Source: CDC COVID data tracker, accessed 12/30/2021
From time to time I’ve been tracking COVID-19 hospitalizations from a data file collected and maintained by the U.S. Department of Health and Human Services. You can see a daily summary of the data by state, at this link. DHHS also provides a weekly report showing individual hospitals, at this link.
In addition to counts of COVID-19 admissions and patients, that file also gathers some measures of how stressed the hospital system is. For example, it captures the fraction of hospitals that report having a critical staffing shortage; it can be used to show the fraction of beds and staffed ICU beds currently occupied; and so on.
Because US DHHS has been collecting that data for some time now, we can compare the Omicron wave to prior waves, in terms of how much stress it’s putting on the hospital system.
Here’s a graph showing occupancy rates for all beds and staffed ICU beds (top lines), and COVID-19 patients’ share of all beds and staffed ICU beds (bottom lines). The graph starts on 12/1/2020 – in the 2020 winter wave. At runs more-or-less through yesterday. (N.B. occupancy rate is just what you’d think — it’s the fraction of staffed beds that are occupied by a patient.)
Source: Calculated from this US DHHS file.
The top lines show that hospitals have been running around 70% to 80% of bed capacity throughout the pandemic. Near as I can recall, this is just a bit higher than normal. I recall that for most of my career, the U.S. hospital occupancy rate stood between 65% and 70%.
Now let me draw your attention to the note in black, on the graph above. I’m pointing at the yellow line, which shows COVID-19 cases as a percent of all staffed ICU beds. At the height of last year’s winter wave, COVID-19 cases were occupying almost one-third of all the staffed ICU beds in the U.S.
What I think is more interesting, though, is that Delta wave came quite close to that. The Delta wave resulted in far fewer cases and hospitalizations. But Delta cases must have been much more severe, on average, compared to prior variants. Despite the considerably smaller number of cases and admissions, Delta patients still occupied 30% of all staffed ICU beds at the peak of the Delta wave.
As you can see from the right edge of the graph, the current mix of Delta and Omicron cases occupies just over 20% of all staffed ICU beds now. And so while Omicron is shattering case count records, it’s not breaking ICU use records.
Are these statistics capable of identifying a hospital system in crisis? (I mean, maybe everything always looks more-or-less normal from this file. In which case, it’s useless). Here’s the same graph for Idaho. The period during which they declared state-wide crisis standards of care clearly stands out. At some point, more than 90% of all ICU beds were full, and COVID cases were occupying full 60% of all staffed ICU beds.
Source: Calculated from this US DHHS file.
OK, so it’s seems pretty clear that graphing a few simple occupancy measures can identify periods of extreme stress in a hospital system.
Now let’s do New York and Washington DC. As those are the two states with highest daily new COVID-19 case counts.
For New York (below) despite the spike in cases to more than 200 new COVID-19 cases per 100K residents per day, the hospital situation remains better than it was at the peak of the 2020 winter wave. At present, COVID cases occupy about 25% of all ICU beds, compared to 30% last winter.
For Washington, DC, this most recent spike to more than 300 new cases / 100K residents per day has brought COVID-19 ICU use back to the level seen in last winter’s wave. About one-quarter of all inpatient beds and staffed ICU beds are now occupied with COVID-19 cases.
I guess my takeaway from all this is, no crisis yet. Nothing Idaho-like, in the two states hardest-hit by Omicron. And, secondarily, Delta really did appear to result in more severe cases, on average. As Omicron gradually pushes out Delta, that, in isolation, will be a good thing.
