COVID-19

[AMS]

The UK official figures at least state clearly that the deaths figure is only for people hospitalised.



But the media never conveys this. It is always presented as the actual number.

Presumably in the initial weeks the hospitalisation rate was quite high, but over the next month a higher ratio will die in at home and in nursing homes - because doctors will not recommend transfers to overloaded hospitals.

One issue is that until there is a widely available and clinically useful antibody test it may not be possible to know whether someone died of Covid or another illness.

So long as tests are a scarce and vital resource it’ll be difficult to justify using them on the deceased.

Taking (body fluid) samples and preserving them for subsequent analysis when tests are readily available and cheap shouldn't be too controversial, though.

Definitely. RNA can be a bit fragile, but antibody titres in a frozen blood sample should work fine once the assays are available.

[Stranger Mouse]

One issue is that until there is a widely available and clinically useful antibody test it may not be possible to know whether someone died of Covid or another illness.

So long as tests are a scarce and vital resource it’ll be difficult to justify using them on the deceased.

Taking (body fluid) samples and preserving them for subsequent analysis when tests are readily available and cheap shouldn't be too controversial, though.

Definitely. RNA can be a bit fragile, but antibody titres in a frozen blood sample should work fine once the assays are available.

At least they won’t have to dig into the Alaskan permafrost https://www.bbc.co.uk/programmes/w3csyx3g

[EACLucifer]

The original plan was to be highly tuned to the facts on the ground, tuning up or down isolation measures at the exact moment - keep the infections suppressed but enable the economy to tick over. A nice idea on paper, but would never have been possible in real life even with huge detail on cases.

That was a complete nonsense plan which could never have worked. There are two fundamental problems with it. Firstly, there is a delay between a policy adjustment and its effect, so it is not possible to evaluate the policy until some time has passed, when it may be to late to change it. This is why you can fly a model aircraft by remote control, you might be able to remotely drive a vehicle on the moon, but it's completely impossible to drive one on Mars. Secondly, in a population where a large majority are still susceptible, the growth will be exponential, so a small error in policy will quickly magnify. Combining these effects means that for a highly infectious disease the policy must be to take extreme measures quickly as otherwise the spread is too wide to control.

Pretty much this. All of the above ^

With the lag time of infection to crisis of illness we are dealing with here, small differences in transmission speed can easily lead to enormous uncertainty in size of peak.

Disturbingly, the government still seem to be acting as though suppression and testing don't exist as approaches - despite working radically better in Korea than our idiotic approach has here - and still don't seem to understand that delaying the problem is useful, as it buys time to prepare.

[sTeamTraen]

Excess all-cause mortality will eventually show up here...

https://www.euromomo.eu/index.html

Thanks.

For the US, a great site that has pretty much all the public health data you could ever want is https://wonder.cdc.gov. I just love running queries on there through their slightly clunky interface; you can almost hear the tape spools spinning on the mainframe when you click to submit.

[lpm]

One issue is that until there is a widely available and clinically useful antibody test it may not be possible to know whether someone died of Covid or another illness.

So long as tests are a scarce and vital resource it’ll be difficult to justify using them on the deceased.

Taking (body fluid) samples and preserving them for subsequent analysis when tests are readily available and cheap shouldn't be too controversial, though.

Definitely. RNA can be a bit fragile, but antibody titres in a frozen blood sample should work fine once the assays are available.

No need for anything complicated.

Just keep a list of suspected deaths, where the care home manager or local GP say they think it was probably Covid. Two counts, official definites and judgmental unconfirmeds.

[bolo]

The Washington Post homepage has COVID-19 counters for the local area. As of tonight:

D.C. proper - 346 known cases, 5 deaths (1.4%)
Maryland - 1,240 and 16 (1.3%)
Virginia - 891and 25 (2.8%)

Is the higher death rate in Virginia likely to be because of

• statistical noise?
• worse diagnosis of nonfatal cases?
• a more virulent variant of the virus?
• less access to adequate medical care?
• cockend?

