Taking it to the Streets - Covid-19 #12

[Mudguard]

20 minutes ago, Maithanet said:

Reading that article, I'm fairly sure that the 31% is percentage of positive cases divided by total tests, not the percentage of all New Yorkers who had coronavirus.  The fact that number is going down is good, but it isn't really relevant to the question of herd immunity. 

It's not clear from the article, but the title of the article itself is "31 Percent Of New York City Residents Have Tested Positive For COVID-19."  I assumed they were talking about antibody test results, but it's not specified.  Looking at other articles online, I think you might be right though.  I didn't watch Cuomo's presentation, so I can't make any judgments one way or the other.

[Zorral]

33 minutes ago, Mudguard said:

It's not clear from the article, but the title of the article itself is "31 Percent Of New York City Residents Have Tested Positive For COVID-19."  I assumed they were talking about antibody test results, but it's not specified.  Looking at other articles online, I think you might be right though.  I didn't watch Cuomo's presentation, so I can't make any judgments one way or the other.

Um it did mean the test for antibodies suggested that 31% of NYC residents have them, i.e. were infected, either because they had it and recovered, or showed no symptoms.  This is up from 21 % on Friday, and 25% as of the weekend.  One in three residents of NYC, in other words.

But since Maithenet has informed everyone from his lofty position of expertise in -- well, whatever he's expert at -- my opinions are without merit and not to be believed, and deserve to be called out for that, here's somebody else's.

https://www.washingtonpost.com/health/antibody-tests-support-whats-been-obvious-covid-19-is-much-more-lethal-than-flu/2020/04/28/2fc215d8-87f7-11ea-ac8a-fe9b8088e101_story.html

 

Quote

 

 

Results from coronavirus antibody tests have started to trickle in, and they bolster the consensus among disease experts that the virus is significantly more lethal than seasonal flu and has seeded the most disruptive pandemic in the last century.

“I think it is the worst pandemic since 1918,” said Cecile Viboud, an epidemiologist at the National Institutes of Health’s Fogarty International Center, alluding to the “Great Influenza” pandemic that claimed an estimated 675,000 lives in the United States.

The new serological data, which is provisional, suggests that coronavirus infections greatly outnumber confirmed covid-19 cases, potentially by a factor of 10 or more. Many people experience mild symptoms or none at all, and never get the standard diagnostic test with a swab up the nose, so they’re missed in the official covid-19 case counts.

 

 

[Werthead]

26 minutes ago, Mudguard said:

It's not clear from the article, but the title of the article itself is "31 Percent Of New York City Residents Have Tested Positive For COVID-19."  I assumed they were talking about antibody test results, but it's not specified.  Looking at other articles online, I think you might be right though.  I didn't watch Cuomo's presentation, so I can't make any judgments one way or the other.

This much more pessimistic interpretation of the figures suggests they are unreliable, as the samples were somewhat self-selecting from people who were much more likely to be exposed to the virus in the first place. There's also a significant chance that whilst a lot of the positives came from people with a coronavirus, it might not have been this one, and even if the figures were accurate, the viral load in many of the cases may have been far too low to actually trigger the disease, rendering the fact they've had the virus to be practically worthless (since there is nothing stopping them getting the disease when exposed to a higher viral load).

Trying to manage samples like opinion polls is pretty silly and quite dangerous, and not likely to give actionable data. What you need to do is mass testing and contact tracing. If you can't do that, no figures are reliable and you can't end the lockdown until you have reliable figures, otherwise the risk of a mass outbreak even worse than the first one is much greater.

Even our notably block-headed Prime Minister acknowledged this yesterday (after his Road to Damascus conversion having actually had the virus, and an apparently bad case of it), noting that the "moment of maximum danger" will be when countries try to come out of lockdown too early, the virus spikes and the political will to reinstate the lockdown may be lacking, especially because multiple lockdowns will be much more costly to the economy than a single, long one. It's at that point you get a total meltdown as health care systems are overwhelmed. The precedent of the Spanish flu where the second wave was far deadlier than the first one still rings true, although the circumstances were somewhat different (in the case of the US, since it's more of a resurgence of the first wave).

The United States trying to end lockdowns this week is totally insane and if those states proceed, the rest of the world needs to simply block people from the US flying to their countries. Even Britain is now seriously discussing imposing a mandatory two-week quarantine on all new arrivals into the country (which they should have done two months ago) and I can see that being the next idea to spread globally after other countries have had some success with it.

