Thanks for sharing.
I found the peek in all-case mortality also very interesting, because that way counting is much more unambiguous: dead is dead.
They showed a clear diversion from the "average" in recent weeks, but... they did not show the stdev for the averages. Finally I found a chart that shows that "outliers" are not uncommon.
There is a very noticeable spike in the worst hit countries: Italy, Spain, France, Belgium, Netherlands and the UK. The cumulative excess deaths for all Europe in the last two months matches the COVID-19 reported deaths (around 100000).
This article claims that most Covid-19 hotspots have significantly more excess deaths than reported covid deaths. Suggesting that there is a lot of underreporting. NYC being a notable exception. I think several countries count only corona deaths in hospitals, but systematically miss all deaths in care facilities. https://www.spiegel.de/wissenschaft/corona-todesfaelle-wie-v... (charts should be readable despite any language barrier)
(I don't defend either view. As the old joke goes, don't trust any statistics you haven't manipulated yourself :) Statistics and causality have always been difficult, even more so in exceptional time with imperfect short-term data only.)
Another coincidence is that the country that is most suspected of under-reporting, Germany, is also the least represented in EuroMOMO. There is data for only 2 regions.
Italy, Spain and France, the most significant contributors both in COVID-19 related deaths and excess mortality in general, all count deaths in care facilities now. I don't know about the UK though. Deaths at home are probably not counted, but according to the authorities, they are a minority: COVID-19 does not happen suddenly and people normally have time to go to the hospital. Still significant though.
My gut feeling is that there are actually ~50% more deaths than reported. But we'll have to wait for at least a few months to get proper statistics.
https://www.corriere.it/politica/20_marzo_26/the-real-death-...
New York has been criticized for retroactively adding older cases. The Johns Hopkins data just tacks it on to the end of their time series, creating a weird spike and screwing up what the data represents:
Where's the noticeable spike for Ireland? Where's the spike for Portugal? Where's the spike for Luxembourg? Where's the data for the rest of Germany outside of Hesse and Berlin? Where's the spike for Austria? (And this is not a criticism of them, but they only track Western Europe by the looks of things.)
Don't know where Euromomo is getting its data from but I suggest to you that it's incomplete.
This NYT article[0] (other publications like The Economist[1] have arrived at similar numbers) show that the cumulative excess deaths for France, Netherlands, Switzerland, Spain, and England & Wales sometimes far exceeds reported Covid-19 deaths.
Of the countries you mention only Belgium is actually reporting accurately. (As is Sweden btw, a country you do not mention.) Note that both are smallish countries.
As of 14 hours ago Chris Giles (FT economics editor) tweeted[2] "A cautious estimate of the total number of UK excess deaths linked to coronavirus stands today at 42,700" (He updates this most days.) Worldometer[3] currently has the UK on 20,319 deaths. Quite a difference.
In fact, the numbers show that in Europe actual deaths are between 1.4 [Swiss] and 2.1 [British] times higher than the reported numbers for countries that are under-reporting.
Btw, you say that there are 100,000 deaths? EU 27 has 96,533 deaths as of this moment. EU+UK has 116,852 deaths as of this moment. And Europe in its entirety[4][5] has 122,568. (Am tracking these figures using a spreadsheet.)
Let's say that the adjustment we have to make is between 1.4 and 2.1 and let's ignore population size and pick 1.75 and then lower that to take into account that some countries are accurately reporting and let's err on the conservative side so let's choose 1.66… repeating as our adjustment rate, agreed? This gives us an estimated excess # of deaths for Europe of 204,280. Twice the figure you've given.
[0] https://www.nytimes.com/interactive/2020/04/21/world/coronav...
[1] https://www.economist.com/graphic-detail/2020/04/16/tracking...
[2] https://twitter.com/ChrisGiles_/status/1254105061745098752?s...
[3] https://www.worldometers.info/coronavirus/country/uk/
[4] https://en.wikipedia.org/wiki/Europe [5] https://www.reddit.com/r/europe/new/
(Europe is generally taken to extend from the Atlantic states of Ireland, Portugal, and Iceland in the West to the Ural and Caucasus Mountains in the east, from Scandinavia in the north to Italy and Greece in the south.)
Admissions at hospitals have collapsed: in the UK they halved. Admissions due to respiratory illnesses however didn't really go up, not surprising when you consider the small absolute numbers. There is now a massive backlog of operations and diagnostics for cancer that health systems will struggle to clear in time.
There's a story with some analysis of that problem here:
https://www.telegraph.co.uk/news/2020/04/26/what-second-coro...
