COVID19

[Banquo]

God help the poor bastards in refugee camps the world over when it starts going through them.

I was thinking the same for the Brazilian favellas and Indian slums. I suppose the average there is lower but it still won’t be pretty.

Townships in SA....I don't think they have a chance if it gets a grip in those.

[Sandydragon]

God help the poor bastards in refugee camps the world over when it starts going through them.

I was thinking the same for the Brazilian favellas and Indian slums. I suppose the average there is lower but it still won’t be pretty.

Townships in SA....I don't think they have a chance if it gets a grip in those.

South Africa at least seems to be taking this seriously. The Brazilian Trump less so.

[Banquo]

I was thinking the same for the Brazilian favellas and Indian slums. I suppose the average there is lower but it still won’t be pretty.

Townships in SA....I don't think they have a chance if it gets a grip in those.

South Africa at least seems to be taking this seriously. The Brazilian Trump less so.

Seems is the operative word. I'm not sure I believe a word of what is being said. Mind that applies in many cases.

[Mellsblue]

Hope:
COVID patients in critical care beds drop for the first time.

Not sure how this correlates with peak deaths. Italy’s peaked before ICU numbers followed. In Spain ICU numbers are still going up despite deaths (hopefully) peaking.

[Donny osmond]

Hope:
COVID patients in critical care beds drop for the first time.

Not sure how this correlates with peak deaths. Italy’s peaked before ICU numbers followed. In Spain ICU numbers are still going up despite deaths (hopefully) peaking.

Depends on therapies I guess. Italy (I'm told) have/had a policy of basically getting everyone on a ventilator as a blanket rule, whereas I think that is now not considered the best therapy for cv19 patients, at least not initially. I imagine the prevailing therapy for cv19 patients will dictate how many ICU beds are being used? Shooting from the hip here.

Sent from my CPH1951 using Tapatalk

[Digby]

I think I've heard in this country patients on ventilators and/or requiring adrenaline to boost low BP were for whatever reason also being hydrated less but that's caused problems with kidney function and they're now hydrating more.

[Mellsblue]

Less hope:
I haven’t seen the ICU numbers myself but read that despite having less cases in ICU the U.K. deaths are still comparable to Italy, Spain and France.

[Mellsblue]

Latest confirmed death log. To keep MP happy, I feel duty bound to point out that Denmark have pretty stringent social distancing measures and are doing a lot better than fellow Scandi Sweden.

[morepork]

I will peer review you until you love me Mells.

[Mellsblue]

I will peer review you until you love me Mells.

More like Stockholm syndrome.......

[Son of Mathonwy]

I'll ask it again:
D) OK, you tell me which repressive methods the Chinese used which made the difference between success and failure. And how did South Korea succeed without them?

Where are you getting this stuff about the app? Don't you understand that the main part of contact tracing is simply asking the patient where they've been and who they were in contact with? The high-tech approach helps but it's just a small part of it.

You're still not getting my old post about the death rates. Can't tell if you're misremembering or just not understanding it.

I’ll say it again, quarantine measures involving martial law, for one. We get annoyed when Derbyshire police use drones to shame people not adhering to rules and Northants police threaten to check people’s shopping backs. China would see that as a monumental step forward in human rights. That’s before they’ve been lying throughout this whole thing and you still want to believe their figures.

South Korea has managed fine without martial law, so how can you argue that martial law was essential to China's success?

The S Koreans have been tracking and tracing via people’s phones, ie with an app, whether new or piggy backed I’m not sure. The problem with simply asking the patient where they’ve been and who they’ve been in contact with is you can have no symptoms for days or even weeks. People will then need to remember who they’ve seen for the last 5-14 days, including those they sat next to in the train, stood next to in the shops etc etc. The issue with even a small number of people not complying is that this virus has such a high R value that it’s spreads very quickly, coupled with no symptoms for so long one non-compliant person can keep spreading whilst asymptomstic and those they infect will and so on and so on.
It also seems that S Korea has a virtually cashless society and has been tracking the population that way, too, ass well as using CCTV to track and trace. All very Orwellian and something probably unacceptable here and possibly illegal.
It also seems that Seoul has now closed bars, nightclubs etc.....

