Tuesday, March 17, 2020

Some of Today's Covid-19 Graphs

If you're a songwriter you can probably get away with rhyming "covid" with "morbid."

Here's a comparison of US cases of Covid-19 per capita compared to Italy's. There's still an order-and-a-half of magnitude between them, so I've divided Italy's by (an arbitrary) 35 to get them on the same graph:

Here is a graph showing how many days ago US cases today matched Italy's then. Again, per capita:

The number of days keeps increasing, so perhaps we're not on Italy's path after all. 

By the way, China had only one new case in the last day.

In the US, the number of cases is still increasing exponentially (essentially):

So far the death rate in the US is about half that of the rest of the world, but the numbers here are relatively small:

Maybe on definition of exponential growth are numbers where when you look at them you say, wow, that can't possibly continue, only to say the same thing again the next day -- wow, now that can't possibly continue! And so on.


Layzej said...

Half of the new cases in Ontario can be traced to the USA. Nonetheless, Canada has just closed its borders to all countries but the USA. I suspect they are balancing economic impacts vs health impacts. I hope they strike the right balance.

David Appell said...

Canada's cases per capita is up to 81% of the US's now.

Canada now has 441 cases with 4 deaths (up from 1 death yesterday). The virus is probably already circulating inside your country and closing the border would be too late anyway. Do you think?

David Appell said...
This comment has been removed by the author.
David Appell said...

Duh, now that I reread your comment obviously the virus is circulating within Canada.

Layzej said...

Closing the border prevents us from importing new cases. Social distancing will reduce local spread. We'll likely have 20,000 cases before the benefits of these measures are seen. Not good.

David in Cal said...

Interesting background on why Italy has so many cases

The reason [Iran and Italy] are suffering the most outside China is mainly due to their close ties with Beijing, primarily through the “One Belt and One Road” (OBOR) initiative...

Lombardy and Tuscany are the two regions that saw the most Chinese investment. Nearly a year later, the first Wuhan coronavirus infection case in Italy was reported in the Lombardy region on Feb. 21. Today, Italy is experiencing the worst coronavirus outbreak outside China, and Lombardy is the hardest-hit region in the country.



Layzej said...

The US and Canada have agreed to close our shared border. That was Canada's final (and biggest) gap. Now we just hunker down and wait.

It looks like we'll have ~20,000 cases in Canada by March 29th, at which point we should start to see a tapering off.

David Appell said...

Where does your 20,000 number come from?

Layzej said...

It's a wild ass guess based on our current rate of increase. The error bars are pretty wide on this one.

We didn't really get serious about mitigation until a few days ago. My thinking is that the number of cases will continue to grow at the present rate for two weeks. At that point everyone who is sick should already be showing symptoms. As long as we keep the number of new cases down we should be able to see a tapering off by that time.

My WAG for the USA is close to 200,000 cases by that time. I hope I'm wrong.

David Appell said...

Thanks. I hope you WAGs are at least the right order of magnitude. I hope.

David Appell said...


They call it the "Wuhan virus," which now is just a way to piss liberals off.

Italy and Iran got unlucky. Are we now supposed to avoid trading with all countries because a virus might come from any one of them?

The last paragraph of that article is ridiculous and takes political advantage of the situation, which is vulgar.

Layzej said...

I hope you WAGs are at least the right order of magnitude. I hope.

Are you thinking the number could be much higher?

Ned said...

It's the Federalist. Of course it's hideous.

Ned said...

On projections: I think in countries where testing has been inadequate (US, others) the reported number of cases is almost useless. The reported number of deaths is probably a bit more accurate, though still an underestimate since not all deaths will have been correctly attributed to CV.

Here's a simple model, based on current deaths. Assume those represent total cases from two weeks previous, scaled by a 1% to 2% death rate. That means two weeks ago there were 50 to 100 times as many cases as the current number of deaths. Assume the true number of cases (not reported cases) is doubling each 6 days, or increasing 5-fold in 14 days. Then current cases = 250 to 500 times the current number of deaths.


USA = 124 deaths (now) x 250 to 500 = 31,000 to 62,000 actual cases (now)
vs 8019 reported cases
Canada = 9 deaths (now) = 2250 to 4500 cases (now)
vs 648 reported cases

And two weeks from now:

USA = 155,000 to 310,000 actual cases
Canada = 11,250 to 22,500 actual cases

So that matches Layzej's estimates reasonably well.

Obviously this method only works while the disease is in its exponential growth phase ... hopefully soon the "doubling every 6 days" thing will be slowing down substantially.

David Appell said...

Thanks. Even 155 K is still hard to believe.

Entropic man said...

I have been putting similar numbers on UK comment threads and encountering similar incredulity.

It is a bit like climate change. People are reluctant to accept how bad things can get.

Layzej said...

It is a bit like climate change. People are reluctant to accept how bad things can get.

The hockey stick curve on most of David's recent COVID graphs is familiar as well. I'm generally optimistic about climate change and I was fairly optimistic about managing COVID as well.

The problem is that we don't see the impact of COVID mitigations for another two weeks. By that time it's too late to find out you didn't do quite enough.

This has me somewhat less optimistic about climate change. With COVID we at least have China and Italy as warnings of what may be in store for us. Will we find out too late that we missed our chance on climate?

Layzej said...

It's starting to look like something <10,000 cases in Canada by March 29th. The growth rate has been trending down day after day. Yesterday was 10% rather than the 33% we started at. It doesn't look like testing has slowed own. 98,824 tests completed at the national lab as of the 22nd vs 54,956 on the 19th.

On the other hand, there was an influx of folks returning from spring break over the last week. Hopefully they are self isolating as instructed.

Layzej said...

Looks like I was off by a few days on my US estimate. As of the 29th US was at 142,070 and growing at ~20,000/day.

I was off by quite a bit more on my Canadian guess. As of the 29th Canada was at 6,320 and growing at ~700/day.