[Bird on a Fire]

So far AFAICT variation in fatalities has been linked to underlying demographic variation in the population. Is Virginia significantly older, fatter, iller or more male-skewed?

The USA will be interesting though, as so far COVID has only affected countries with a healthcare system. We're about to gain good insight into how it will spread when it reaches less developed countries.

[dyqik]

The Washington Post homepage has COVID-19 counters for the local area. As of tonight:

D.C. proper - 346 known cases, 5 deaths (1.4%)
Maryland - 1,240 and 16 (1.3%)
Virginia - 891and 25 (2.8%)

Is the higher death rate in Virginia likely to be because of

• statistical noise?
• worse diagnosis of nonfatal cases?
• a more virulent variant of the virus?
• less access to adequate medical care?
• cockend?

Differing requirements for testing and small number statistics is most likely.

Add in a couple of outbreaks in nursing homes or similar, and you can easily push the death rate per confirmed cases up by a lot.

[secret squirrel]

The Washington Post homepage has COVID-19 counters for the local area. As of tonight:

D.C. proper - 346 known cases, 5 deaths (1.4%)
Maryland - 1,240 and 16 (1.3%)
Virginia - 891and 25 (2.8%)

Is the higher death rate in Virginia likely to be because of

• statistical noise?
• worse diagnosis of nonfatal cases?
• a more virulent variant of the virus?
• less access to adequate medical care?
• cockend?

Differing requirements for testing and small number statistics is most likely.

Add in a couple of outbreaks in nursing homes or similar, and you can easily push the death rate per confirmed cases up by a lot.

A back of the envelope calculation. If we assume for the sake of argument that the base death rate is the same everywhere we can model it as a binomial distribution. Using each of these three data points to estimate the standard deviation of the population* gives results of .3%, .5% and .6%. So all these results are with a couple of standard deviations of each other. This doesn't prove anything as such, but it indicates that you might reasonably expect to see this kind of variation even without any underlying difference.

* Which we can assume to be normal, by the Central Limit Theorem.

[EllyCat]

The USA will be interesting though, as so far COVID has only affected countries with a healthcare system. We're about to gain good insight into how it will spread when it reaches less developed countries.

I know it shouldn’t, but that did amuse me.

[FlammableFlower]

The message might finally be getting through to Donald... it just took a friend in a coma and images of refrigerated trucks filled with body bags in Queens and the point that this is just the beginning to get there.

[El Pollo Diablo]

Just posting this in case it's useful at a later point. If anyone else has good comparative statistics on OECD countries or a similar group then please link.

https://www.oecd.org/health/health-syst ... artset.pdf

[greyspoke]

The message might finally be getting through to Donald... it just took a friend in a coma and images of refrigerated trucks filled with body bags in Queens and the point that this is just the beginning to get there.

A quote from that article:

Birx said the task force reviewed 12 models of the coronavirus’s spread in the United States, and it predicted between 1.6 million and 2.2 million fatalities, a worst-case scenario, if Americans did not practice social distancing and take other mitigation measures.

From what I can gather, the criticism of the UK approach is that too few independent models were plugged into the decision-making. But it may be that the report actually means one model, 12 different scenarios.

[Woodchopper]

A third of coronavirus cases may be ‘silent carriers’, classified Chinese data suggests
More than 43,000 people in China had tested positive without immediate symptoms by the end of February and were quarantined
It is still unclear what role asymptomatic transmission is playing in the global pandemic

https://www.scmp.com/news/china/society ... classified

“The number of novel coronavirus (Covid-19) cases worldwide continues to grow, and the gap between reports from China and statistical estimates of incidence based on cases diagnosed outside China indicates that a substantial number of cases are underdiagnosed,” a group of Japanese experts led by Hiroshi Nishiura, an epidemiologist at Hokkaido University, wrote in a letter to the International Journal of Infectious Diseases in February.

Based on their research, Nishiura put the proportion of asymptomatic Japanese patients evacuated from Wuhan, ground zero of the outbreak in China, at 30.8 per cent – similar to the classified Chinese government data.