I'm not saying it's not useful, but it's just one of several methods for adding to info gathered directly from the patient. And most would be fine about it - if they're happy with Facebook and Google tracking them, why would they complain about the government doing it for a time-limited period to save lives and the economy?

The only old post I remember you putting up about death rates is that we’d all be dead within weeks.

That's not what I said. No worries, it's on page 12 of this thread, just take a look.

[Son of Mathonwy]

Comparing countries on per capita deaths would be better.

The graph suggests Germany and Holland are at a similar level whereas, per capita, Holland's deaths are 4.7x Germany's. It makes the USA look worse than it is; China too, of course.
Ignoring tiny territories, Spain and Italy have the worst per capita numbers, followed by Belgium, France, Holland then UK.
Denmark looks like it's doing a great job but in fact, per capita, it's worse than Germany.

There’s plenty of reasons why a per capita isn’t particularly useful. Here’s one reason why:

Basically, it’s log scale, ie how well is a country managing to contain the virus.
The other reason is it drops the US to near the bottom of the chart and they can claim they no longer have a problem. The US is a huge country with pockets of dense population. The govt and their advisers can totally feck up in stopping the virus spread but a per capita graph wouldn’t show that as whilst NY is in a world of trouble, and now the global epicentre, other places within the US aren’t anywhere near as bad.
ie different countries face different issues in combating the spread.

I appreciate it's a log scale. My point is that it's based on the absolute numbers.
It would be better if it was the log of the per capita numbers.
That would still allow us to read the exponential rate of growth as the gradient of the curves, but also it would stop misleading us into thinking that, say, Denmark is doing unusually well when it isn't, or that Germany and Holland are similar when Germany's doing much better. In general, the per capita numbers (whether on a log scale or not) level the playing field for smaller and larger countries.

As to whether or not biasing the graph against the US (and China) is a good thing - I guess that's a matter of choice for the newspaper producing the graph. I wouldn't choose it - it means you're not comparing like with like.

[Son of Mathonwy]

Indeed and a number of scientists are starting to break ranks and suggest that we could return to school and limited work. How anyone is supposed to make sense of this. Some reporting today on allowing under 25s to return to work if they don’t live with vulnerable adults, but how that can be managed or how businesses would operate is another matter.

This stuff is worrying. I know it's like the only thing the journalists can think to ask is "when will the lockdown end???", but we can't stop it until the epidemic is under control - and we are nowhere near that. Luckily the government is showing some sense on this point.

[Mellsblue]

I'll ask it again:
D) OK, you tell me which repressive methods the Chinese used which made the difference between success and failure. And how did South Korea succeed without them?

Where are you getting this stuff about the app? Don't you understand that the main part of contact tracing is simply asking the patient where they've been and who they were in contact with? The high-tech approach helps but it's just a small part of it.

You're still not getting my old post about the death rates. Can't tell if you're misremembering or just not understanding it.

I’ll say it again, quarantine measures involving martial law, for one. We get annoyed when Derbyshire police use drones to shame people not adhering to rules and Northants police threaten to check people’s shopping backs. China would see that as a monumental step forward in human rights. That’s before they’ve been lying throughout this whole thing and you still want to believe their figures.

South Korea has managed fine without martial law, so how can you argue that martial law was essential to China's success?

The S Koreans have been tracking and tracing via people’s phones, ie with an app, whether new or piggy backed I’m not sure. The problem with simply asking the patient where they’ve been and who they’ve been in contact with is you can have no symptoms for days or even weeks. People will then need to remember who they’ve seen for the last 5-14 days, including those they sat next to in the train, stood next to in the shops etc etc. The issue with even a small number of people not complying is that this virus has such a high R value that it’s spreads very quickly, coupled with no symptoms for so long one non-compliant person can keep spreading whilst asymptomstic and those they infect will and so on and so on.
It also seems that S Korea has a virtually cashless society and has been tracking the population that way, too, ass well as using CCTV to track and trace. All very Orwellian and something probably unacceptable here and possibly illegal.
It also seems that Seoul has now closed bars, nightclubs etc.....

I'm not saying it's not useful, but it's just one of several methods for adding to info gathered directly from the patient. And most would be fine about it - if they're happy with Facebook and Google tracking them, why would they complain about the government doing it for a time-limited period to save lives and the economy?

The only old post I remember you putting up about death rates is that we’d all be dead within weeks.