[davos]

2 hours ago, Kalbear said:

If accurate, that's remarkable and a very good sign for life going forward. As horrible as this has been this might mean that the mortality and high severity rate is far less horrible than we thought. 

It also means that the spread is far greater than we thought. 

What it means is that this fucker is terrifying contagious and we are extraordinarily lucky as a species that it is not more lethal.  If the Covid-19 virus had case fatality rate of say, its cousin, the SARS virus, which had somewhere up around 9 to 10% fatality rate overall, we would be looking at tens of millions of deaths worldwide and shutdowns that, breakdown of governments and civil order, and the kind of economic devastation that is hard to really grasp.   Back in 03 we were fortunate that SARS just was not able to spread in the aggressive manner.  We have been fortunate since 2012 that MERS, which has a 35% fatality rate, does spread particularly well at all.   And we are nearly as fortunate that this virus isn't more lethal.  There is not saying that luck will hold if and when another novel corona virus emerges.  There maybe factors that make increased lethality less able to spread, but I would not count on it.  

[Buckwheat]

They are slowly opening the services here - car repair shops have opened last week, some services like hairdressers', libraries and others are opening next week. But we are still not supposed to leave our commune except for work or other essential business, and no talk about when the schools are going to be opened.

There has been a very low number of new cases in the past few days. They are trying to run a study on how many asymptomatic people have been infected and they sent invitations to 3000 random people in the country to be tested as part of the study - however, less than half of these even responded to the invitation. I personally can hardly understand why people wouldn't want to help the scientists by offering to be tested, if that is what will possibly help make further decisions and help find out what is going on with the spread, but people are apparently most worried that the test hurts.  

[Werthead]

2 minutes ago, davos said:

What it means is that this fucker is terrifying contagious and we are extraordinarily lucky as a species that it is not more lethal.  If the Covid-19 virus had case fatality rate of say, its cousin, the SARS virus, which had somewhere up around 9 to 10% fatality rate overall, we would be looking at tens of millions of deaths worldwide and shutdowns that, breakdown of governments and civil order, and the kind of economic devastation that is hard to really grasp.   Back in 03 we were fortunate that SARS just was not able to spread in the aggressive manner.  We have been fortunate since 2012 that MERS, which has a 35% fatality rate, does spread particularly well at all.   And we are nearly as fortunate that this virus isn't more lethal.  There is not saying that luck will hold if and when another novel corona virus emerges.  There maybe factors that make increased lethality less able to spread, but I would not count on it.  

There is some evidence that the more lethal a virus is, the less effective it is at spreading quickly. More lethal viruses act fast and overwhelm the body's defences before it can react, which dramatically reduces both the time between infection and death but also the time available to spread the infection. The most lethal pandemic-causing virus we know of appears to be the 1911 Chinese epidemic which was very close to 100% fatal, but fortunately (!) acted so fast it was actually quite hard to spread, and both China and global authorities reacted surprisingly efficiently to help stamp it out.

None of that is a total given, though, and a slower-acting, more infectious and high-lethal virus is certainly possible (the 1918 pandemic fits more into that category). You're right that this is a warning shot across our bows and a highlight that we may have been extraordinarily lucky to go so long (a full century) without a major global pandemic with even a moderate mortality rate taking place. It's possible we and our descendants will end up looking at the 1920-2020 period as an extraordinary one of relative peace and security from such problems, not a status quo which can ever be restored fully.

[Mudguard]

1 minute ago, davos said:

What it means is that this fucker is terrifying contagious and we are extraordinarily lucky as a species that it is not more lethal.  If the Covid-19 virus had case fatality rate of say, its cousin, the SARS virus, which had somewhere up around 9 to 10% fatality rate overall, we would be looking at tens of millions of deaths worldwide and shutdowns that, breakdown of governments and civil order, and the kind of economic devastation that is hard to really grasp.   Back in 03 we were fortunate that SARS just was not able to spread in the aggressive manner.  We have been fortunate since 2012 that MERS, which has a 35% fatality rate, does spread particularly well at all.   And we are nearly as fortunate that this virus isn't more lethal.  There is not saying that luck will hold if and when another novel corona virus emerges.  There maybe factors that make increased lethality less able to spread, but I would not count on it.  

I've been wondering about the CFR of 10% for the first SARS for a while now.  It's very possible that the actual fatality rate for the first SARS is much less than the reported CFR if many infected cases were not detected.  Not sure whether antibody surveillance testing was done to find out this information. 

