Things continue to settle. The U.S. is down to 22 new COVID-19 cases per 100K population per day, a 33% reduction over the past seven days.
A week ago, that figure was 43% every seven days. So the rate of decline is definitely beginning to slow. Maybe we can now start speculating on where that’s going to settle out — what will be our “endemic” level of new Omicron cases.
Data source for this and other graphs of new case counts: Calculated from The New York Times. (2021). Coronavirus (Covid-19) Data in the United States. Retrieved 2/26/2022, from https://github.com/nytimes/covid-19-data.” The NY Times U.S. tracking page may be found at https://www.nytimes.com/interactive/2020/us/coronavirus-us-cases.html
One sign of the slow return to normalcy
If I had to pick a milestone to mark the slow return to normalcy, it’s that COVID-19 is no longer pressing on hospital ICU capacity in any state. As of this week, COVID-19 cases occupy less than 30% of ICU beds in every state. The last time that was true was July 2021.
Source: Calculated from US DHHS uniform hospital data set.
It really has been that long a slog for hospital personnel.
And for that, please don’t forget to thank any unvaccinated individuals you may happen to know. The figure below, from observational data, clearly reflects both the direct impact of vaccination and some behavioral factors. But at some level, a body is a body. You really don’t have to care whether it’s foolishness about vaccination or foolishness about other things that puts any particular body in the ICU. The expense and risk caused by the small unvaccinated portion of the adult population is hugely disproportionate to their number.
Source: Taken directly from the CDC COVID data tracker, accessed 2-26-2022
Are we at the “same risk as flu” level yet?
At the end of January, I figured that at 40 cases / 100K / day, a vaccinated and boostered person’s risk of hospitalization and death from COVID-19 was no higher than the same risks from flu, in a typical flu season (Post #1400-4).
I used flu as a benchmark because we’re all familiar with it, and that level of risk seems like a natural point for judging the level of COVID-19 hygiene you wish to maintain. At first blush, once the risk from COVID falls below that for flu, plausibly, you wouldn’t take any more precautions against COVID than you would against flu. In theory, at least.
If that were true — if my 40 case benchmark were accurate — then almost all states are at the level where COVID-19 poses no more risk to a boostered person than typical seasonal flu.
But even if you focus narrowly on risk of hospitalization and death, a lot of information has come out since I did that calculation. Mostly, the Omicron case rates of hospitalization and death have risen, even as the reported cases have fallen. And every time I re-check that, they continue to rise as the Omicron wave ebbs. That didn’t happen with (e.g.) the end of the Delta wave. So something has changed.
Source: Calculated from data from the CDC COVID data tracker. Deaths offset by two weeks to account for typical lag between diagnosis and death for decedents.
I had three possible explanations of that, discussed in various earlier posts. listed here in order of my guess as to which is most important, empirically:
- Waning immunity among the elderly.
- Loss of cases from the official counts due to home testing.
- Expansion of a potentially-more-virulent strain of COVID-19.
If true, any significant decline in immunity some four to five months after the last booster shot is a particularly critical issue for assessing your own risks. It’s not as if that decline in immunity is a surprise. It’s only a question of how large a magnitude that is, and how much more slowly your protection against severe disease fades, relative to your protection against any infection.
(To see the full CDC discussion of this, read this page. It’s long, and written with all the typical scientific caveats, but you’ll get the clear impression that immunity fades rapidly enough that it’s noticeable within a typical study’s six-month time window.)
But as of today, the CDC isn’t saying much about fading immunity, in terms of its recommendations. And they’re probably not going to say that unless and until there’s a big uptick in cases. Arguably, as long as new cases continue to fall faster than immunity in the elderly falls, the elderly face falling risks. And in that situation, I think it’s unlikely that the CDC would make a recommendation for a fourth booster shot. Additionally, it would just be bad optics to start recommending a fourth booster now.
All things considered, I can think of just two situations under which we’ll be hearing from CDC about waning immunity and the need for another booster. One would be the next COVID-19 wave, if any. A second would be an actual rise in hospitalizations or deaths among the elderly, showing that the falling overall case counts are not enough to offset waning immunity in the elderly.
And the other will be this fall, when CDC is busy talking up the flu vaccine anyway. It’s still a pretty good bet that we’re going to have an annual winter COVID-19 season, just as we have an annual winter flu season. That seems to be typical, in the Northern Hemisphere, for most of the other coronaviruses. Given our past two years’ history, I’d be surprised if that weren’t true about COVID-19.
Seasonality of human coronaviruses (other than COVID-19) in Stockholm, Sweden:
Given all that, my best guess is that any push for additional boosters will come this fall, in the context of something we’re all familiar with: Flu season.
In any case, no matter how I do the math — either the current 22 cases per 100K per day, or the current 5700 hospitalizations per day — my best guess remains that they typical vaccinated-and-boostered person faces less risk of hospitalization and death from COVID-19 than from typical seasonal flu. (The hospitalizations benchmark is based on the analysis of Post #1430.) All of those numbers are now somewhat in doubt, but my sense is that we’re far enough below the “equal risk” thresholds that the average vaccinated and boostered person still faces less risk of infection or severe disease from COVID-19 than from flu.
I say this, and I grasp it intellectually. But it will be a while before my behavior fully reflects that. Plus, if the signs are still up in the stores requesting that patrons wear masks — well, I’ll wear a mask. Why not? I have what I hope is a more-than-lifetime supply of N95 and equivalent masks now.
State-level correlation of new cases and hospitalization case rate
In at least one prior post (Post #1391), I noted that COVID-19 case hospitalization rates varied widely across states.
In fact, the only aspect of state-to-state variation in hospitalization practices for COVID that made clear sense is that high-elevation states had a higher ICU use per hospitalized case. There’s just plain less oxygen in the air in the Mountain states than at sea level, and that that show up plainly in ICU for respiratory conditions, as well as (e.g.) home oxygen costs in the Medicare population.
Other than that, the COVID-19 case hospitalization rates looked like somebody threw darts at the map. There didn’t seem to be any obvious pattern that could see.
Currently, one peculiarity I’m seeing is that the handful of states with continue high new case loads (e.g., Idaho) don’t have particularly high rates of new hospitalizations. And so, as my last exercise for today, I want to check the state-level correlation between the new case rate ( cases/100K/day) and the case hospitalization rate (hospitalizations / new case).
Well, that’s about as clear as a graph gets, when it comes to health care data. That looks very much like the burden of hospitalization is about the same across the states, and the only thing that’s varying is the number of cases being reported.
I’ll definitely have to take a closer look at this, if only to make sure I didn’t mess anything up in the calculation. At the minimum, this certainly makes me take a step back and wonder what, exactly, is driving the state-to-state variation in daily new cases being reported.
I guess a small effect like this was to be expected. After all, every case hospitalized for COVID-19 will be a case that’s reported. But I never imagined anything like the nice, tight cluster of dots above, or the steep slope of the implied curve. It’s as if almost all the variation in case counts is just noise.