In the past the recommendations of epidemiologists have ended up killing a lot more than they saved, with the 2001 foot and mouth epidemic in the UK being a classic example. It's likely it will be true again this time.
It is interesting though that the median undercount is converging to ~10-20x. Perhaps the protocols across regions are similar enough that the confirmed case counts are somewhat comparable.
Unrealistically low death stats coming from Turkey compared to cases easily refute that argument.
Dead is dead, unless the state finds a way to claim that it was not a COVID19 dead.
This is what “dead is dead” mean. One can argue what should count as a COVID19 case, and how exactly we are counting. There is a lot less argument over who is dead and who is not.
It seems I actually missed the mention of "all-cause" while reading the comment.
It's just a matter of demanding tests to declare as a COVID death and do not providing enough tests.
Brazil, for example, has an artificially low count of cases due to the lack of tests and a similarly low number of deaths. However, cases of death by "pneumonia", generic types of SARS and "unexplained respiratory diseases" skyrocketed: https://oglobo.globo.com/sociedade/coronavirus/alem-da-covid...
The 1.66%, otoh, seems reasonably in line or at least compatible with what's been observed in Korea and elsewhere.
Given age is going to skew things a good deal, it seems like a picture is emerging but not that new a picture. An IFR of even 1% is pretty bad, especially given these statistics show how infectious this virus is.
Most of those tests were done on April 4th and 5th which was 3 weeks after Austria started relatively strict lockdown measures, which also impacts that number, as this will result in the test to find an even lower number of positive people.
To be clear: it would be the most dangerous general epidemic disease since the advent of vaccination, and by a significant amount. You need to go back to measles and polio to find general population outbreaks that were more lethal.
Not every death is the same - a 80 year old with weak immune system could have lived 5 years longer without corona, but a healthy 20 year-old dying from cytokine storm caused by influenza has lost potentially 60 years of healthy life - the loss is much worse.
Seeing those numbers makes me more supportive of the isolation measures.
In a triage situation, where you have to decide between different people dying, such choices are unavoidable or necessary. But I want highlight that you are talking about this stuff to say that it's OK to plan for the death of "a 80 year old with weak immune system could have lived 5 years longer" versus no death at all. And that's not OK.
One choice is providing some opportunity for further life beyond what's expected. That's generally considered something society likes but isn't obligated to provide. Society doesn't obligate it's member to spend money developing some miracle-extend that gives someone five more years.
The other choice is taking life that would normally be expected. That is something that society very much frowns on. If you could protect someone and you don't do it 'cause it would cost you money, you may wind-up in jail for murder.
Very quantitatively oriented people seem to have a hard time grasping why there's a difference here. But I think it's very rational in an evolutionary game-theoretic compact kind of way. Everyone is a member of society and values everyone else's life highly, more highly than immediate material things though maybe not more highly than other people's lives. This gives member of society basic security - you are thinking my insulin might worth just stealing and selling on the open market, me murdering you first might be my best strategy. You can see where things break down? The "social contract" is kind of the way around this.
And not doing anything (pretend it's just the flu) will result in 50M dead world wide. Everyone worried about a new depression should realize one is going to happen no matter what we do now. The only thing we can do is act in a humane fashion.
Where do you get your 50M worldwide figure? When a new flu appears, Neil Ferguson claims his 3K lies of undocumented C code forecast 200M will die. These numbers are all speculation and worse predictors than throwing darts at a board behind your back.
If you look worldwide, there are 7.8B people. If herd immunity takes 60% of the population becoming infected, that's 4.6B infections. With an IFR of 1%, that's 46.8M deaths. 460M hospitalizations (where possible).
Even say the IFR is overstated as some like. Say it's a magnitude less, comparable to the flu at .1% Now you are down to 4.7M deaths, but still the 460M hospitalizations. Still one of the most serious crises in the last 100 years.
OK, so how many "influenza death equivalents" are we looking at? What's your metric for how bad this is? I mean, I think that's a little ghoulish, obviously, but if people really want to make this argument I'd really like to see the kinds of well-founded numbers that the experts are producing. Medical ethics is hardly a new field, after all. You'd think someone would have pulled some analysis off the shelf.
Instead, the people pushing "these people would have died anyway" seem to be almost exclusively political actors (or their proxies on social media sites like this one) with a goal of either defending the inaction of the current administration or pushing a policy goal that necessarily sets the virus loose on the public.
But if you really want to make a numerate case for not trying to save the old and sick, I'd genuinely and carefully read it.
