Post #1333: COVID-19 trend post-Thanksgiving

Posted on November 30, 2021

Today we get an initial look at the post-Thanksgiving trend in daily new COVID-19 cases.  So far, this seems to be shaping up for a much milder winter wave than last year.

And, as a bonus, I’ll add that if Omicron follows the path set by Delta, it’ll be too late to contribute to this year’s winter wave.  If the U.S. sees its first few cases of omicron today, and if omicron spreads at about the same rate that delta spread, it will be mid-March 2022 before Omicron is the dominant strain in the U.S.

Lot of “ifs” in that last sentence, but that’s the best estimate I can come up with at the moment.


  Trend to 11/29/2021

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 11/30/2021, 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.

It’s tough to make sense of the raw data because the holidays perturb the data.  The dip at the end of the line above is mainly due to the Thanksgiving holiday.

Most of the decline in cases is due to a temporary non-reporting of new cases.  State health department will eventually work through the backlog of reported new cases once they re-open after Thanksgiving.

But as we saw last year, in addition, holidays result in a (typically small) actual decline in new cases, even after states have fully “caught up” with any backlogs.  There’s no way to know why that occurs, but I’ve always assumed that  people who are who are only marginally ill don’t bother with getting tested over the holiday.


Hospitalizations as a better measure of true trend.

For that reason, a better way to get an instant reading on the underlying trend is to look at hospitalizations.  Unlike those who are just a little ill, people who are ill enough to require an inpatient level of care really don’t have a lot of choice about being hospitalized or not.  Hence the hospitalization data are typically less affected by holiday-related issues.

Source:  CDC COVID data tracker, accessed 11/30/2021 

Based on what the CDC is showing, there was a modest decline in new COVID-19 hospitalizations over the Thanksgiving holiday (orange line, above)   It’s still too soon to say whether that’s temporary or not, but that’s certainly a better estimate of the actual underlying trend than the raw new case count.


This year is qualitatively different from last year

This year’s COVID-19 hospitalization trend for the Thanksgiving weekend iss qualitatively different from what occurred around Thanksgiving last year.  For comparison, here’s Thanksgiving 2020 (below).  There was a one-day drop in new admissions on Thanksgiving itself (reported with a one-day lag in these data).  But that was it.  Then there was a return to the a strong upward trend, paralleling the long-run trend in new cases (red line).

Now I’m going to show the U.S. and six geographic regions, comparing this year’s winter wave to last year’s winter wave.  In a nutshell, the bigger the space between the two lines, the more this year differs from last year.

And, to get right to it, what I think I’m seeing for the U.S. as a whole, and in each region, is a growing gap between last year’s level of new cases and this year’s level.  The lines appear to be diverging.

 

There remain large sections of the country where there’s no hint of a winter wave yet.  And where there has been a winter wave — Northeast and Midwest — it’s not keeping pace with last year.

The upshot is that this is shaping up to be a much milder winter wave than last year.


When is Omicron due?

What about Omicron?  Apparently that’s quite a bit more contagious than the current (delta) variant.  When are we going to start seeing the effects of that?

The best I can do is assume omicron:delta :: delta:alpha.  That is, let me assume omicron is as big a step up in transmissibility as delta was.  And then go back and look at how long at how long it took for delta to take over from alpha.

Unfortunately, CDC doesn’t show the historical data on variants, so I have to piece this together from various news reports:

  • ~3/15/2021:  First known U.S. cases.
  • ~4/1/2021: 0.1%
  • ~5/1/2021: 1.3%
  • 6/5/2021:  10%
  • 6/19/2021:  20%
  • 7/3/2021:  50%
  • 7/17/2021:  83%
  • 9/11/2021:  99%

As you can see, it took Delta three and a half months to become the dominant strain.  That’s from the first detected case, to the point where it accounted for half of new cases.  (And then maybe another month and a half before it displaced more-or-less all the other strains.)  If omicron spreads at that rate, and we find the first few U.S. cases this week, then it will be mid-March 2022 before omicron becomes the new dominant strain in the U.S.  Assuming last year is a reasonable guide, we’ll be well past the peak of the 2021-22 winter wave by that point.