According to worldometers, there's been about 213,000 deaths from about 3,130,000 cases, for a CFR of about 7%, and the number of deaths will only go up as the over 2 million active and nonrecovered cases resolve themselves.  By the time this is all over, the CFR based on the confirmed PCR tests could easily be around 10% or even higher.

[Fragile Bird]

Donald Trump keeps raving about the fact that the US has done more tests than any other country in the world, which is why the US has more cases than any other country in the world. 

We know per capita testing is lower than many other countries, but back on page 11 of this thread I looked at deaths in the US and deaths in the big case European countries that had 328M in population, like the US (more deaths than the US), and then at 20 different European countries using Italy and Germany and a bunch of smaller countries (fewer deaths).

The US has tested 5,919,847 to date. Using the big case countries (Italy, Spain, France, Germany, the UK plus Norway), they have conducted 6,658,163 tests. Then I looked at the 20 countries, and they've tested 6,907,689. I thought it would be fewer, since so many small countries are in the list.

Just to set the record straight.

[The Anti-Targ]

8 hours ago, Ran said:

Speaking of testing, I forgot to follow up with the results of the testing of staff at Danderyd Hospital in Stockholm. This is a test from the Royal Institute of Technology in Stockholm  with a private partner, SciLifeLab (who are also involved in a study to use wastewater to monitor levels of infection in communities) which they say has been 100% on positive and negative samples they've thrown at it, because of it checking across so many of the viral proteins rather than just a couple.

In any case, per their report, 20% of the 527 hospital workers tested positive. This was from samples taken a week ago, and so represents infections from ~3 weeks ago (or potentially earlier). Interestingly, they plan to make a long-term study of this particular hospital:

 

Antibody or antigen testing? One report I read said typical antibody production starts (i.e. is detectable by reliable tests) within 10 days of infection. Antigen should also be detectable within this timeframe. AFAIK, it's only symptoms that have a variable appearance timeframe.

[The Anti-Targ]

6 hours ago, Mudguard said:

I've been wondering about the CFR of 10% for the first SARS for a while now.  It's very possible that the actual fatality rate for the first SARS is much less than the reported CFR if many infected cases were not detected.  Not sure whether antibody surveillance testing was done to find out this information. 

According to worldometers, there's been about 213,000 deaths from about 3,130,000 cases, for a CFR of about 7%, and the number of deaths will only go up as the over 2 million active and nonrecovered cases resolve themselves.  By the time this is all over, the CFR based on the confirmed PCR tests could easily be around 10% or even higher.

I don't know a lot about SARS1 and MERS, but what I have picked up suggests that both viruses are highly pathogenic and only symptomatic people were infectious. So these attributes make both of these diseases much easier to track, control and eradicate. It is likely the true CFR is not far off the reported CFR. You could have been sitting on a plane next to someone brewing the infection of either of these diseases, but if they were physically well you were at no risk of infection.

Covid-19 is a very different beast with the vast majority of infections being mild to asymptomatic and so the vast majority of infected people are flying under the radar, and people are infectious before / without showing symptoms. These attributes make Covid-19 a total bastard to track and control. Testing even this early in the pandemic is already very reliably telling us in populations where community spread went uncontrolled for a period of time before lockdowns the true infection rate is likely to be 10x, 20x or maybe even 30x the official confirmed infection rate. which means we know with quite good consistency that the true CFR is 1% or less where the population is in a reasonable state of health, there is a decent healthcare system and people live in sanitary conditions with access to food and clean water, ie. no population-wide exacerbating risk factors. Then you have places like New Zealand and Australia that appear to have a really tight handle on the virus (eradication is a distinct possibility), and in both countries the CFR on reported cases is 1.3% (1.289% NZ, 1.306% AU). NZ and Aus will have missed cases no doubt, and at most with the amount of testing we've both been doing my guess is that we have picked up at least half of all infections. Iceland too has done ridiculous amounts of testing with a CFR of 0.55%. Iceland is a small sample size but my semi-educated guess is the global CFR in developed countries will be pretty close to that number. So that's 3 small(ish) developed countries with good healthcare and health monitoring systems and extensive testing that are close to kicking the virus in the teeth all with CFRs not far from 1%. So, countries with a high apparent CFR have a pretty shit handle on the actual number of cases in the country. That may be intended or unintended depending on the control (or uncontrol) policy they are trying to implement.

