Working out true spread and death rate of COVID-19

[Jozhua]

There seems to be a lot of holes in our testing and reporting of both cases and reporting of deaths due to COVID-19. This has made coming out with more concrete spread/death rates difficult.

Having this data is important, as it allows health officials to know what policies to enact and the exact risk to the public. Up to 80% of cases are estimated to be asymptomatic, with only 20% showing symptoms. Testing of pregnant women coming to give birth in an NYC hospital may shed some light on this situation:
New England Journal of Medicine - Universal Screening for SARS-CoV-2 in Women Admitted for Delivery

Thus, 29 of the 33 patients who were positive for SARS-CoV-2 at admission (87.9%) had no symptoms of Covid-19 at presentation.

Of the 29 women who had been asymptomatic but who were positive for SARS-CoV-2 on admission, fever developed in 3 (10%) before postpartum discharge (median length of stay, 2 days). Two of these patients received antibiotics for presumed endomyometritis (although 1 patient did not have localizing symptoms), and 1 patient was presumed to be febrile due to Covid-19 and received supportive care. One patient with a swab that was negative for SARS-CoV-2 on admission became symptomatic postpartum; repeat SARS-CoV-2 testing 3 days after the initial test was positive.

If this turns out to be the case, deaths may spike for the next month or so, but then the population may gain herd immunity with a much lower human cost associated with it. If not, lockdowns will be necessary to contain the spread, if possible. Then there's the risk of asymptomatic spread, etc...

Really, an antibodies test will help us understand the true scale and come up with better estimates for cases, fatalities and case fatalities.

Also, I've noticed an uptick in the case fatality ratio in the UK, which seems to point to only more serious cases being tested, which means there is an even higher amount going unreported than before. Over the last couple of weeks, case fatality has increased from 5%ish to 20%ish... Also, due to not having a representative sample of the population routinely tested, we can't even rely on the trends of case numbers to try and work out what's going on.

I think the stable linear chart gives a false impression of containment too, as we essentially hit our testing limits to publish about 4-5000 new positive cases a day.

I'd be interested to hear everyone's thoughts, estimations and new data that perhaps samples representitive spreads of the population and sees what proportion is infected, and how many of those go onto develop more severe symptoms. Also can't wait for this antibodies test, although looks like that's going to be May time at the earliest...

 

[6862]

An interesting report, but one flaw I spotted was that the median length of stay was 2 days, which is not necessarily long enough to display symptoms. Of course some of those women will have stayed longer, but on such a small sample size I fear that the almost 90% asymptomatic figure is unreliable. I agree that it would be good if the numbers already infected are orders of magnitude greater than we know about, and it does seem likely that some mass countrywide antibody testing (random sampling of the population?) will be necessary.

 

[Bantamzen]

In the public domain at least there is still a lot of missing or unpublished data. So its going to be near on impossible to work out just what the spread / mortality rate might even be. Very occasionally the government and/or media drip feed some information around the figures & demographics of the virus. We know for example that elderly people are more vulnerable (which begs the question why has it been allowed to spread like wildfire in care homes?), and more recently there are some signs that people from certain ethnic backgrounds seem to have a higher risk of mortality (although it is not clear if this is a genetic issue or a cultural one). But as yet the bigger picture is not clear, and not helped by a convoluted reporting mechanism that means deaths from several weeks ago are recorded in the latest day's numbers. But this way go some way to explain the uptick, what we really ought to be doing is recasting previous day's figures with these late reported deaths marked against the actual day of passing. Otherwise it skews the model's forecasting, and that makes decision making more difficult.

What is of course needed is greater testing both for the virus & any associated anti-bodies. There is increasing speculation that the virus may have been in the general population longer than we currently believe, which means that the spread / mortality rate and peaks may not be accurate under the current datasets. Unfortunately without anti-body testing this would be almost impossible to prove, so getting this into action as soon as possible is vital to gaining a better understand of just exactly where we are.

 

[Mag_seven]

The government talk about flattening the curve, but does anyone know exactly what is on the X axis and Y axis of that graph. As I understand it the X axis is time and the Y axis is the number of infections. If we could see the actual numbers on that graph then we would know just how many people the government expects to get infected (Y Axis) and how long the epidemic is going to last in this country (X axis).

 

[Mogster]

There seems to be a lot of holes in our testing and reporting of both cases and reporting of deaths due to COVID-19. This has made coming out with more concrete spread/death rates difficult.