That's not what I said. No worries, it's on page 12 of this thread, just take a look.

I won’t engage any further on any arguments that China are some sort of example to follow. They have lied continuously and had a martial law lockdown beyond anything a liberal democracy would tolerate.

Re: tracking. I doubt people would be happy about it, see reaction to over zealous police. You’re equating an app people download of their own volition, that has social upsides (let’s not go into the numerous and weighty downsides) compared to state snooping of your movements. Apples v pears. At the end of the day neither of us truly knows so let’s avoid going round in a circle. You seems to have ignored the bits about tracking people on CCTV and via card purchases. There are very strict European laws about how people’s data can be used. You seem to completely misunderstand how the S Korean track and trace is/was working. They’ve also now shutdown Seoul’s night time economy so even their Orwellian response might not be as foolproof as you think. One article I read stated this was due to one idiot super spreader working his way round numerous clubs in one night, ie it spreads like wild fire and only needs a few morons to ruin it for everyone.

I’m not trawling back through your posts. You had a complete meltdown weeks ago and your ability to be levelled headed and understand the nuances of it all still seems to be missing.

[Mellsblue]

Comparing countries on per capita deaths would be better.

The graph suggests Germany and Holland are at a similar level whereas, per capita, Holland's deaths are 4.7x Germany's. It makes the USA look worse than it is; China too, of course.
Ignoring tiny territories, Spain and Italy have the worst per capita numbers, followed by Belgium, France, Holland then UK.
Denmark looks like it's doing a great job but in fact, per capita, it's worse than Germany.

There’s plenty of reasons why a per capita isn’t particularly useful. Here’s one reason why:

Basically, it’s log scale, ie how well is a country managing to contain the virus.
The other reason is it drops the US to near the bottom of the chart and they can claim they no longer have a problem. The US is a huge country with pockets of dense population. The govt and their advisers can totally feck up in stopping the virus spread but a per capita graph wouldn’t show that as whilst NY is in a world of trouble, and now the global epicentre, other places within the US aren’t anywhere near as bad.
ie different countries face different issues in combating the spread.

I appreciate it's a log scale. My point is that it's based on the absolute numbers.
It would be better if it was the log of the per capita numbers.
That would still allow us to read the exponential rate of growth as the gradient of the curves, but also it would stop misleading us into thinking that, say, Denmark is doing unusually well when it isn't, or that Germany and Holland are similar when Germany's doing much better. In general, the per capita numbers (whether on a log scale or not) level the playing field for smaller and larger countries.

As to whether or not biasing the graph against the US (and China) is a good thing - I guess that's a matter of choice for the newspaper producing the graph. I wouldn't choose it - it means you're not comparing like with like.

Have you read the whole thread? It states clearly why per capita is misleading from an expert in such things. You probably know better, though.

[Digby]

Indeed and a number of scientists are starting to break ranks and suggest that we could return to school and limited work. How anyone is supposed to make sense of this. Some reporting today on allowing under 25s to return to work if they don’t live with vulnerable adults, but how that can be managed or how businesses would operate is another matter.

This stuff is worrying. I know it's like the only thing the journalists can think to ask is "when will the lockdown end???", but we can't stop it until the epidemic is under control - and we are nowhere near that. Luckily the government is showing some sense on this point.

(a) it's not the only thing they're asking, there are lots of questions on what does the plan to reopen look like even if they can't say when, and that's just on coming out of lockdown

and (b) the government is starting to wobble, this is already costing a lot more than they figured, and what they figured was horrendous

(c) the slightly interesting thing is this may have gone beyond government dictate, we might actually have a civil decision to stay at home to protect the NHS as things stand and the government are going to need something to change behaviour around that beyond saying we didn't mean/expect quite this many of you to stay at home

[Son of Mathonwy]

I appreciate it's a log scale. My point is that it's based on the absolute numbers.
It would be better if it was the log of the per capita numbers.
That would still allow us to read the exponential rate of growth as the gradient of the curves, but also it would stop misleading us into thinking that, say, Denmark is doing unusually well when it isn't, or that Germany and Holland are similar when Germany's doing much better. In general, the per capita numbers (whether on a log scale or not) level the playing field for smaller and larger countries.

