10.12.20

COVID19 in the US

Posted in Maps at 12:09 pm by ducky

I have been working pretty intimately with US COVID19 cases and deaths data for the past six weeks because of my COVID map application, and here are my big takeaways:

It sucks to go first.

New York and New Orleans got absolutely hammered by COVID early on, and they had a huge number of both cases and deaths. (Yellow is confirmed cases, red is confirmed deaths. Dots are individual days; lines are seven-day rolling averages. Cases are per ten thousand people while deaths are per one hundred thousand people. All of the graphs in this post are at the same scale.)

New York had a huge spike in confirmed cases, followed almost immediately by a huge spike in deaths, so immediately that it’s hard to tell the red line from the yellow line. (I believe that the odd spike in mid-May is due to a reporting issue; about 1500 deaths were reported on that day, but I bet they didn’t happen on that day.)

Note that you need to be a little careful with your interpretation here because there were almost no tests available outside of research labs until mid-March, and very limited tests for quite some time afterwards. The case count is almost certainly an undercount; there is almost no lag between the case peak and the death peak because people didn’t get diagnosed until they were close to death.

Similarly, New Orlean’s first spike was followed almost immediately by a huge spike in deaths. But look at the later two spikes: deaths did not jump significantly after the later two spikes.

I see this over and over. For example, Miami got hit very hard in July:

While Miami did see an increase in deaths, it was not nearly as bad as the deaths New York and Miami saw in April.

Why is this? Unclear. I was not the only person to notice it: here’s a Washington Post article about the phenomenon.

These are a bunch of possibilities:

  1. The virus has mutated and gotten less deadly. The article above says that genetic studies do not bear that out, but doesn’t give any more details.
  2. More testing. While it is true that more testing makes the case count higher, that isn’t a big enough effect to account for the numbers.
  3. The virus is hitting younger people. Another way to look at it might be “we have learned how to protect nursing homes”. I don’t have enough data to be able to tell how big a factor that is.
  4. More people have (at least limited) immunity due to prior non-COVID19 corona virus infections. (Many “common cold” viruses are coronaviruses.) I don’t think that would explain it, unless you think that people in New Orleans didn’t have colds before March but did later on.
  5. We are better at treating the disease. This is certainly true. Giving steroids at the right time appears to be significantly better than not.
  6. People are getting infected lower doses of virus and hence are not getting as sick. This also appears to be the case. The viral load of people showing up at hospitals has been dropping, according to the Washington Post article. This is almost certainly due to people wearing masks, standing six feet apart, avoiding restaurants, etc.

The virus moves.

COVID-19 hotspots move around. First, it was New York.

13 April 2020

(This is a cartogram, where counties’ areas are proportional to their population. Big regions mean cities; small regions are rural. Yellow is good, brown is bad.)

In the summer the Sunbelt got hit.

26 July 2020

Currently, the hotspots are in the northern central-west US — Wisconsin, Wyoming, Montana, Utah, etc., and scattered rural areas.

1 October 2020

There are very few places where the virus has not gotten to — yet. Places that have had outbreaks, particularly which have had nasty outbreaks (I’m looking at you, Northeast), are mostly doing a good job of managing now.

It almost seems like places aren’t able to learn from everyone else’s misery, they need to have a nasty outbreak themselves to truly believe that it can happen to them.

Reopening universities in person was a bad idea.

Cases spiked in early September in college towns:

Home of the University of Illinois
Home of the University of Iowa
Home of Indiana University
Home of the University of Missouri
Home of the University of Pennsylvania

Go VT, NH, ME, PA!

The far northeastern states have done an outstanding job, with almost no cases — yet.

Aside from some minor outbreaks early on and an outbreak right now in Centre County — which is almost certainly due to Penn State University reopening (see above) — Pennsylvania has done a really good job — so far.

Western Oregon and Western Washington have also done a good job — so far. Colorado has mostly done a good job — so far.