Having this data is important, as it allows health officials to know what policies to enact and the exact risk to the public. Up to 80% of cases are estimated to be asymptomatic, with only 20% showing symptoms. Testing of pregnant women coming to give birth in an NYC hospital may shed some light on this situation:
New England Journal of Medicine - Universal Screening for SARS-CoV-2 in Women Admitted for Delivery



If this turns out to be the case, deaths may spike for the next month or so, but then the population may gain herd immunity with a much lower human cost associated with it. If not, lockdowns will be necessary to contain the spread, if possible. Then there's the risk of asymptomatic spread, etc...

Really, an antibodies test will help us understand the true scale and come up with better estimates for cases, fatalities and case fatalities.

Also, I've noticed an uptick in the case fatality ratio in the UK, which seems to point to only more serious cases being tested, which means there is an even higher amount going unreported than before. Over the last couple of weeks, case fatality has increased from 5%ish to 20%ish... Also, due to not having a representative sample of the population routinely tested, we can't even rely on the trends of case numbers to try and work out what's going on.

I think the stable linear chart gives a false impression of containment too, as we essentially hit our testing limits to publish about 4-5000 new positive cases a day.

I'd be interested to hear everyone's thoughts, estimations and new data that perhaps samples representitive spreads of the population and sees what proportion is infected, and how many of those go onto develop more severe symptoms. Also can't wait for this antibodies test, although looks like that's going to be May time at the earliest...

An important point is that as they were testing for viral RNA these results are just a snapshot of who’s infectious at the time they are tested. It also suggests very recent infection, at least within 14 days or so.

Finding detectable viral RNA is usually viewed as an indicator of the individual shedding viable virus and being able to infect others. People don’t tend to become chronically infected during influenza like illnesses and RNA will only be present for a few days, then people recover and RNA disappears as the virus has been destroyed by the hosts immune system and cleared from the persons body. Detectable virus specific antibody will usually appear around the time viral RNA disappears as the persons immune system fights the infection.

An antibody test would detect people who have been infected, cleared the virus and recovered. It seems developing and mass producing a suitably specific and sensitive antibody test is proving difficult though.

 

[Domh245]

The government talk about flattening the curve, but does anyone know exactly what is on the X axis and Y axis of that graph. As I understand it the X axis is time and the Y axis is the number of infections. If we could see the actual numbers on that graph then we would know just how many people the government expects to get infected (Y Axis) and how long the epidemic is going to last in this country (X axis).

The actual numbers aren't known, knowable, nor are they particularly important - so long as the number of current infections requiring hospital treatment doesn't exceed hospital capacity we're 'on track' - people will die, but it's making sure that nobody dies who could have survived if they'd been able to be hospitalised.

 

[Mogster]

The worrying possibility raised by last weeks ONS stats is that unusual numbers of people may be dying not only from Covid infection but because of secondary factors caused by the national preoccupation with Covid.

Suggestions seem to be people failing to attend hospital for both routine or urgent care.

 

[Jozhua]

The worrying possibility raised by last weeks ONS stats is that unusual numbers of people may be dying not only from Covid infection but because of secondary factors caused by the national preoccupation with Covid.

Suggestions seem to be people failing to attend hospital for both routine or urgent care.

That's what worries me, if this turns out to be the peak of the virus, then everyone who has had appointments cancelled or been too afraid to turn up will also miss out on vital healthcare.

 

[Trackman]

The worrying possibility raised by last weeks ONS stats is that unusual numbers of people may be dying not only from Covid infection but because of secondary factors caused by the national preoccupation with Covid.

Suggestions seem to be people failing to attend hospital for both routine or urgent care.

It was on the news about A&E visits down by about 30%. One thing they found out that the number of people visiting A&E with suspected heart attacks as halved. The Doctor from the British Heart Foundation said it puts thousands at risk. He also said people are not turning up for appointments either.
Makes you think what other serious conditions people are not going to A&E with or keeping appointments.

 

[yorkie]

That's what worries me, if this turns out to be the peak of the virus, then everyone who has had appointments cancelled or been too afraid to turn up will also miss out on vital healthcare.

True; for the vast majority of the population, the biggest risk is not directly from Covid-19 but indirectly from the effects of mitigation, either by the individual concerned (such as people not going to hospital because of a fear of getting the virus) or due to restrictions on normality imposed to reduce the spread of the virus. But I think that is probably a separate topic to the one posed in this thread, so it may be worth creating a new thread to discuss the details of this. There are various news articles which could be referred to.

... Then there's the risk of asymptomatic spread, etc...