So you would need to estimate the number of years lost vs the economic damage. This is impossible to get right on both sides but at least it gives you a framework.
Basically everyone who dies from this has a preexisting condition, but basically everyone in the US will develop at least one of the big three (hypertension, diabetes, or obesity) at some point.
Kinda sucks if you have any of those conditions right now though.
You chose to eat X and not exercise for years/decades. A middle class American has enough education and purchasing power to know and behave accordingly.
Does it suck? Yes, but i find it incredibly unfair and hypocritical towards the rest of the world by ruining their lives based on the extremely old and or fat/unhealthy population.
And your comment comes across as extremely crass and insensitive.
The elderly and immunocompromised obviously die more to almost every illness. But the effect is really pronounced with covid. And most other viral infections tend to kill children at higher rates too, and covid very notably does not. It's definitely an interesting aspect of the disease, though it's produce a kind of horrifying calculus among a lot of the right wing in the US.
What do you mean? There were vaccines in 1918.
https://www.oxfordbiosystems.com/COVID-19-Rapid-test
"In order to test the detection sensitivity and specificity of the COVID-19 IgG-IgM combined antibody test, blood samples were collected from COVID-19 patients from multiple hospitals and Chinese CDC laboratories. The tests were done separately at each site. A total of 525 cases were tested: 397 (positive) clinically confirmed (including PCR test) SARS-CoV-2-infected patients and 128 non- SARS-CoV-2-infected patients (128 negative). The testing results of vein blood without viral inactivation were summarized in the Table 1. Of the 397 blood samples from SARS-CoV-2-infected patients, 352 tested positive, resulting in a sensitivity of 88.66%. Twelve of the blood samples from the 128 non-SARS-CoV-2 infection patients tested positive, generating a specificity of 90.63%."
That gives us 62% false positive ratio according to (where a study finds the prevalence to be 6% of subjects using the test):
http://vassarstats.net/clin2.html
In some cases we have research being carried out with such low positive results that they can entirely be accounted for by the low specificity. So for example if you took samples from 100 people, based on 90% specificity, even if everyone had never had corona, 10 could be found positive.
Credit to this post:
https://old.reddit.com/r/COVID19/comments/g7f373/second_roun...
However it should be noted the article in question for this submission does not mention the type of test used.
Edit: I'm looking at the reddit post but I have a lot of reservations with the "prevalence 0.06", unless we'll use the test to test absolutely everybody and not only people who are suspect. Has that calculator been validated as well?
If the test was 12 false positives in 128 negatives, how come they can claim the false positive rate is 60%?
https://www.miamidade.gov/releases/2020-04-24-sample-testing...
"Our data from this week and last tell a very similar story. In both weeks, 6% of participants tested positive for COVID-19 antibodies, which equates to 165,000 Miami-Dade County residents"
That is what the commentator is referring to in the linked post.
So if you plug their own figures into the calculator:
Sensitivity .8866 Specificity .9063
and a Prevalence of .06 based on the study, you get the 62% false positive rate.
As the prevalence increases, as with the NYC study which found the positive rate to be 21% (prevalence), the false positive rate decreases, down to 28% of the NYC study.
The Antibody test is a serology test which measures the amount of antibodies or proteins present in the blood when the body is responding to a specific infection. This test hasn’t been reviewed by the FDA. Negative results don’t rule out SARS-CoV-2 infection, particularly in those who have been in contact with the virus. Follow-up testing with a molecular diagnostic lab should be considered to rule out infection in these individuals. Results from antibody testing shouldn’t be used as the sole basis to diagnose or exclude SARS-CoV-2 infection. Positive results may be due to past or present infection with non-SARS-CoV-2 coronavirus strains, such as coronavirus HKU1, NL63, OC43, or 229E.
https://tildes.net/~health.coronavirus/o6a/coronavirus_antib...
I don't know what tests the other studies used.
https://www.reddit.com/r/LockdownSkepticism/comments/g6eqtt/...
It may be helpful to you.
Thank you for tracking these metrics.
Specifically, the Premier Biotech/Hangzhou Biotest Biotech test was validated by a Chinese provincial CDC and found 4 false positives out of 150. [1]
It was also validated by the COVID-19 Testing project and found 3 false positives out of 108. [2]
The Biomedomics test used in the Miami Dade study was also validated by the COVID-19 Testing Project and found 14 false positives out of 107. [2]
Hence I would recommend taking the results of the California and Florida studies with a huge grain of salt as the prevalence rates they found were within the false positive rates of the tests used.
[1] https://imgcdn.mckesson.com/CumulusWeb/Click_and_learn/COVID...