Stream: Covid-19 Response
Topic: CDC Page
Hans Buitendijk (Mar 28 2020 at 00:39):
CDC Page with info (https://www.cdc.gov/nhsn/acute-care-hospital/covid19/index.html) including .csv (https://www.cdc.gov/nhsn/pdfs/covid19/covid19-test-csv-import.csv) and additional guidance links around that.
Abbie Watson (Mar 28 2020 at 01:11):
Not exactly harmonized with FHIR or the Epic initiatives. Will need to create a CSV to FHIR mapping function.
Hans Buitendijk (Mar 28 2020 at 01:25):
Or others. But this helps focus on the scope of FHIR profiles version 1 to enable a FHIR formatted .xml, .json to have consistent mapping and next step to share data in FHIR format, progressing towards next steps.
Josh Mandel (Mar 28 2020 at 01:35):
I'm kinda dismayed that the documented process to onboard a facility (before reports can be sent) involves over 2 hours of active time filling out signup forms, and two 24h waiting periods.
Josh Mandel (Mar 28 2020 at 01:38):
https://www.cdc.gov/nhsn/ltach/enroll.html
Michael Donnelly (Mar 28 2020 at 11:41):
The description for the field "numVent" says "MECHANICAL VENTILATORS: Total number of ventilators available" - any idea whether this is actually intended to be the number of open ventilators that are available for a patient to start using or the number of operational ventilators including those in use right now?
Josh Mandel (Mar 28 2020 at 15:30):
There are two distinct fields. One for total ventilators and one for ventilators in use.
Josh Mandel (Mar 28 2020 at 15:31):
https://www.cdc.gov/nhsn/pdfs/covid19/57.130-toi-508.pdf documents these toward the bottom
Michael Donnelly (Mar 28 2020 at 15:34):
Ah, yep, that does it. I was looking at https://www.cdc.gov/nhsn/pdfs/covid19/import-covid19-data-508.pdf which wasn't as clear.
Keith Boone (Mar 28 2020 at 19:14):
Abigail Watson said:
Not exactly harmonized with FHIR or the Epic initiatives. Will need to create a CSV to FHIR mapping function.
And a FHIR to CSV mapping function. I think we can probably address some part of that via Accept header, but there's another part of that needing to address columns to use. BUT, whatever we do, we can using FHIR Mapping stuff to document it.
Michael Donnelly (Mar 28 2020 at 19:46):
I found the different things being measured confusing, so I tried to map it out.
| Population | Location | Criteria for Inclusion | CDC | Epic |
|---|---|---|---|---|
| Beds | Hospital | Total | numTotBeds | |
| Beds | ICU | Available | Open ICU Beds | |
| Beds | ICU | In use | numICUBedsOcc | |
| Beds | ICU | Total | numICUBeds | |
| Beds | Inpatient | Available | Open Non-ICU Acute Beds | |
| Beds | Inpatient | In use | numBedsOcc | |
| Beds | Inpatient | Total | numbeds | |
| Patients | Hospital | Known or suspected to have COVID-19 | numC19HospPats | |
| Patients | Hospital | Known to have COVID-19 | Currently Admitted COVID-19 Patients | |
| Patients | ICU | Known to have COVID-19 | COVID-19 Occupied ICU Beds | |
| Ventilated patients | Hospital | Known or suspected to have COVID-19 | numC19MechVentPats | |
| Ventilated patients | Hospital | Known to have COVID-19 | Currently Admitted COVID-19 Patients on Ventilators | |
| Ventilators | Hospital | Available | Ventilators Available | |
| Ventilators | Hospital | In use | numVentUse | |
| Ventilators | Hospital | Total | numVent | Total Ventilators |
Josh Mandel (Mar 28 2020 at 19:55):
it's kind of striking that there's almost no overlap between the measures defined by these two groups. Of course they're all very close :-)
Josh Mandel (Mar 28 2020 at 19:55):
The table is superbly helpful.
Hans Buitendijk (Mar 28 2020 at 23:53):
On https://github.com/AudaciousInquiry/saner-ig/wiki/Bed-Classification the comparison is even more stark. Will have similar updates for https://github.com/AudaciousInquiry/saner-ig/wiki/Patient-Summary shortly.