Really, an antibodies test will help us understand the true scale and come up with better estimates for cases, fatalities and case fatalities..

And on that subject, this is interesting:

No matter how you crunch the numbers, this pandemic is only just getting started | William Hanage

People are understandably looking for good news. But the truth is, we’re nowhere near controlling coronavirus, says Harvard professor Dr William Hanage

www.theguardian.com

An editorial in the British Medical Journal has reported data from China suggesting that as many as four in five cases of Sars-CoV-2 infection could be asymptomatic. It then goes on to quote people from the Centre for Evidence-Based Medicine in Oxford, who say that if this is true “What the hell are we locking down for?”

and here it is: https://www.bmj.com/content/369/bmj.m1375

...A total of 130 of 166 new infections (78%) identified in the 24 hours to the afternoon of Wednesday 1 April were asymptomatic, said China’s National Health Commission....

...other researchers, including Sergio Romagnani, a professor of clinical immunology at the University of Florence, have said they have evidence that most people infected by the virus do not show symptoms. ...

If this is true, it makes it harder to identify people who do have the virus, but on the other hand it will mean many more people will have the virus than previously believed and this would mean the death rate is potentially much lower than previously believed.

 

[Meerkat]

I am wary about the asymptomatic figures. Is that asymptomatic just at test, or are they followed up and we know they got rid of the virus without ever displaying symptoms?
If it’s the former then circumstances would bias the results - if you are testing based on good contact tracing then you will find newly infected people still incubating, if you are testing at an airport the symptomatic probably won’t have been allowed on the plane at the other end.
And again noisy experts create a fuss whilst ignoring the pressing issue - the asymptomatic level is irrelevant if the small percentage of serious cases happen in short enough period to overwhelm the health system.

 

[edwin_m]

The asympomatic level is important for contact tracing, which is one possible part of an exit strategy. If there are large numbers of them, then contact tracing all of them by manual means is impossible. Even tracing by app becomes more difficult because those without symptoms won't report themselves as potentially infected.

 

[Meerkat]

The asympomatic level is important for contact tracing, which is one possible part of an exit strategy. If there are large numbers of them, then contact tracing all of them by manual means is impossible. Even tracing by app becomes more difficult because those without symptoms won't report themselves as potentially infected.

Do we know how contagious asymptomatic people are, particularly under distancing rules?
It still isn’t a showstopper with enough tracing and testing capability - the asymptomatic are going to be linked to a symptomatic person at some point so if you can test the contacts chains quick enough....

 

[Jozhua]

I am wary about the asymptomatic figures. Is that asymptomatic just at test, or are they followed up and we know they got rid of the virus without ever displaying symptoms?
If it’s the former then circumstances would bias the results - if you are testing based on good contact tracing then you will find newly infected people still incubating, if you are testing at an airport the symptomatic probably won’t have been allowed on the plane at the other end.
And again noisy experts create a fuss whilst ignoring the pressing issue - the asymptomatic level is irrelevant if the small percentage of serious cases happen in short enough period to overwhelm the health system.

Yeah, we have limited information as to those who are symptomatic and go on to develop symptoms later, which makes it very hard to make estimates. Another example of how poor testing and tracking has been.

Even given the limitations on testing resources, the ones we do have seem to have been allocated poorly. I still believe getting a representative sample of people and testing them every 5-7 days, along with tracking symptoms would really help in trying to come up with some figures and modelling. Obviously that testing capacity is needed for healthcare, as well as the sick, but weighing that up against getting some solid evidence and prevent far more deaths from under/over-reaction is worth it I think. Just takes some initiative.

The asympomatic level is important for contact tracing, which is one possible part of an exit strategy. If there are large numbers of them, then contact tracing all of them by manual means is impossible. Even tracing by app becomes more difficult because those without symptoms won't report themselves as potentially infected.

Agree 100%. The testing needs to be sorted for any scheme like this to be properly effective.

Symptom tracking also appears to be a complete mess, we need to be doing far more in following up with patients and asking if they can remember symptoms from different stages of the infection and see if trends can be found.

 

[Meerkat]

Even given the limitations on testing resources, the ones we do have seem to have been allocated poorly

There’s no way you could justify significant research testing when they couldnt even test the people being sent into care homes.

 

[edwin_m]

Do we know how contagious asymptomatic people are, particularly under distancing rules?
It still isn’t a showstopper with enough tracing and testing capability - the asymptomatic are going to be linked to a symptomatic person at some point so if you can test the contacts chains quick enough....

