Share on Facebook Share on Twitter

Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Ebb & Flow of COVID; Data sites
#1
California is clearly coming down off its highest level of new COVID infections.

Great, but why?

When this wave of infections started, some thought that being indoors for winter was the trigger.  Well, that still applies now.
So what is going on?  In California, we've had 3 waves, each larger than the previous.  Indeed, the trough after the 2nd wave was higher than the peak of the first in number of cases to date.

I find it easiest to look at the COVID waves in terms of R.
The state of California publishes their estimates of R (for the state, each county and regions) but only for the last 3 months.  (Look at "Nowcasts")
A group of people at UCSF published R estimates for the SF Bay Area and other selected areas of California.  These show the full history for the year, but the graphs are limted in size.
The Centre for the Mathematical Modelling of Infectious Diseases presents a map of the US with the current R for the states shown.  Other countries are available.

Digging into the CMMID data takes you to a Harvard database site with a number of interesting datasets.  In particular, there is a large (2.8MB) image with the R graphs for each state over the last 4 months.
Here's a smaller portion of it.
[Image: R-Oct5-Jan30.png]

Note that the vertical axis for each state is a slightly different scale.  Notice how each of the curves is different.  California sustained a long period of a fairly constant R, before the downturn about late November or early December.  (The curves can shift forward/backward in time,  depending on what you measure (test results, hospitalizations, or deaths) and whether you normalize those backward to estimated exposure dates.)

The application of restrictions is clearly part or much of the reason for R to drop.  As I've noted before, it seems the slopes are fairly constant, but it is uncommon for R to stay constant.  More of the time, R is dropping or rising relatively steady.  We've been unable to find an acceptable set of restrictions so we're swinging rather wildly due to overcompensation with a few months period.

If people behaved consistently to a set of restrictions, R should be constant.  But that doesn't happen.  R falls for a while and then rises again.  Some of that is clearly due to official removal of restrictions.  (For instance, California removing the Regional Stay at Home Order even though new cases are about the same as when it was instituted, and hospitalization is worse.)
Some of it is that people start cheating on the rules (for instance, the basketball coaches, non-playing players, and officials that I mentioned), wanting to be on vacation from COVID.

I tend to think we've got one more wave in front of us, but the vaccines will reduce the deaths of the older people.  I expect we'll be unable to tell which of the people not wearing masks and gathering in groups (or teams) think they are immune because they had COVID a first time, or had a vaccine, or just think it won't happen to them.

Another interesting data site is the Delphi Group at CMU.
Reply
#2
We don't understand enough about several variables to understand each little wave over time. Human behavior, viral behavior in different weather and the many ways that viral mutations may affect transmission are all somewhat unknowns.
Reply
#3
Agreed. These factors also change over time. Everything does:

* coronavirus fatigue
* weather
* holiday or not holiday times
* properties of variants
* demographics of where the virus lurks

The situation has so many unknowns, it's hard to predict based on past info.
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)