But official figures from South Korea – which had carried out nearly 300,000 tests on all close contacts of its confirmed cases as of Wednesday – are the most comparable to China’s. More than 20 per cent of the asymptomatic cases reported to the Korea Centres for Disease Control and Prevention remained without symptoms until they were discharged from hospital.

“Korea currently has a significantly higher rate of asymptomatic cases than other countries, perhaps due to our extensive testing,” Jeong Eun-kyeong, director of South Korea’s CDC, told a press briefing on March 16.

Another useful point of reference is the data collected from the Diamond Princess cruise ship, which was quarantined for weeks in Yokohama, Japan. All of its passengers and crew were tested, with 712 people testing positive – 334 of whom were asymptomatic, according to official Japanese figures.

An EU report has put the proportion of asymptomatic cases in Italy at 44 per cent, but in most parts of the country people without symptoms are not tested.

In Hong Kong, 16 of the 138 confirmed cases as of March 14 were asymptomatic or presymptomatic, according to Ho Pak-leung, a professor with the microbiology department of the University of Hong Kong.

All of these numbers point to a significantly higher ratio of asymptomatic cases than indicated by data publicly released by China so far. There were 889 asymptomatic patients among the 44,672 confirmed cases as of February 11, epidemiologists from the Chinese Centre for Disease Control and Prevention wrote in a paper published online in JAMA Network Open on February 24.

More at the link.

[Martin Y]

Another rather pointless government-bashing piece on facebook from Best for Britain, quoting a (paywalled) Telegraph piece saying "Germany will issue coronavirus antibody certificates to allow quarantined to re-enter society". Cue much underinformed wot-I-reckon about how useless we are compared to everyone else in the whole world.

Anyway, it's at least an interesting thought that when there actually *is* an antibody test we might formally certify people or, rather more practically, informally adopt some kind of badge which says "it's okay, I've had it".

[FlammableFlower]

Another rather pointless government-bashing piece on facebook from Best for Britain, quoting a (paywalled) Telegraph piece saying "Germany will issue coronavirus antibody certificates to allow quarantined to re-enter society". Cue much underinformed wot-I-reckon about how useless we are compared to everyone else in the whole world.

Anyway, it's at least an interesting thought that when there actually *is* an antibody test we might formally certify people or, rather more practically, informally adopt some kind of badge which says "it's okay, I've had it".

Bit like one of those stickers kids get for being brave at the dentists?

[lpm]

Another rather pointless government-bashing piece on facebook from Best for Britain, quoting a (paywalled) Telegraph piece saying "Germany will issue coronavirus antibody certificates to allow quarantined to re-enter society". Cue much underinformed wot-I-reckon about how useless we are compared to everyone else in the whole world.

Anyway, it's at least an interesting thought that when there actually *is* an antibody test we might formally certify people or, rather more practically, informally adopt some kind of badge which says "it's okay, I've had it".

The idea of The Immune having more rights is simultaneously sensible and horrific.

[Martin Y]

Another rather pointless government-bashing piece on facebook from Best for Britain, quoting a (paywalled) Telegraph piece saying "Germany will issue coronavirus antibody certificates to allow quarantined to re-enter society". Cue much underinformed wot-I-reckon about how useless we are compared to everyone else in the whole world.

Anyway, it's at least an interesting thought that when there actually *is* an antibody test we might formally certify people or, rather more practically, informally adopt some kind of badge which says "it's okay, I've had it".

Bit like one of those stickers kids get for being brave at the dentists?

That's the kind of thing. More practical than a hen-party-style springy wire halo restyled as a crown, but less fun.

[greyspoke]

Another rather pointless government-bashing piece on facebook from Best for Britain, quoting a (paywalled) Telegraph piece saying "Germany will issue coronavirus antibody certificates to allow quarantined to re-enter society". Cue much underinformed wot-I-reckon about how useless we are compared to everyone else in the whole world.

Anyway, it's at least an interesting thought that when there actually *is* an antibody test we might formally certify people or, rather more practically, informally adopt some kind of badge which says "it's okay, I've had it".