Stories currently that they have tested all who were on board the US aircraft carrier and 60% of those infected are asymptomatic. This is of course a young and fit population and says nothing about how many will get symptoms in the future, and (depending what test they used) probably nothing about how many of the apparently uninfected have been infected and recovered.

But I'm pretty sure a significant proportion of those infected will develop no easily detectable symptoms, and some of those will be infectious for a period. If that's any more than a very small number then contact tracing won't be very effective, unless you automatically alert contacts of contacts as well, which would probably result in far too many false alarms.

 

[Mogster]

I am wary about the asymptomatic figures. Is that asymptomatic just at test, or are they followed up and we know they got rid of the virus without ever displaying symptoms?
If it’s the former then circumstances would bias the results - if you are testing based on good contact tracing then you will find newly infected people still incubating, if you are testing at an airport the symptomatic probably won’t have been allowed on the plane at the other end.
And again noisy experts create a fuss whilst ignoring the pressing issue - the asymptomatic level is irrelevant if the small percentage of serious cases happen in short enough period to overwhelm the health system.

“Asymptomatic“ should mean being infected and then clearing the virus while having no observable symptoms. If you are tested you should have detectable RNA and later antibodies as normal.If you are tested without symptoms and provide a positive test and then develop symptoms the next day then on the previous day you were really “pre-symptomatic” which is different.

The period between infection and producing a positive test is often referred to as the “window” period. In that time you can be infected/infectious but the infection is too early to be detected by testing.

The question is how infectious are asymptomatic or pre-symptomatic people? Early on the WHO were stating that from Chinese data true asymptomatic infection was rare and so unimportant as a mode of infection. We don’t seem to have made much progress in deciding if that’s actually true. If there have been large numbers of very mild or asymptomatic infections then obviously that changes the death rate in particular.

 

[Jozhua]

There’s no way you could justify significant research testing when they couldnt even test the people being sent into care homes.

I don't know, you see at a certain point, if someone shows all the symptoms then what difference does it make if they test positive or not?

Being able to understand the infection and how it spreads has the potential to save far more lives. Healthcare systems do it all the time, with given resources and funding, they have to decide what is worth using those on and what isn't. I have however, been wondering if countries with spare testing capacity could use it to test a random (otherwise untested) sample of a population in the middle of an outbreak. The test results likely will take longer to come back, but if it is more for scientific purposes that's fine.

Stories currently that they have tested all who were on board the US aircraft carrier and 60% of those infected are asymptomatic. This is of course a young and fit population and says nothing about how many will get symptoms in the future, and (depending what test they used) probably nothing about how many of the apparently uninfected have been infected and recovered.

But I'm pretty sure a significant proportion of those infected will develop no easily detectable symptoms, and some of those will be infectious for a period. If that's any more than a very small number then contact tracing won't be very effective, unless you automatically alert contacts of contacts as well, which would probably result in far too many false alarms.

I think these outbreaks in confined spaces such as a ship, might be a good case study for what things might look like over a wider population. Obviously the infection rate will be higher on a ship, but estimates for things like death rates could be helpful. Ships like the Diamond Princess have been studied, although it is worth considering the age groups and risk factors of those onboard. Cruise ships tend to have older passengers, whilst military ships will have mostly younger crewmembers on board.

In regards to a contact tracing system, it will very likely be ineffective with current levels of testing. If you take Germany's death rate of 2% and use it to estimate the UK's true cases from deaths, you find the UK's actual number of cases to be 685,000 (as of yesterday). Even then, Germany still won't be catching every case, so it could be closer to a million. (Being conservative about it...) This is also likely to mean most people will inevitably come close by to a positive case anyway, so we might all just be back to square one!

 

[Jozhua]

The good news is that antigen testing capacity in the UK has improved to 38,000 tests per day. Yesterday 10,000 tests went unused though, so testing is being rolled out to more frontline staff.

In merely 1750 days, the whole country can be tested ...an improvement from 4,750 just a few days ago though.

 

[Jozhua]

https://www.bbc.co.uk/news/uk-52390970

Twenty thousand households in England are being contacted to take part in a study to track coronavirus in the general population.

The study aims to improve understanding of infection levels and how many people may have immunity to the virus.

Volunteers will provide nose and throat swabs on a regular basis to see whether they have the virus.

The findings will help to inform the government's strategy for easing the lockdown.

I have to give credit where credit's due, the government is really improving on their efforts to fight this. A study to track infections on a large scale of 20,000 people is a fantastic idea and hopefully will better inform policy going forwards.

