Hospitalization (Omicron)

[M_T]

I've not seen good enough coverage of what hospitalizations are looking like, so I thought I'd try to find it.
TL/DR:  The percentage of those with COVID that are admitted is up, but I didn't find data clarifying whether current hospitalizations are climbing.  If they aren't climbing, then the increasing percentage just represents the community-level of infection.

Santa Clara County (my home county):  The county used to post county-wide capacities, taking into account staffing levels.  That data is gone.

California:
I found this data for California counties, but it is a couple of weeks old (Dec. 24 as I write this).   I don't know whether they are describing capacity at (then-current) staffing levels, or at normal staffing levels.  EDIT:  For ICU, they indicate this is staffed beds.  But I compared against San Mateo County's data, which had no more than 65 ICU beds since Dec. 30 with no more than 6 available on any day while this site ("week of Dec. 31") showed 78 ICU beds with 23 available.

As of January 4, the California Department of Health indicates good numbers (> 15%) of staffed ICU beds available in each region.  And, that's without any widespread postponement of procedures.  Two regions are under 20% available. The state as a whole is at 20.1% available, which is at the bottom edge of where it has been since 1-Dec, but not dropping much (yet?).  The number in the ICU with COVID is steadily increasing (Dec. 1: 16.6%; Jan 4: 24.9%), but that may be because of the community rate of infection ("incidental COVID").

The CDPH did not show current hospitalization numbers that I found.  Month-old data are meaningless as there were so few Omicron cases then.


Other:
Texas Medical Center (Houston):  They have 3+ levels of ICU commitments.  Normally they run at Phase I.  If there are serious outbreaks, they can go up to Phase III, with provision for going beyond that.  Currently, they are at 90% of Phase I beds in use, of which 20.4% have COVID.  I do not believe they have required postponing of procedures at this time.
NOTE: I believe these numbers are assuming full staffing.  If, say, 20% of their staff were out sick, the number of STAFFED ICU beds might be lower than shown in these graphs.  My guess is they are actually pushing into their Phase II.

TMC is showing the number of COVID cases in the newly hospitalized to be at the same level as the peak of Delta in August (ie, the highest ever). These numbers are a little uncertain because of issues in reports over New Years.


CDC. You can see graphs of the number of COVID-19-positive admissions (per 100K) for jurisdictions and age groups.  The daily new COVID-19-positive admissions for the US has just passed the August-Sep 2021 peak but not yet reached the peak of a year ago.  California numbers are lower but climbing faster.  The US went up by 70% week over week for the week ending Jan. 2, while California was up by nearly 110%.

Again, the data doesn't show numbers as to how many are getting admitted for COVID symptoms or complications, and how many have COVID incidental to their reason for being admitted.

Another chart shows 10% of Emergency Department visits have COVID, about a third higher than Aug-Sep 2021.

Healthcare workers with COVID have exceeded the highs of a year ago.

[ChrisGreene]

I don't know the sources he uses, but John Burn-Murdoch seems to have data on cases and hospitalization:

https://twitter.com/jburnmurdoch/status/...2954221570

His analysis has been very good in the past month. He seems to know how to adjust for confounding factors.

His post refers to this a a visualization of hospital data.

https://alexanderjxchen.github.io/circuitbreaker/

I have not yet invested the time to learn what the data sources are and if there are any shenanigans going on, so consume this info with skepticism.

[M_T]

Thanks for the pointer.

Spot checking... The visualization says San Mateo County (CA) is At Capacity.
That appears to be correct when compared to
https://www.smchealth.org/data-dashboard/hospital-data
(As of Jan. 6: ICU: 1 available, 44 non-COVID, 11 COVID; Acute Care: 9 available, 350 non-COVID, 54 COVID)

It appears the SMC data is using Staffed Beds, meaning it drops if staff is unavailable. On Dec. 28, there were only 41 staffed ICU beds and 326 acute care beds. On Jan 6, there were 56 and 413.

However, "At Capacity" doesn't mean you'll be turned away. They have 67 unused ICU Surge Beds, and 134 unused Acute Care Surge Beds.

[ChrisGreene]

I worry about the San Mateo numbers. When "staffed" beds are used up, what happens? Is this where "crisis standards of care" kick in (whatever that means). I assume the care is substandard in either one or two ways:

1) Staff are pulled in from other functions in the medical system, meaning that the care givers might not have as much experience in COVID care.
2) The same staff care for more people than is normally allowed.

[M_T]

His post refers to this a a visualization of hospital data.

https://alexanderjxchen.github.io/circuitbreaker/

I have not yet invested the time to learn what the data sources are and if there are any shenanigans going on, so consume this info with skepticism.

I'm less than happy with the presentation of the data.  Unfortunately I didn't capture a picture of the data yesterday, but it looks very different today.   My guess is they are using too short an interval for deciding whether a county or state is "Has capacity", "Unsustainable", or "Forecasted to Exceed Capacity".  They list 38 states (and thus all those counties) as having 0 new cases yesterday.  Their definition of Unsustainable is given as "The single day case count yesterday is on a trend to the area reaching a circuit breaker."

I'm sure those states that have "zero" new cases yesterday will have a higher than normal count on other days, leading to wild swings between Has Capacity and Unsustainable.

By the way, an email address at Harvard is given for "discussing the data".

I see they mention some change on January 7.  Maybe they'll get it stable, but it isn't now.  (Btw, this site is apparently new, so that's Jan. 7, 2022, not 2021.)