Post #1353: COVID-19 trend to 12/17/2021

Posted on December 18, 2021


In the U.S., the winter wave continues to be mainly a wave in the Northeast states.  That area is now worse off (in terms of new cases per day) than it was at this time last year.  Rhode Island is still the only only state with new COVID-19 case rates in excess of 100 / 100K / day.

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 12/18/2021, from”  The NY Times U.S. tracking page may be found at

At this point, almost all states in that region have between 60 and 100 new cases / 100K / day, and all are seeing rising rates.

For the U.S. as a whole, the winter wave stalled last week as states with declining new case rates balanced states with increases.  The balance shifted to a small net increase at the end of the week, as parts of much of South Atlantic region began to see increased case growth.  Parts of the Midwest continued to have high case rates, but for the most part those rates were stable (i.e., they did not contribute to overall growth in daily new cases).

In the South Atlantic, the majority of the states saw increases.  Most critically, Florida and Georgia began an upturn.  Those are high-population states and have a large impact on the average.

Also of note, DC’s case rate jumped.  That matters, I think, because DC is an fairly dense urban area with a lot of foreign travel into the city.  As with New York and Los Angeles, when we get a new variant, DC always seems to get it early.  I have to wonder if that jump is the first indication of a high-Omicron area in the U.S.

Where do we stand on Omicron?

Best guess, as of today (12/18/2021), Omicron accounts for one-quarter of new U.S. COVID-19 cases. 

For the week ending 12/11/2021, the CDC estimated that 2.9% of cases were Omicron. That won’t be updated again into next Tuesday (per the CDC COVID tracker website).

In the meantime, all we can do is project the figure forward using the estimated 2.5 day doubling time.  Using that doubling time, the average 2.9 percent for the week ending 12/11/2021 translates to 3.5 percent on that final day.   Starting from that, we should be right around 25% today.  Omicron should account for more than 50% of new cases as of next Tuesday.  If you assume a slightly longer 3-day doubling time, it would take another couple of days to do that.

That’s all a projection.  We won’t get another data-based estimate until next Tuesday.  This rate of growth implies that the CDC should come out on Tuesday and announce that Omicron was 12.3% of new U.S. cases for the week ending 12/18/2021.  If the announced number is in that ballpark, then the timetable above is approximately correct.

An important caveat about dates and timing.

In normal times, the exact timing of events, down to the day, hardly matters.  But when the doubling time of this new variant is 2.5 days, just a handful of days slippage in the timing can grossly change the estimates.

There are two issues here.  One is that the CDC’s weekly estimate is, itself, a projection based on older data and growth rate (per the CDC).  Just assume for the moment that the CDC”s projections are accurate.

The more important point is that the CDC doesn’t know when a person was infected.  They only know the date on which a specimen was taken. And so, while I have not been able to track it down, I believe that every date the CDC uses in its estimates of Omicron as percent of total is the date on which the sample was taken.

This matters because it takes time for infected individuals to develop symptoms and arrange to get tested.  Last time I looked, it was something like a median of four to five days between infection and onset of symptoms. 

In other words, when the CDC gives us a percentage of cases that are Omicron, and a date, I believe they are telling us something about the infections that took place roughly five days before that date, on average. 

And that means that the population actually walking around and spreading COVID, as of that date, will have experienced two additional doubling times of Omicron.

In other words, your on-the-ground, in-the-crowd risk of contracting Omicron (as opposed to Delta) is actually much higher than what is shown in the table above.  Something like 4x higher, at the lower end.  That’s because those dates are the dates that samples were taken, typically five days after the date of infection.  Omicron would have undergone two more doublings during that five day period.

And so you have to keep two different time-streams clear in your mind.  One is for the actual act of infection — the point at which you actually inhale some COVID-19 in some public indoor space.  The second is for the reporting of that infection.  That is, the date on which the infection eventually gets found out, gets reported in the data, results in a hospitalization, and so on.  Included in that reporting stream is the CDC’s projection of Omicron as percent of total.

Best guess, that second stream — the stream of case reporting — is about five days behind that first stream — your actual infection risk.  And when Omicron has a 2.5 day doubling time, the upshot of all that is that, in all likelihood, Omicron is already accounting for more than 50% of all the infections that occur each day in the U.S.  But that fact won’t make its way into the reported data until the middle of next week.

If all these projections are correct, then just walking around in a crowd today, more than half of what you’ll be exposed to is Omicron.  Today.  Every statistic you see today will be based on a far lower proportion than that.  But the actual in-the-air proportion of viruses in the U.S. on average, should already be 50% or more Omicron.