As testing expands to 100,000 a week, hopefully by the end of the month, we can expect to track things much better.

In the meantime I've been doing some of my own maths. If you take death rates of other countries and use them to estimate cases from UK deaths, you get some interesting results. A 2% rate, as seen in Germany suggests we have 900,000 cases.

A 0.8% rate, as seen in Singapore early on suggests we have 2,225,000 cases.

Obviously these figures are very speculative, but they are interesting as comparison, to perhaps see what would happen if testing was available on a very large scale.

 

[leightonbd]

The CMO for Scotland, before she resigned to spend more time with her holiday home, was busy putting some big numbers into circulation: 65,000 cases (infected people) in Scotland as at 27 March (the end of lockdown week 1; see first link). I’ve no idea how that estimate was formed.

Four weeks later and Scotland has 1600 COVID-linked deaths, or 1120 Coronavirus deaths (see second link). How might that reconcile back to the 65,000? Let’s say R was 1 throughout the lockdown period (a simplification) meaning the 65,000 became 130,000: if this is right the death rate is either side of 1%. But was the 65,000 right, or just scaremongering - I find it hard to uncover any support for the number.

Knowing how widespread it is/ has been in the population is so key to taking decisions on what next: I’m amazed Government let the testing initiative drop. What’s announced yesterday/ today is coming so late.

Coronavirus: Stay home plea as Scotland deaths rise to 33

Eight more people with Covid-19 die in Scotland as the FM urges people to observe lockdown rules this weekend.

www.bbc.co.uk

Coronavirus in Scotland - Restrictions 'for rest of year' - BBC News

Nicola Sturgeon says lifting lockdown is not a "flick of the switch moment" - measures need to be eased in a phased and careful manner.

www.bbc.co.uk

 

[Jozhua]

The CMO for Scotland, before she resigned to spend more time with her holiday home, was busy putting some big numbers into circulation: 65,000 cases (infected people) in Scotland as at 27 March (the end of lockdown week 1; see first link). I’ve no idea how that estimate was formed.

Four weeks later and Scotland has 1600 COVID-linked deaths, or 1120 Coronavirus deaths (see second link). How might that reconcile back to the 65,000? Let’s say R was 1 throughout the lockdown period (a simplification) meaning the 65,000 became 130,000: if this is right the death rate is either side of 1%. But was the 65,000 right, or just scaremongering - I find it hard to uncover any support for the number.

Knowing how widespread it is/ has been in the population is so key to taking decisions on what next: I’m amazed Government let the testing initiative drop. What’s announced yesterday/ today is coming so late.

Coronavirus: Stay home plea as Scotland deaths rise to 33

Eight more people with Covid-19 die in Scotland as the FM urges people to observe lockdown rules this weekend.

www.bbc.co.uk

Coronavirus in Scotland - Restrictions 'for rest of year' - BBC News

Nicola Sturgeon says lifting lockdown is not a "flick of the switch moment" - measures need to be eased in a phased and careful manner.

www.bbc.co.uk

I agree the government left things too late, but now the situation is improving.

I know there's been a lot of silly WHO bashing, but if they had declared an emergency earlier, then countries may have taken it more seriously and prepared testing facilities earlier.

A lot of the outbreak we are seeing now is also linked to the cases in Italy, so this really highlights the need to take the threat seriously across the world.

 

[eastdyke]

There’s no way you could justify significant research testing when they couldnt even test the people being sent into care homes.

You sure as heck can justify 'significant' research testing when the current situation continues to impact both the health and wealth of the Country! Those impacts are likely to continue to be felt for the remainder of my lifetime and of many many others too.
Up to now we have only tested subsets (in some cases subsets of subsets) of the population. That simply does not give information necessary for planning ahead.
I think that the announcement today goes at least some way to address research testing. A sample size of 20,000 households is a good number. I haven't looked at the details.

Coronavirus: Study to track infection and immunity levels

The findings will help to inform the government's strategy for easing the lockdown.

www.bbc.co.uk

Twenty thousand households in England are being contacted to take part in a study to track coronavirus in the general population.
The study aims to improve understanding of infection levels and how many people may have immunity to the virus.
Volunteers will provide nose and throat swabs on a regular basis to see whether they have the virus.

I have to assume (shouldn't I know ..) that households will be chosen at random and that testing will be carefully controlled. Not sure how 'immunity' will be tested either.

 

[yorkie]

This sounds promising...