[Altherion]

10 hours ago, Zorral said:

It's going to be gruesome here.

https://gothamist.com/news/some-conservatives-and-conspiracists-say-its-time-reopen-nyc

You can always find a small fraction of the population with a bizarre view. Nobody is reopening New York City anytime soon. The state as a whole will remain "paused" for a little over two weeks and the entire downstate area (NYC, Long Island, Westchester, etc.) will almost certainly not reopen until June (or maybe even later). Governor Cuomo pretty much said so this afternoon.

[Ran]

3 hours ago, The Anti-Targ said:

Antibody or antigen testing? One report I read said typical antibody production starts (i.e. is detectable by reliable tests) within 10 days of infection. Antigen should also be detectable within this timeframe. AFAIK, it's only symptoms that have a variable appearance timeframe.

IgG antibodies form approximately 12 days after infection begins (or at least immune response to an infection begins), so these are the ones KTH were testing for. I'm not sure if they were also testing for IgM, which form faster but also disappear faster, to try and get a sense of how long those last.

[Jeor]

2 hours ago, Altherion said:

You can always find a small fraction of the population with a bizarre view. Nobody is reopening New York City anytime soon. The state as a whole will remain "paused" for a little over two weeks and the entire downstate area (NYC, Long Island, Westchester, etc.) will almost certainly not reopen until June (or maybe even later). Governor Cuomo pretty much said so this afternoon.

I agree with this, I think if they do reopen NYC is probably going to do it reasonably cautiously. @Zorral, I know you don't agree, but I do think there will be some natural caution around this, so it hopefully won't go to massive crowds again. Having read about your situation in the other COVID-19 thread I'm really sorry to hear about your circumstances and I can only but dimly imagine how dark the situation is.

But if there is any hope, it is that some of the population in New York will be careful, even if things do reopen, so hopefully the full congestion of the city doesn't return. The subway is going to be a real challenge, though - I've visited the city a few times and I get the impression that most New Yorkers don't own cars so are entirely reliant on the subway for transport. That will be tough.

[Which Tyler]

On 4/28/2020 at 4:19 AM, Jeor said:

Nothing surprises me about Trump, really. In fact, it would have surprised me if he hadn't said something twisted about the deaths.

Overall there have been higher death counts in cities than ones directly attributed to COVID-19, and I know there's been debate about the UK's statistic of in-hospital deaths. But to be honest, counting deaths due to COVID-19 is a very tricky business. How many of those deaths were "directly" attributable to the coronavirus (if such a judgement can indeed be made)? And then how many are deaths from pre-existing conditions (e.g. heart trouble) but the coronavirus just tipped them over the edge, when the same coronavirus in a healthy person would have been a non-fatal outcome? Hard to say, so I don't really know what to think of all those counts.

It doesn't matter (IMO)

The important number is how many people died, who wouldn't have died were it not for coronavirus.

Which includes deaths by COVID19, deaths where COVID19 was also present, and deaths caused by non-, delayed- or reduced- care due to coronavirus.

Excess deaths (compared to the average for that time period) is the only stat that matters here.

 

I've out the latest such stats for England and Wales up in the UK politics thread (including similar-but-not-the-same stats for France, Italy, Spain, Germany and Sweden)

[Clueless Northman]

1 hour ago, Which Tyler said:

Excess deaths (compared to the average for that time period) is the only stat that matters here.

Actually, you would still underestimate the number of deaths, because the lockdown definitely reduced mortality, less accidents, close to no flu deaths.

 

12 hours ago, davos said:

And we are nearly as fortunate that this virus isn't more lethal.  There is not saying that luck will hold if and when another novel corona virus emerges.  There maybe factors that make increased lethality less able to spread, but I would not count on it.  

This is my main fear. We quite fucked up our reaction this time. We have to learn and put into place a far better system to face the next one, because odds are that the next major pandemic - in a few decades - will be far deadlier, and we just can't fail then.

 

12 hours ago, Werthead said:

The United States trying to end lockdowns this week is totally insane and if those states proceed, the rest of the world needs to simply block people from the US flying to their countries. Even Britain is now seriously discussing imposing a mandatory two-week quarantine on all new arrivals into the country (which they should have done two months ago) and I can see that being the next idea to spread globally after other countries have had some success with it.

The bunch of ideological raving idiots leading continental Europe is already considering reopening the inner borders. Nice, but we're nowhere near eradication, and most have barely begun to ease out the lockdowns. Heck, the only country that plans to do complete testing is tiny Luxembourg with 700K people. All the others are still testing ridiculously few people - either for virus or antibodies - compared to what's needed to have a good view of the current situation.