Scrutable

Posted: **Thu Apr 02, 2020 10:11 am**

Woodchopper wrote: Thu Apr 02, 2020 7:21 am From the FT

FT new.jpg

At this point the curves in France, Spain and Italy were starting to flatten, but they're not yet doing so in the UK and US.

There may be a need for the UK to bring in additional mitigation measures (not just more enforcement of social distancing, but also things like adding public transport capacity so essential workers don't need to be crowded together when commuting).

The US looks f.cked.

These probably ought to be given as %of country population or something (beds? money? cases? (or maybe someone can do something clever and do country border independant areas of equal size?) because of course San Marino will never reach the same number deaths as China.

It looks like being really strongly authoritarian very early has been the only thing that worked. All others look the same to me at least at log scale.

Posted: **Thu Apr 02, 2020 10:42 am**

Lew Dolby wrote: Thu Apr 02, 2020 9:13 am Woodchopper's FT article claims cases are plateau-ing - but they're plotted on exponential graphs.

On linear axes they look like they'd still be rising pretty fast.

Would be interesting in seeing them graphed like that or the actual figures to see for ourselves.

You can toggle linear vs exponential axes on this visualisation site
https://aatishb.com/covidtrends/

ETA screenshot

Posted: **Thu Apr 02, 2020 10:45 am**

science_fox wrote: Thu Apr 02, 2020 10:11 am
These probably ought to be given as %of country population or something (beds? money? cases? (or maybe someone can do something clever and do country border independant areas of equal size?) because of course San Marino will never reach the same number deaths as China.

I think that would be misleading at this stage, it will keep doubling at a certain rate irrespective of the population size, until a significant percentage of the population is immune - that's why the graphs start at a number of cases / deaths

Posted: **Thu Apr 02, 2020 10:50 am**

TopBadger wrote: Thu Apr 02, 2020 10:09 am
lpm wrote: Thu Apr 02, 2020 8:27 am Two months after we discussed oxygen... two weeks after shpalman raised the alert about very high oxygen use in Italy hospitals... the UK is discovering it could be the constraining factor in numbers treated.

https://www.theguardian.com/society/202 ... s-patients

A major NHS hospital almost ran out of oxygen for its Covid-19 patients on ventilators because it was treating so many people with the disease who needed help to breathe.

The incident, which occurred at a London teaching hospital last weekend, has prompted NHS bosses to urgently warn all NHS trusts in England to limit the number of people they put on mechanical ventilators and continuous positive airway pressure (CPAP) machines.

NHS England was so concerned by what happened that it told hospital bosses in a letter on Monday that the risk constituted a “critical safety concern” which could have major consequences for all patients relying on oxygen to stay alive. It told them to take a series of urgent actions to reduce the risk of their own oxygen supply suddenly running short because of heavy demand.

They advise trusts in the letter to:

Calculate the maximum number of patients who can be treated with high-flow [oxygen] devices such as wall CPAP and communication of this to the relevant clinical teams.

Implement safety measures to prevent accidental O2 system failure (such as limiting the number of these devices available for clinical use).

So how are oxygen supplies looking? Is there much point ordering umpteen thousand ventilators if we've not the oxygen to put through them?

Anecdata Alert: My uncle is in hospital with COVID-19 and he's hopefully on the mend. Once the worst of it was over they took him straight off the oxygen to "see how he does", he's still struggling to breathe but can just about manage without so ... no oxygen for him.

They are trialling him on remdesivir to see if that has any benefit too.

Posted: **Thu Apr 02, 2020 11:07 am**

So what's the bottleneck in liquid oxygen supply? Capacity of the air distillation plants, or capacity of the delivery network? My money would be on the latter. (No shortage of the raw material obviously.)

The BOC website gives a list of uses, including as a supplement in gas burners for steel and glass manufacture, which could take a step back for a bit.

Posted: **Thu Apr 02, 2020 11:15 am**

PeteB wrote: Thu Apr 02, 2020 10:45 am
science_fox wrote: Thu Apr 02, 2020 10:11 am
These probably ought to be given as %of country population or something (beds? money? cases? (or maybe someone can do something clever and do country border independant areas of equal size?) because of course San Marino will never reach the same number deaths as China.
I think that would be misleading at this stage, it will keep doubling at a certain rate irrespective of the population size, until a significant percentage of the population is immune - that's why the graphs start at a number of cases / deaths

I was thinking about this. Caused by some people on the twitters going on about how the graphs by country are not informative as countries are different sizes, when that particular comparison by country graph was published. I can see the logic for a single centre of growth. But when a country has several centres (several major cities), some of which may have got their first cases independently (ie not spread from the first city to get it), will this not complicate matters?

Posted: **Thu Apr 02, 2020 11:47 am**

greyspoke wrote: Thu Apr 02, 2020 11:15 am I was thinking about this. Caused by some people on the twitters going on about how the graphs by country are not informative as countries are different sizes, when that particular comparison by country graph was published. I can see the logic for a single centre of growth. But when a country has several centres (several major cities), some of which may have got their first cases independently (ie not spread from the first city to get it), will this not complicate matters?