Keith Boone (Mar 29 2020 at 05:40):
Yeah, that's such a mish-mash that I want us to find a way to restructure it using the terminology hierarchy to make more sense.
Michael Donnelly (Mar 29 2020 at 15:22):
Which terminology do you mean, @Keith Boone ?
Keith Boone (Mar 29 2020 at 15:23):
Across the measures and strata it's hard to understand what is going on. There IS a hierarchy in all of this.
Michael Donnelly (Mar 29 2020 at 15:24):
Absolutely.
Michael Donnelly (Mar 29 2020 at 15:24):
Is there a specific set of terms you want to use?
Michael Donnelly (Mar 29 2020 at 15:24):
Something like the table above, or different from that?
Michael Donnelly (Mar 29 2020 at 15:25):
Whatever you're thinking, I'd like to start mapping Epic's metrics into it, so we can identify gaps.
Michael Donnelly (Mar 29 2020 at 15:25):
We have other metrics that aren't currently part of the COVID-19 reporting set that we could pull in.
Michael Donnelly (Mar 29 2020 at 15:37):
I took a stab at updating this with Cerner's metrics:
| Population | Location | Criteria for Inclusion | CDC | Epic | Cerner |
|---|---|---|---|---|---|
| Beds | Hospital | Total | numTotBeds | Hospital Beds | |
| Beds | ICU | Available | * | Open ICU Beds | Adult Critical Care Beds Available |
| Beds | ICU | In use | numICUBedsOcc | Adult Critical Care Beds Occupied | |
| Beds | ICU | Total | numICUBeds | * | |
| Beds | Inpatient | Available | * | Open Non-ICU Acute Beds | General Acute Beds Available |
| Beds | Inpatient | In use | numBedsOcc | General Acute Beds Occupied | |
| Beds | Inpatient | Total | numbeds | * | |
| Patients | Hospital | Known or suspected to have COVID-19 | numC19HospPats | ||
| Patients | Hospital | Known to have COVID-19 | Currently Admitted COVID-19 Patients | ||
| Patients | ICU | Known to have COVID-19 | COVID-19 Occupied ICU Beds | ||
| Ventilated patients | Hospital | Known or suspected to have COVID-19 | numC19MechVentPats | ||
| Ventilated patients | Hospital | Known to have COVID-19 | Currently Admitted COVID-19 Patients on Ventilators | Total Number of Ventilators in Use for COVID-19 | |
| Ventilators | Hospital | Available | * | Ventilators Available | |
| Ventilators | Hospital | In use | numVentUse | * | Total Number of Ventilators in Use |
| Ventilators | Hospital | Total | numVent | Total Ventilators |
Michael Donnelly (Mar 29 2020 at 15:38):
@Hans Buitendijk I guessed at a couple of things, so I bet there's a mistake or two in there.
Michael Donnelly (Mar 29 2020 at 15:39):
An asterix means this value could trivially be calculated from other values present.
David Pyke (Mar 30 2020 at 16:44):
@Hans Buitendijk @Michael Donnelly @Keith Boone I have shared a google sheet with you. Go crazy with alignment
Michael Donnelly (Mar 30 2020 at 16:45):
Cool. And good idea - that's a much better place for it.
Josh Mandel (Mar 30 2020 at 16:45):
Can you link here?
Michael Donnelly (Mar 30 2020 at 16:46):
David Pyke (Mar 30 2020 at 16:46):
Sheet available for viewing at https://docs.google.com/spreadsheets/d/1r8RFNkqzHQeelM6wHuuoURcyh57PCvaPrj-eYNySrUQ/edit?usp=sharing
Josh Mandel (Mar 30 2020 at 16:46):
Josh Mandel (Mar 30 2020 at 16:46):
(If you can enable "anyone can comment" that'd be great.)
David Pyke (Mar 30 2020 at 16:47):
Fixed
Carl Anderson (Mar 30 2020 at 18:47):
I'm a little bit troubled that the CDC measures don't seem to be collecting information about sick providers. It's likely that response coordinators will want to know which hospitals have higher rates of infected doctors/nurses so they might prioritize distribution of PPE / reinforcements. Did I miss something?