The idea of The Immune having more rights is simultaneously sensible and horrific.

If you have a certificate to prove you are infected, you will be a valuable commodity.

[Martin Y]

Another rather pointless government-bashing piece on facebook from Best for Britain, quoting a (paywalled) Telegraph piece saying "Germany will issue coronavirus antibody certificates to allow quarantined to re-enter society". Cue much underinformed wot-I-reckon about how useless we are compared to everyone else in the whole world.

Anyway, it's at least an interesting thought that when there actually *is* an antibody test we might formally certify people or, rather more practically, informally adopt some kind of badge which says "it's okay, I've had it".

The idea of The Immune having more rights is simultaneously sensible and horrific.

With their superpowers come great responsibilities. But of course it hasn't yet been quite established that they will be immune.

[El Pollo Diablo]

Just gave blood. Donation healthcare guy said that stocks are really high at the moment, despite donations being a bit down just now. What will all the surgery being cancelled, they're not using a lot of it.

Still, they've always got a diversification outlet into black pudding if things don't pick up.

[raven]

The USA will be interesting though, as so far COVID has only affected countries with a healthcare system. We're about to gain good insight into how it will spread when it reaches less developed countries.

Here's hoping the USA will eventually be an example of how little rubbish politicians matter once they step out of the way and let the competent people take over...

India is bl..dy worrying though. How not to put a lockdown in place: Give everyone just 4 hours notice so that crowds fill the streets as everyone frantically tries to get where they want to be. Don't think about your considerable in-country migrant workforce who now have no income, no food, and no way to travel the long distances back home.

South Africa have cases in townships. Then there's Iraq. Yemen. Libya. Syria. It just doesn't bear thinking about, how bad it's going to get.

[Bird on a Fire]

The USA will be interesting though, as so far COVID has only affected countries with a healthcare system. We're about to gain good insight into how it will spread when it reaches less developed countries.

Here's hoping the USA will eventually be an example of how little rubbish politicians matter once they step out of the way and let the competent people take over...

India is bl..dy worrying though. How not to put a lockdown in place: Give everyone just 4 hours notice so that crowds fill the streets as everyone frantically tries to get where they want to be. Don't think about your considerable in-country migrant workforce who now have no income, no food, and no way to travel the long distances back home.

South Africa have cases in townships. Then there's Iraq. Yemen. Libya. Syria. It just doesn't bear thinking about, how bad it's going to get.

Yes indeed. I have a colleague who works in Guinea Bissau, which is low-development by African standards. It took her about a week to get out of the country, as she decided to pre-emptively evacuate, and that was with the help of a lot of money, good insurance and assistance from a government department.

In a lot of places, self- or even household isolation will be impossible. For example, in Guinea Bissau (and many other places) houses are used pretty much exclusively for sleeping. Cooking happens outside. Eating happens outside, often in dining areas shared by several households. Everybody eats with their hands from the same bowl. Protein depends on somebody catching an animal. Fresh vegetables depend on somebody picking one. There are very few medical staff with very few resources.

That's the norm for a huge number of people worldwide. Their only protection is that comparatively few people visit those regions. A whole load of people fleeing urban centres to return to their village will spread the infection far and wide, as we saw in Italy already, and most of the suppression techniques used so far will be impossible.

[AMS]

Though it probably won't compensate for the weaker healthcare systems, the one thing that less developed countries may have in their favour is demographics, with a higher proportion of their populations being younger. (Eg quick wiki says 44% of Nigeria's population is under 14, only 3% over 65, 2010 figures.)

[AMS]

Just gave blood. Donation healthcare guy said that stocks are really high at the moment, despite donations being a bit down just now. What will all the surgery being cancelled, they're not using a lot of it.

Still, they've always got a diversification outlet into black pudding if things don't pick up.

Surplus / out-of-date donor blood is useful for scientific research requiring immune system cells, possibly including COVID-19 research. It's an important resource for cancer immunotherapy pre-clinical work for example.