>> The actual fatality rate of Covid-19 is the region of 0.1%
> How can we be sure until we have widespread testing?
I think we can be sure that this is an incorrect number, with the data we have right now. At least 8448 people have died of COVID-19 in New York City, and if only 0.1% of the infected die, then we have 8,448,000 infected New Yorkers. This is the entire population of the city (including those who fled to other states) and would mean that we've reached herd immunity and can stop worrying about it now.
> At least 8448 people have died of COVID-19 in New York City
Just to add: be very careful to compare deaths across countries. Every country seems to count deaths differently. Some only count it if you didn't have any other condition. E.g. you had something with your lungs? That death might not be reported.
In Netherlands they compared the amount of people dying per week. There was a sudden huge increase of deaths. It seemed to indicate the number of deaths for Netherlands was 2x as high as they were reporting. Though they did report the deaths with a very clear remark that the numbers were highly likely to be too low. Unfortunately such remarks were often ignored by the public, leading to eventual distrust of any figure. Interestingly, the press at various times did make it clear that the figure could be unreliable. Seems people often just read a headline :-p
PS: My point is that the 8448 might be greatly higher. Depends on how the 8448 was collected/reported.
>it's a fair bit nastier, certainly in some cases, or the hospitals wouldn't be full
While I agree that it probably is indeed nastier than your average flu, it actually wouldn't have to be much nastier to achieve health system saturation in places where it hits groups that are particularly vulnerable (the elderly mainly). Almost all healthcare systems operate on fairly thin margins of tolerance for overload, and a virus of even mild percentages for hospital-worthy or lethal cases could easily, temporarily overload clinical infrastructure if it comes out of the blue and is moving through a population that was entirely virgin to it before the infection in question arrived. This lack of immunity among all parts of society means massive numerical surges in cases very quickly and consequently, major numerical surges in that small fraction of cases that end up being bad enough to warrant hospitalization.
We won't be sure of such statistics whether or not we have widespread testing. Even for pandemics in the past, we just don't know how many people caught them, not even a single significant digit. The best we can do is make reasoned guesses based on the available evidence.
I recommend watching the whole video, as it is by far the clearest argument in favor of the "herd immunity" approach I've seen.
The problem with the approach is that it assumes several things that, shall we say, there is not yet evidence for.
1. The disease is relatively mild, certainly less than an order of magnitude more fatal than influenza.
2. It is possible to protect older people and other vulnerable populations while the disease spreads through most of the rest of society.
3. After people recover from the infection, there is lasting immunity.
If these three things are true (and some other things that I'm not going to argue with, including that it will take a long time to get a vaccine), then herd immunity is a reasonable strategy. But let's look at each in turn.
1. The death rate in New York City is already 0.1% of the entire population. Even under very strict assumptions, that we're exactly at the peak of a totally symmetrical curve, and that herd immunity results in 50% of the population being infected, that results in a lower bound of the IFR at 0.4%. Lombardy gives similar results (0.12% of the population directly attributed to Covid-19)
2. Prof. Giesecke admitted that they failed to do so in Sweden. It is not clear how this could be done, as elderly people do not live in a bubble, but rather have lots of workers coming in and out of the facility to help take care of them.
3. SARS-CoV-2 is a new virus, and we just don't have the data yet. You can extrapolate from existing coronaviruses, which suggest that immunity will last at least a year for most people, but there are worrying signs (low antibody production, reports of reinfection that might or might not be testing artifacts). We just don't know.
To me, it was a gamble, and we'll know before long whether it pays off. There were some things that bothered me about Prof. Giesecke, such as his dismissal of the experience of China ("it's a different world"), and other things that made a lot of sense. People who want to believe will be citing this video as authoritative support for their beliefs. If these assumptions turn out not to be consistent with evidence, then people will be citing this video as a case study in how smart people can get stuff horribly wrong.
Every one of the “reports of reinfection” I have seen was along the lines of: tested positive continuously for 2 weeks, then tested negative twice, then tested positive again a couple days later. There is no evidence in any of the cases I saw that the person was exposed to another infected person during that time period.
This should not be called “reinfection” (even though there are a bunch of sloppy media headlines calling it that) but rather “poor test sensitivity during the last stages of the infection”.
About 1. not even counting death, we don't know what the long term effects on the lung capacity of the people who recovered are. Also, everyone reports "80% of mild cases", with a definition of mild that includes things that reasonable people would definitely not call mild. Essentially, as long as doctors don't tell you you have to be on O2, you have a mild case. People with mild cases may well end up with damaged lungs.
Yes, that is the main issue with our strategy in Sweden. We failed to keep it away from seniors' homes. The strategy seems to have worked just fine otherwise. It stopped our hospitals from getting overwhelmed, espcially since we managed to contain it in Stockholm for the most part. But we did not have resources to test people working with the elderly.
it's even worse than not testing enough in care homes: "Its advice to the care workers and nurses looking after older people such as Bondesson’s 69-year-old mother is that they should not wear protective masks or use other protective equipment unless they are dealing with a resident in the home they have reason to suspect is infected."
https://www.theguardian.com/world/2020/apr/19/anger-in-swede...
The responses in the US, UK and Sweden (and probably others) are downright criminal and should put politicians in jail.
https://unherd.com/thepost/coming-up-epidemiologist-prof-joh...