The growth rate will be the same in both places regardless of the absolute number, so it does not matter. Every day the total gets multiplied by the same factor R in each city, and since city1*R + city2*R = (city1+city2)*R you can just lump everything together.

Only if there is reason to believe there are different growth rates, e.g. if one region is locked down and the other is not, would you need to model it separately.

(Of course, there is more to a model than just one growth rate, but generally since everything is proportional to population size the same reasoning holds even for most more complicated models.)

Posted: **Thu Apr 02, 2020 11:57 am**

56% of Britons believe that the 23rd of March was too late to introduce strict measures; 35% believe that it was the right time. Only 4% believe it was too soon.

That suggests to me that if they would have been introduced earlier, there would anyway have been a decent amount of public support.

Posted: **Thu Apr 02, 2020 11:59 am**

jaap wrote: Thu Apr 02, 2020 11:47 am
greyspoke wrote: Thu Apr 02, 2020 11:15 am I was thinking about this. Caused by some people on the twitters going on about how the graphs by country are not informative as countries are different sizes, when that particular comparison by country graph was published. I can see the logic for a single centre of growth. But when a country has several centres (several major cities), some of which may have got their first cases independently (ie not spread from the first city to get it), will this not complicate matters?
The growth rate will be the same in both places regardless of the absolute number, so it does not matter. Every day the total gets multiplied by the same factor R in each city, and since city1*R + city2*R = (city1+city2)*R you can just lump everything together.

Only if there is reason to believe there are different growth rates, e.g. if one region is locked down and the other is not, would you need to model it separately.

... and in a semi-log plot, R gives you the slope of the line, so different total populations just give a fixed vertical offset between lines.

Posted: **Thu Apr 02, 2020 12:07 pm**

One of the biggest employers in Massachusetts has announced it's moving to a 4 day work week to cut costs (i.e. staff pay), due to reduced income due to the coronavirus.

It's Boston Scientific, which makes medical devices. Income is down because elective surgeries are cancelled. I don't know if they make ventilators, but I'm sure they could make parts for them.

Posted: **Thu Apr 02, 2020 12:26 pm**

Italy's data does look like it's turning a corner. As for the US...

: Capture+_2020-04-02-13-23-17-1.png (126.18 KiB) Viewed 7478 times

Posted: **Thu Apr 02, 2020 12:43 pm**

jaap wrote: Thu Apr 02, 2020 11:47 am
greyspoke wrote: Thu Apr 02, 2020 11:15 am I was thinking about this. Caused by some people on the twitters going on about how the graphs by country are not informative as countries are different sizes, when that particular comparison by country graph was published. I can see the logic for a single centre of growth. But when a country has several centres (several major cities), some of which may have got their first cases independently (ie not spread from the first city to get it), will this not complicate matters?
The growth rate will be the same in both places regardless of the absolute number, so it does not matter. Every day the total gets multiplied by the same factor R in each city, and since city1*R + city2*R = (city1+city2)*R you can just lump everything together.

Only if there is reason to believe there are different growth rates, e.g. if one region is locked down and the other is not, would you need to model it separately.

Well yes, but that was not exactly my point. It was the values of city1 etc. which were the issue. Or, to put it another way, how would the figures look if you decided to split a large country up into several smaller ones for the purposes of the graph. The thing I didn't join up to this until just now was the starting points chosen for the x-axis on that graph, the time when k (three or five I think, can't find it now) cases were discovered in that country. That time will be earlier when you are looking at the single big country model (say, the UK) than when you are looking at the small countries (say England, Scotland, Wales and NI) separately. Or to put it another way, SUM[city0... cityn]=k.

So looking at the UK, England saw k cases a few days before any were seen in the others, so the England graph will be lower (for times after cases started cropping up elsewhere in the UK) but start in the same place, the graphs for Wales etc. will contain much lower numbers but with a few days shift to the left. Hmmmm.

Posted: **Thu Apr 02, 2020 12:58 pm**

So maybe k should have been made proportional to the population of the country?

Posted: **Thu Apr 02, 2020 1:22 pm**

shpalman wrote: Thu Apr 02, 2020 11:57 am 56% of Britons believe that the 23rd of March was too late to introduce strict measures; 35% believe that it was the right time. Only 4% believe it was too soon.

That suggests to me that if they would have been introduced earlier, there would anyway have been a decent amount of public support.

That's presumably 56% of Britons being polled now, though. I would have been interested to see the numbers at the time. On 23 March the UK had about 5000 total cases and about 230 deaths. I think you might have struggled to get a majority, or even a third, to support a lockdown. Only a week earlier, there were around 1200 cases and 21 deaths, and a lot of people would have seen it as an Italian problem.

Posted: **Thu Apr 02, 2020 1:23 pm**

El Pollo Diablo wrote: Wed Apr 01, 2020 1:35 pm The latest numbers are in. 563 new deaths, taking the total to 2,352. This is pretty much exactly in line with the trend (the doubling rate hasn't changed at all from yesterday). Expected rise tomorrow is around 600. At this rate, we should hit 10,000 deaths in less than a week.

