Stream: Covid-19 Response
Topic: Metrics for Hospital Capacity
Michael Donnelly (Mar 26 2020 at 16:31):
COVID-19-Metric-Specifications-2020-03-26-c.pdf
- Open ICU Beds - the total number of open ICU beds at the point of reporting.
- Open Non-ICU Acute Beds - the total number of open non-intensive, acute care beds at the point of reporting.
- Individual healthcare organizations define which departments are for intensive and non-intensive acute care
- ICU Active Ventilator Episodes - the number of patients currently on a ventilator in intensive care units during the reporting period. Nurses document whether a patient is on a ventilator and when it is turned on and turned off
- ICU Total Ventilators - the number of ventilators available for use in ICUs at the time of reporting. Individual healthcare organizations manually specify how many ventilators they have.
Michael Donnelly (Mar 26 2020 at 16:38):
Open bed
An open bed is any unoccupied, unblocked, and available bed. Beds can be blocked or unavailable for many reasons, such as if a room has multiple beds and one of the beds is occupied by a patient who needs to be isolated or has another need for privacy. A bed might also be unavailable after a patient is discharged but the bed still needs to be cleaned.
Michael Donnelly (Mar 26 2020 at 16:41):
We also have computed metrics. ICU Ventilators Available is ICU Total Ventilators minus ICU Active Ventilator Episodes.
Michael Donnelly (Mar 26 2020 at 16:52):
As @Lloyd McKenzie recommended, "ventilators" actually means "ventilator slots." If one ventilator machine can be used for four patients at a time, it counts as four ventilators.
Michael Donnelly (Mar 26 2020 at 17:00):
@Keith Boone what additional bed management context are you looking for?
Abbie Watson (Mar 26 2020 at 18:30):
Can we get the Cerner folks to weigh in on these definitions? And can we begin thinking about FHIR queries for them? This seems like a lot of Device queries. Maybe Location queries.
Josh Mandel (Mar 26 2020 at 18:36):
This comes back to the question of whether systems are likely to expose the data about devices and locations, or simply expose reports on metrics.
Abbie Watson (Mar 26 2020 at 18:43):
Fair point.
John Moehrke (Mar 26 2020 at 18:44):
two very different security exposures...
Abbie Watson (Mar 26 2020 at 18:57):
I'm just spitballing an idea here... but the Measure resource includes the group and supplementalData fields, which could/should feasibly include the Device and Location data. Neither Measure.group nor Measure.supplementalData are required fields. But they could be, within a profile. And we happen to have a US Core profile that exists at the appropriate level for coordinating all of these activities. Perhaps we could profile the Measure and MeasureReport resources to require inclusion of Device and Location data when reporting ICU bed data. That could harmonize the granular reporting and the metric reporting, and allow slicing/dicing of data after the fact. It could be summarized in a simple rule of "Don't send an ICU metric unless you include the basic underlying device and location counts with it.' Fairly easy to understand and implement.
Josh Mandel (Mar 26 2020 at 19:22):
I agree with all this up to "Don't send an ICU metric unless you include the basic underlying device and location counts with it." -- I'm assuming we'll want people to send summary metrics regardless of whether they're able to include the underlying granular devices/locations. I might just be misunderstanding.
Abbie Watson (Mar 26 2020 at 19:25):
If we profiled the resource to require group or supplementalData to be required, how else would you paraphrase it?
Michael Donnelly (Mar 26 2020 at 19:46):
Would we be running the risk of the perfect being the enemy of the good? If a site is unwilling or unable to expose the granular data but can give totals, do we want the totals?
Vassil Peytchev (Mar 26 2020 at 19:53):
I think this actually corresponds the original suggestion
"Don't send an ICU metric unless you include the basic underlying device and location counts with it."
Although the rest of the exchange is a bit confusing.
David Pyke (Mar 26 2020 at 19:56):
We're not in the position to reject what we can get at this point.
Michael Donnelly (Mar 26 2020 at 20:24):
We added some COVID-19 specific capacity metrics. These will be finalized on Monday, March 30.
- COVID-19 Admissions - total number of COVID-19-positive patients with admissions during the reporting period.
- COVID-19 ICU Occupied Beds - number of ICU beds occupied by COVID-19-positive patients at the point of reporting.
-
COVID-19 Average Length of Stay - average number of days COVID-19-positive patients spent in an inpatient
setting (including the ICU) before being discharged during the reporting period. -
COVID-19 Average ICU Length of Stay - as above but only for ICU
- COVID-19 Currently Admitted Patients - count of currently admitted COVID-19-positive patients at the point of reporting
- COVID-19 Currently Admitted Patients on Ventilators - count of currently admitted COVID-19-positive patients who are also on a ventilator at the point of reporting.
-
COVID-19 Ventilator Hours - average time, in hours, to two decimal points, that admitted COVID-19-positive patients were on ventilators during the reporting period.
-
numerator: total number of hours that patients in the denominator spent on a ventilator during the reporting
period -
denominator: number of COVID-19-positive patients admitted and on ventilators during the reporting period
-
-
COVID-19 ICU Admissions - number of ICU admissions that started during the reporting period for COVID-19-positive patients.
As above, Individual healthcare organizations define which beds are for intensive and non-intensive acute care.
Michael Donnelly (Mar 26 2020 at 20:26):
COVID-19-positive patient
A patient with either or both of the following:
-
A documented COVID-19 infection in the patient’s chart. The infection can have been documented any time
in the 14 days prior to the patient’s admission and does not need to be active. -
A presumptive positive COVID-19 test
Abbie Watson (Mar 28 2020 at 01:14):
@Hans Buitendijk from Cerner just pointed out that the CDC has published guidelines for reporting hospital capacity.
https://www.cdc.gov/nhsn/pdfs/covid19/import-covid19-data-508.pdf
Now to being harmonizing and aligning....
Hans Buitendijk (Mar 28 2020 at 01:27):
I already put links in gitHub, and unless somebody happens to include it in the list here (https://github.com/AudaciousInquiry/saner-ig/wiki/Bed-Classification), I'll include it tomorrow together with the Cerner column update.
Hans Buitendijk (Mar 28 2020 at 01:31):
And if you want to stay under this subject, the page that the CDC guidance is on (https://www.cdc.gov/nhsn/acute-care-hospital/covid19/index.html) has also a link to an upcoming webinar on March 31, 2-3pm ET.
Josh Mandel (Mar 28 2020 at 01:36):
(deleted)
Josh Mandel (Mar 28 2020 at 01:37):
For example,
https://covid19-under-fhir.smilecdr.com/baseR4/MeasureReport?period=2020-03-16
Will get all available MeasureReport-based summaries for for March 16. We don't have a similar way to fetch Groups for a specific reporting date, but we are populating Group.characteristic.period once for each Group.
Last updated: Apr 12 2022 at 19:14 UTC