As to whether or not biasing the graph against the US (and China) is a good thing - I guess that's a matter of choice for the newspaper producing the graph. I wouldn't choose it - it means you're not comparing like with like.

Have you read the whole thread? It states clearly why per capita is misleading from an expert in such things. You probably know better, though.

The thread in twitter? Yes, I've taken a look through it - loads of posts so it's difficult to know what you're referring to. The main thrust of the first posts is that log is better than linear for depicting exponential processes - is that what you mean?

I'm not arguing with that - I have no problem with a log scale. As I've just said, my point is that it would be better to use the per capita numbers rather than the absolute numbers.

[Mellsblue]

I appreciate it's a log scale. My point is that it's based on the absolute numbers.
It would be better if it was the log of the per capita numbers.
That would still allow us to read the exponential rate of growth as the gradient of the curves, but also it would stop misleading us into thinking that, say, Denmark is doing unusually well when it isn't, or that Germany and Holland are similar when Germany's doing much better. In general, the per capita numbers (whether on a log scale or not) level the playing field for smaller and larger countries.

As to whether or not biasing the graph against the US (and China) is a good thing - I guess that's a matter of choice for the newspaper producing the graph. I wouldn't choose it - it means you're not comparing like with like.

Have you read the whole thread? It states clearly why per capita is misleading from an expert in such things. You probably know better, though.

The thread in twitter? Yes, I've taken a look through it - loads of posts so it's difficult to know what you're referring to. The main thrust of the first posts is that log is better than linear for depicting exponential processes - is that what you mean?

I'm not arguing with that - I have no problem with a log scale. As I've just said, my point is that it would be better to use the per capita numbers rather than the absolute numbers.

The whole point is you track the growth. The really big fecker with this COVID is the R0. Controlling spread is key and the end game of controlling spread is controlling the rate at which people die and the log scale curve flattens. The aim of every govt is to “flatten the curve” to control the rate of the spread, to control the numbers in the ICU and to control the number of deaths - the log scale measures this.
Unless you close your country to the entire world it will get there, it’ll probably get to you anyway. At that point it’s all about slowing down the rate of deaths - the length of time being asymptomatic and the time it takes to eventually kill means you don’t know what you’re dealing with until min 3 weeks after the fact.
Under a per capita system, smaller pops look like they’re fecking it up and larger pops look like they’re doing pretty well because even if a smaller country is controlling spread, and therefore deaths, you’ll still look worse than a bigger country who aren’t doing as well.
It’s not perfect, and can be complemented by per capita for context, but it’s the best tool for measuring every govts aim.

[Son of Mathonwy]

Have you read the whole thread? It states clearly why per capita is misleading from an expert in such things. You probably know better, though.

The thread in twitter? Yes, I've taken a look through it - loads of posts so it's difficult to know what you're referring to. The main thrust of the first posts is that log is better than linear for depicting exponential processes - is that what you mean?

I'm not arguing with that - I have no problem with a log scale. As I've just said, my point is that it would be better to use the per capita numbers rather than the absolute numbers.

The whole point is you track the growth. The really big fecker with this COVID is the R0. Controlling spread is key and the end game of controlling spread is controlling the rate at which people die and the log scale curve flattens. The aim of every govt is to “flatten the curve” to control the rate of the spread, to control the numbers in the ICU and to control the number of deaths - the log scale measures this.
Unless you close your country to the entire world it will get there, it’ll probably get to you anyway. At that point it’s all about slowing down the rate of deaths - the length of time being asymptomatic and the time it takes to eventually kill means you don’t know what you’re dealing with until min 3 weeks after the fact.

For what you just said it makes no difference whether you use log of absolute deaths or log of per capita deaths.

Under a per capita system, smaller pops look like they’re fecking it up and larger pops look like they’re doing pretty well because even if a smaller country is controlling spread, and therefore deaths, you’ll still look worse than a bigger country who aren’t doing as well.
It’s not perfect, and can be complemented by per capita for context, but it’s the best tool for measuring every govts aim.

