The U.S. is now down to 38 new COVID-19 cases per 100K population per day.
In some sense, we’ve now completed the U.S. Omicron wave. Today’s case count matches what it was at the start of the Omicron wave, almost two months ago to the day, on 12/17/2021.
And we’ve reached that milestone exactly one month after the peak of the wave on 1/16/2022, at which point the U.S. was seeing almost 250 new COVID-19 case per 100K population per day.
The upshot is that the U.S. Omicron wave was weirdly symmetric. It was one month straight up, then one month straight down.
The only remaining questions are a) is that an accurate reflection of new cases, and b) what happens next?
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 2/17/2022, 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
For the current downward trend there’s no letup in sight. Cases fell 45 percent in the past seven days. We’re now approaching three weeks with an almost-constant percentage rate of decline in reported new cases.
I’ve pointed out that this sustained rapid decrease seems unusual in the context of other countries’ experiences with Omicron. Most countries I’ve looked at have a curve similar to the U.K., where the rate of decline had some sort of slowdown about two weeks after the Omicron peak. The U.S. curve looks nothing like that.
Source: Johns Hopkins University via Google search.
That said, I guess it’s not really all that unusual in the context of the entire U.S. pandemic. We’ve had similar periods of more-or-less constant percentage decline in earlier parts of the pandemic. So this may well be an American thing, as opposed to an Omicron thing. Here’s the history of the pandemic in logs:
Don’t look a gift horse in the mouth?
I keep returning to the charts below, where the Omicron wave appears to be qualitatively different from the Delta wave. (And, as I recall it, from all prior waves). In the past, for any one COVID-19 variant, new hospital admissions closely tracked new cases. Like so, for Delta:
\
Source: Calculated from CDC COVID data tracker data.
That broke down entirely on the downslope of the Omicron wave. Here’s the same chart, as of data reported 2/15/2022:
Source: Calculated from CDC COVID data tracker data.
As a consequence of that breakdown, the reported case hospitalization rate continues to rise. As of 2/15/2022, on paper at least, Omicron appeared to be just about exactly as virulent as Delta was. That 6.3% apparent case hospitalization rate is right up there with the figures for the last part of the Delta wave.
Source: Calculated from CDC COVID data tracker data.
Let me emphasize that there’s nothing up my sleeve here. The chart above comes from doing long division on two data series supplied by the CDC. That’s daily new hospital admissions divided by the seven-day moving average of new cases. (This actually minimizes the issue. If I used the seven-day moving average of new admissions, the current case hospitalization rate would be over 7 percent.)
I don’t know why nobody else has talked about this publicly. I guess if you’re in the administration, and the numbers are going your way, there’s little incentive to question them. If you’re a Republican, you’re happy with any reason to put an end to COVID-19 restrictions, such as they are. So nobody has any strong incentive to ask what’s going on.
Don’t look a gift horse/elephant/donkey in the mouth.
But there’s no way that Omicron could suddenly have gotten that much more virulent without somebody noticing. There has to be something wrong with the numbers. I guess that just doesn’t much matter, right now.
At some point, I considered taking all the other available indicators of true COVID case load and lining them up as above. But after having looked at a few, I’m not sure that any of them will resolve it.
CDC tracks the fraction of hospital OPD visits that are for COVID, but that’s an odd statistic in that it’s not a count of visits, its just the fraction of the total. You can look at the data and see that, all other things equal, that varies seasonally. Hospital OPDs are busier in the winter, and that fraction is accordingly suppressed.
CDC tracks COVID DNA in waste water, but the numbers are crap. They sometimes show monthly changes in the thousands of percent. I don’t think that can be used as a quantitative measure of cases.
The Carnegie-Mellon COVIDcast project tracks individuals who report COVID-like symptoms, via a daily Facebook survey. That said, it looks like the fraction of respondents who report those symptoms is typically 10 to 20 times larger than the fraction that plausibly have a new case of COVID-19.
Eventually, the CDC will update its seroprevalence survey data. Presumably, if a lot of cases are not showing up in the official counts, there will be an increasing gap between the official count and the count based on antibodies in the blood (seroprevalence). But it’ll be a couple of months from now before we’ll see the February 2022 data, and even then, because seroprevalence reflects the entire pandemic history (up to the limits of the sensitivity of the tests), those numbers will change quite slowly.
Beyond that, I’m not seeing any independent source of information that looks even remotely useful for checking the counts.
In the end, I think this is just an un-resolvable uncertainty. Likely, when I do my final flu-versus-COVID comparison, my sole option will be to rely on hospitalizations and deaths, and simply ignore the case counts.
Has the fat lady sung, or not?
Near as I can tell, everyone outside of the CDC is assuming that the show is over, and that we’re done with COVID epidemics. Virginia, for example, has now permanently barred any sort of school mask mandate, starting March 1. Note that they didn’t just end the current set, they prevented any school district from implementing a universal mask mandate ever, no matter what, until such time as they change that legislation.
As far as I can tell, the only basis for that is reasoning-by-analogy. The 1918 flu went away, so this one will too. I believe that’s the entire depth of reasoning.
This seems imprudent to me. And that’s not just because nobody has any data from which to draw any conclusion about whether or not we’re done. Based on the last two years, we really won’t have a clue about that until this summer, when outbreaks will or will not occur in the hot-climate states as people turn on their air conditioning.
Mostly, I think it’s imprudent because COVID-19 is fundamentally different from any of the other endemic diseases we have floating around in the U.S. population. It is:
- Airborne, no close contact needed.
- Exceptionally contagious, particularly the BA.2 (second-son-of-Omicron) variant. Estimated R-nought for that is something like 22, which is more-or-less off the charts. (For reference, R-nought for typical seasonal flu runs around 1.5 to 1.75).
- Able to slip past existing antibody defenses, and
- There is no long-lasting vaccine protecting against any infection.
- There remains a huge pool of human and animal cases within which the virus may continue to mutate.
On that last one, even if we could somehow knock it out of earth’s entire human population, we’re never going to get out out of the deer, and whatever other species it may have spread to.
My bottom line is that a lot of people are acting as if this is over, now and for all eternity. But I would say that’s completely unproven. Crippling our future response, to gain political points now, just does not seem like a very smart policy move.
I guess my bottom line is that we haven’t stopped being stupid yet. We’re not looking forward, and asking how we should best respond if this flares up again. We’re steering with our eyes firmly fixed on the rear-view mirror, and barring certain public health responses that a minority of the population found objectionable. And telling ourselves that it’s OK, because, hey, the pandemic is over, why not pander? Without a thought in the world to what we’ll do if that turns out not to be the case.
So, at the end of the day, after almost two years of pandemic ups and downs, nope, we haven’t learned a damned thing.
