be both a healthcare facility and educational facility. Ideally, these poly-hierarchies should handled via reasoning. This suggests to me that the roles/functions would also need to be in ENVO. Would the medically relevant roles/functions keep an OMRSE identifier and then have ENVO import them for purposes of reasoning?

[pbuttigieg]

Would the medically relevant roles/functions keep an OMRSE identifier and then have ENVO import them for purposes of reasoning?

yes, that makes sense