If a big country and a small country are controlling the spread equally well it will grow at the same rate in each, correct?
Say big country is 10x population of small country.
Assume an equal proportion of their populations are infected initially, say 100 and 10 people respectively.
In a week, they grow at the same rate, so that the new numbers are 200 and 20.
Absolute numbers say the big country's infections increased by 100, and the small one by 10. Does that mean the big country is handling the problem worse? No - the infection rate is the same in each.
But a graph of absolute numbers will make the big country look 10 times worse than the small one.
Per capita they will look the same - which surely is the fair representation since they have the same infection rate.

[Mellsblue]

The thread in twitter? Yes, I've taken a look through it - loads of posts so it's difficult to know what you're referring to. The main thrust of the first posts is that log is better than linear for depicting exponential processes - is that what you mean?

I'm not arguing with that - I have no problem with a log scale. As I've just said, my point is that it would be better to use the per capita numbers rather than the absolute numbers.

The whole point is you track the growth. The really big fecker with this COVID is the R0. Controlling spread is key and the end game of controlling spread is controlling the rate at which people die and the log scale curve flattens. The aim of every govt is to “flatten the curve” to control the rate of the spread, to control the numbers in the ICU and to control the number of deaths - the log scale measures this.
Unless you close your country to the entire world it will get there, it’ll probably get to you anyway. At that point it’s all about slowing down the rate of deaths - the length of time being asymptomatic and the time it takes to eventually kill means you don’t know what you’re dealing with until min 3 weeks after the fact.

For what you just said it makes no difference whether you use log of absolute deaths or log of per capita deaths.

Under a per capita system, smaller pops look like they’re fecking it up and larger pops look like they’re doing pretty well because even if a smaller country is controlling spread, and therefore deaths, you’ll still look worse than a bigger country who aren’t doing as well.
It’s not perfect, and can be complemented by per capita for context, but it’s the best tool for measuring every govts aim.

If a big country and a small country are controlling the spread equally well it will grow at the same rate in each, correct?
Say big country is 10x population of small country.
Assume an equal proportion of their populations are infected initially, say 100 and 10 people respectively.
In a week, they grow at the same rate, so that the new numbers are 200 and 20.
Absolute numbers say the big country's infections increased by 100, and the small one by 10. Does that mean the big country is handling the problem worse? No - the infection rate is the same in each.
But a graph of absolute numbers will make the big country look 10 times worse than the small one.
Per capita they will look the same - which surely is the fair representation since they have the same infection rate.

I don’t know how else to explain it to you but...no. I suppose...does the virus spread less quickly because a country has a lower pop? No it spreads by it’s R0 regardless. The virus’s R0 is the same wherever it is. If the virus takes no account of pop size why would you use per capita as a measure? Al that matters is containing the spread, which ultimately manifests in deaths, so just measure by gross numbers.

[morepork]

Per capita only works if there is a consensus methodology for sampling applied across the board. At this stage of the game the more useful dissection requires normalising testing methodology. This normalising component includes using the same test with the same bio marker. I don’t see anywhere near that level of consensus at the moment, which is a massive confound to modeling. You can’t control multiple variables at the one time, and testing methodology is the greatest issue currently. Achieve testing consensus, then we can have a good old constructive argument.

[morepork]

Ideally, in a cold hearted clinical context we would like to plot total deaths for this time of year against deaths during the onset of community virus spread (again differences in testing are a confound) in a population that has resisted social distancing. An unethical experiment variable yes, but it would sort the argument out. Would require transparency in normal data. All this globalism and no reliable commonly accessible data base. We really do fuck ourselves with the long fetid dick of the free market don’t we?

[canta_brian]

America summed up in one picture.

[Digby]

Ideally, in a cold hearted clinical context we would like to plot total deaths for this time of year against deaths during the onset of community virus spread (again differences in testing are a confound) in a population that has resisted social distancing. An unethical experiment variable yes, but it would sort the argument out. Would require transparency in normal data. All this globalism and no reliable commonly accessible data base. We really do fuck ourselves with the long fetid dick of the free market don’t we?

If you binned the free market I'm going to suggest you'd still have travel around the globe. And I suppose you could eliminate every country in the world and just have the Federation that Star Trek speaks of, but you might not get any country in the world thinking that's a good idea, not even the likes of Yemen, Syria, Wales and North Korea

[Mellsblue]

Hope:
Sarah Gilbert, professor of vaccinology at Oxford University, told The Times she was “80 per cent confident” that the vaccine being developed by her team would work, with human trials due to begin in the next fortnight.