David Pyke (Mar 30 2020 at 18:48):
That would be good feedback to give to the CDC
Virginia Lorenzi (Mar 31 2020 at 14:53):
where is that original spreadsheet?
Josh Mandel (Mar 31 2020 at 15:08):
What "original" do you mean? CDC publishes a template at https://www.cdc.gov/nhsn/pdfs/covid19/import-covid19-data-508.pdf (describing their CSV) if that's what you're looking for. (And a minimal example here.)
John Moehrke (Mar 31 2020 at 15:17):
Pence points at a slightly different https://www.cdc.gov/nhsn/pdfs/covid19/57.130-covid19-pimhc-blank-p.pdf
Josh Mandel (Mar 31 2020 at 15:19):
The PDF you linked to and the CSV are the same form, from CDC; they offer form and spreadsheet based collection (thought Virginia was asking about spreadsheet). Full background is https://www.cdc.gov/nhsn/acute-care-hospital/covid19/index.html (and detailed field definitions are here) and they're running a webinar today if you want to dive in :)
John Moehrke (Mar 31 2020 at 15:22):
it is unfortunate there are two web pages on the CDC asking for the same info... Meaning, it is frustrating to have Pence pointing at a different page, even if the content of the page describes the same thing... confusing, even if 10 minutes of point by point comparison shows it is the same... 10 minutes wasted time by 5000 hospital sites, is alot of wasted time... not counting the waste outside of that group
John Moehrke (Mar 31 2020 at 15:24):
I am conflicted on the change to direct-to-federal reporting. The use of state reporting agencies seems to be overhead, but adding another pathway for data to travel will end up creating issues including duplicate data, missing data, corrupt data, and wasted time.... All things that direct-to-federal is likely intended to avoid.
Josh Mandel (Mar 31 2020 at 16:59):
I think there's only one CDC reporting program. They provide options for how to submit (CSV, PDF, web form), to meet different needs + levels of tech literacy.
Andrea Pitkus, PhD, MLS(ASCP)CM, CSM (Apr 01 2020 at 01:26):
Technically, for public health COVID-19 reporting, I'm aware of ELR (electronic lab reporting) required of performing laboratories by law, eCR(electronic case reporting) required by law of providers, both which are to the appropriate public health jurisdiction, while NHSN and the VP Pence requirements are reports required of hospitals. From a laboratory perspective, it would be nice if each public health jurisdiction could send their pos cases on to HHS (The White House), but also realize they have limited resources and are swamped currently. (ELR is HL7 2.51 messaging, and eCR is CDA based)
Andrea Pitkus, PhD, MLS(ASCP)CM, CSM (Apr 01 2020 at 01:28):
Exactly. It'll likely be in triplicate if you count ELR (lab), eCR (provider) and hospital (VP) reporting of same info.
Austin Kreisler (Apr 10 2020 at 01:09):
NHSN released new COVID-19 pathways this evening for Healthcare worker staffing and healthcare supply. There are data collection forms and CSV file templates available as well. It's all in the same location as before: https://www.cdc.gov/nhsn/acute-care-hospital/covid19/index.html
Josh Mandel (Apr 10 2020 at 01:14):
Ooh, these are interesting, because at first glance they "feel a lot more Questionnaire-y" than the forms we've looked at before, which "feel Measure-y". Obviously I'm speaking imprecisely and there's no hard line...
Keith Boone (Apr 10 2020 at 05:22):
Yes, but they certainly identify ValueSet that we should consider creating for measures. From a measurement perspective, imagine you had 10 of these forms for a given region, how would you report the results in aggregate. I think for each category I'd want at least a total of yes/no responses, and an overall total of numbers of yes and no responses. I feel pretty good about the fact that we had already identified these next two as critical areas.
Austin Kreisler (Apr 29 2020 at 16:23):
NHSN recently released a Long Term Care COVID-19 module here: https://www.cdc.gov/nhsn/ltc/covid19/index.html. Four new forms and four csv's.
Last updated: Apr 12 2022 at 19:14 UTC