569. So better than feared, leading to newspaper idiots writing about how we might have turned the corner.

Posted: **Thu Apr 02, 2020 1:25 pm**

Yep. Tomorrow's forecast is 738 deaths. Doubling rate steady once again - absolutely no change. We're also now pretty much 14 days behind Italy. Again.

Posted: **Thu Apr 02, 2020 1:37 pm**

lpm wrote: Thu Apr 02, 2020 1:23 pm
El Pollo Diablo wrote: Wed Apr 01, 2020 1:35 pm The latest numbers are in. 563 new deaths, taking the total to 2,352. This is pretty much exactly in line with the trend (the doubling rate hasn't changed at all from yesterday). Expected rise tomorrow is around 600. At this rate, we should hit 10,000 deaths in less than a week.
569. So better than feared, leading to newspaper idiots writing about how we might have turned the corner.

Only curve that doesn't turn corners is a straight line. We're still turning the corner. Just in the wrong direction.

Posted: **Thu Apr 02, 2020 1:50 pm**

sTeamTraen wrote: Thu Apr 02, 2020 1:22 pm
shpalman wrote: Thu Apr 02, 2020 11:57 am 56% of Britons believe that the 23rd of March was too late to introduce strict measures; 35% believe that it was the right time. Only 4% believe it was too soon.

That suggests to me that if they would have been introduced earlier, there would anyway have been a decent amount of public support.
That's presumably 56% of Britons being polled now, though. I would have been interested to see the numbers at the time. On 23 March the UK had about 5000 total cases and about 230 deaths. I think you might have struggled to get a majority, or even a third, to support a lockdown. Only a week earlier, there were around 1200 cases and 21 deaths, and a lot of people would have seen it as an Italian problem.

I humbly suggest that if the lockdown had have started say a week earlier and you polled Britons now, you would still get a decent proportion saying it was the right time, fewer saying that it was too late, and a few more saying it was too soon.

Posted: **Thu Apr 02, 2020 2:14 pm**

AMS wrote: Thu Apr 02, 2020 12:26 pm Italy's data does look like it's turning a corner. As for the US...

I think comparing confirmed cases is misleading, because of the very different detection rates between countries

Posted: **Thu Apr 02, 2020 2:24 pm**

Yes, death is the only answer. Sorry, deaths. Deaths.

Posted: **Thu Apr 02, 2020 2:28 pm**

And even death doesn't end all our problems - different measures of deadification in different countries.

It's a bit silly that the charts often show 3 countries with faked numbers - China, Iran and USA. Not coincidentally, the 3 countries with the worst outbreaks.

Posted: **Thu Apr 02, 2020 2:31 pm**

El Pollo Diablo wrote: Thu Apr 02, 2020 2:24 pm Yes, death is the only answer. Sorry, deaths. Deaths.

And even then there are differences between countries, and changes within countries over time.

I assume at some point people will start modelling the detection processes as well as the disease processes to gain unbiased estimates of true numbers of cases.

Posted: **Thu Apr 02, 2020 2:33 pm**

lpm wrote: Thu Apr 02, 2020 1:23 pm 569. So better than feared, leading to newspaper idiots writing about how we might have turned the corner.

Didn't take long. Up pops a BBC idiot to over-interpret one day's figures.

Hopes for a plateau in UK figures
Nick Triggle
Health Correspondent

Today's daily figures for the number of new cases, deaths and tests virtually mirror those released yesterday. At this stage, when we are on an upwards trajectory, the fact there have not been significant increases in terms of cases and deaths can perhaps be interpreted as a good sign... There will be hope this is where it plateaus, before dropping down.

Posted: **Thu Apr 02, 2020 2:49 pm**

In which Simon Jenkins tries to apologise but can’t quite manage to.

Posted: **Thu Apr 02, 2020 3:01 pm**

lpm wrote: Thu Apr 02, 2020 2:33 pm
lpm wrote: Thu Apr 02, 2020 1:23 pm 569. So better than feared, leading to newspaper idiots writing about how we might have turned the corner.
Didn't take long. Up pops a BBC idiot to over-interpret one day's figures.
Hopes for a plateau in UK figures
Nick Triggle
Health Correspondent

Today's daily figures for the number of new cases, deaths and tests virtually mirror those released yesterday. At this stage, when we are on an upwards trajectory, the fact there have not been significant increases in terms of cases and deaths can perhaps be interpreted as a good sign... There will be hope this is where it plateaus, before dropping down.

God almighty.

No awareness of yesterday's being a slight reversion to the mean, and therefore slightly over expectation. Or of uncertainty. Or randomness.

On only 14 of the 31 days of March did Italy see a new daily high in terms of numbers of deaths. They still saw an exponential rise, but one which had relative good days and bad days. I C&P'd a graph from here, but not before highlighting with red dots the days that an Italian version of Nick Triggle would've spaffed himself daft at the rise being less than or not that much bigger than previous days.

: Triggle, muppet.png (20.76 KiB) Viewed 7312 times

Scrutable

COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19

Re: COVID-19