1 in 5 New Yorkers May Have Had Covid-19, Antibody Tests Suggest (Published 2020)

Accurate antibody testing is a critical tool to determine if the pandemic has slowed enough to begin restarting the economy.

www.nytimes.com

One of every five New York City residents tested positive for antibodies to the coronavirus, according to preliminary test results described by Gov. Andrew M. Cuomo on Thursday, suggesting the virus had spread far more widely than known.

The results also provided the tantalizing prospect that many New Yorkers who never knew they had been infected — possibly as many as 2.7 million, the governor said — had already encountered the virus, and survived. Mr. Cuomo also suggested the death rate was far lower than believed....

However there are questions over the reliability of the tests. But if it does turn out such a large number had it, then that means the death rate may be even lower than previously thought. The article suggests it could possibly be as low as 0.5%.

 

[Meerkat]

You sure as heck can justify 'significant' research testing when the current situation continues to impact both the health and wealth of the Country!

You couldn’t pull 20k tests for research when that was more than the total number of tests being done!

 

[eastdyke]

I know there's been a lot of silly WHO bashing, ...........

Politically motivated perhaps? And nothing to do with the Marxist ideologies of its President Director-General and his friendships with China, Russia and at one time Robert Mugabe?

You couldn’t pull 20k tests for research when that was more than the total number of tests being done!

Nobody said that they would be done in 1 day!

 

[eastdyke]

To look at 'numbers', here are some more.
ECDC [European Centre for Disease Prevention and Control (An agency of the European Union)] publishes a daily bulletin of 'cases' and deaths.

The European Respiratory Virus Surveillance Summary (ERVISS)

This interactive dashboard provides a weekly integrated epidemiological summary for influenza, RSV and SARS-CoV-2.

www.ecdc.europa.eu

Situation update for the the EU/EEA and the UK

Each Country has its own definition of a case and the deaths may or may not include those occurring outside hospitals.
At least the numbers in the table are quoted 'per 100.000 population' rather than the bald totals that we get from both Government and Media (for impact, I know ..).
Currently (and this will of course rise each day) note the deaths per 100,000 population in particular Spain, Italy, Belgium, Germany Austria and Poland when compared to UK.
Part of the linked table:

EU/EEA and the UK	Sum of Cases	Sum of Deaths	Reported cases per 100 000 population	Reported deaths per 100 000 population
Spain	208389	21717	446.0023	46.47958
Italy	187327	25085	309.9835	41.50996
Germany	148046	5094	178.5237	6.142684
United_Kingdom	133495	18100	200.7776	27.22255
France	119151	21340	177.8712	31.85681
Belgium	41889	6262	366.7374	54.8237
Netherlands	34842	4054	202.2051	23.52734
Portugal	21982	785	213.796	7.634878
Ireland	16671	769	343.4837	15.84422
Sweden	16004	1937	157.1612	19.02157
Austria	14924	494	168.6892	5.58379
Poland	10169	426	26.77564	1.121686

 

[johnnychips]

There has been criticism in the Belgian press that it will be perceived as a hotbed of CV and this will put off future tourists. As you surmised, it is the reporting of deaths that influence the statistic: the Belgian method does not just include hospital deaths.

 

[Jozhua]

This sounds promising...

1 in 5 New Yorkers May Have Had Covid-19, Antibody Tests Suggest (Published 2020)

Accurate antibody testing is a critical tool to determine if the pandemic has slowed enough to begin restarting the economy.

www.nytimes.com

However there are questions over the reliability of the tests. But if it does turn out such a large number had it, then that means the death rate may be even lower than previously thought. The article suggests it could possibly be as low as 0.5%.

That is interesting, I wonder what proportion of the UK would have it in a 0.5% fatality scenario, and assuming deaths are uncounted in care homes.

It's around 2.7 million with the current reported death rate, so given the lag time of reporting and non reporting of care homes, at the current time it could be 3.5-4 million.

I wonder at what point in regards to death rate and level of infection they will be willing to accept to change policies. Guess it's a balance of the health effects of lockdown vs Coronavirus...

There has been criticism in the Belgian press that it will be perceived as a hotbed of CV and this will put off future tourists. As you surmised, it is the reporting of deaths that influence the statistic: the Belgian method does not just include hospital deaths.

LOL, the Belgian media is being silly for suggesting that, I mean Italy will likely take the largest hit to tourism, being the first truly developed country to have a widespread outbreak.

Right now, the countries with the most accurate data, who have the best policies will be the ones looked upon well by the international community, including potential visitors.