minten

It seems a bit harsh to constantly review Sweden on the number of deaths compared to their neighbours as their policy was never aimed at less corona death in the first place and they were upfront about this. If someone wants to compare, you have to look at the broader picture (economic growth, quality of life), which you cannot do until much later. Despite what many predicted, their IC capacity seems to be holding.

Why are some of you people so obsessed with being proven right (or: making sure we all know somebody else was wrong)? It's so painfully obvious none of us were right, not even remotely, and nobody was "less wrong" than anybody else. As data comes in the truth appears to be more and more somewhere in the middle between what we were all saying. Can't we just accept that and move on by posting interesting articles etc... I mean, do whatever you want, but it's so stupid.

Preliminary NY sttudy says 21.2% of NYC has antibodies.

We all know we won't care because it won't be us. It's a point i've been trying to make since the start.

I don't want to come across as cynical or anything (and I'm definitely not making any immoral suggestions), but if you look at the profile of that 1% I'm not so sure that would be bad for the economy as a whole at all.

https://www.medrxiv.org/content/10.1101/2020.04.14.20062463v1 another antibody test study: " Under the three scenarios for test performance characteristics, the population prevalence of COVID-19 in Santa Clara ranged from 2.49% (95CI 1.80-3.17%) to 4.16% (2.58-5.70%). "

Looks promising. No control arm in the study is a caveat. Market seems to absolutely love this. GILD up 15% in after hours, SPY up 3.5%. Wow.

Is 3% much above the false positive rate for these anti-body tests? The researchers claim the anti-body test is not accurate enough to tell a certain person on an individual basis whether or not they have antibodies because of the false positives, but by looking at an entire group the results are reliable for that group. I can only assume that means they are correcting for the false positive rate. This is a study by a research institute. We'd have to wait for the final results what they did exactly, but doesn't sound like this was a study done by some group of overenthusiastic amateurs who were not aware of this.

There's definitely a time and place for questioning (and even for fund cutting if need be), I totally agree; But surely we can also all agree that time is not now.

This is interesting. Netherlands used policies more or less intermediate compared to places like Austria on one side and Italy on the other with, as expected, more or less intermediate results. Netherlands has also produced interesting work on influenza vaccine effectiveness and is a good relative European student in terms of historical flu vaccinate rates. It looks like (picture developing) that the CV does behave (intrinsic features) similarly to the influenza virus, with the main difference being that the population tends to have a much lower natural immunity to it and there is no vaccine, not even a partially effective one. Reasonable extrapolation of data in the Netherlands suggests that the eventual death rate from CV (with some social distancing and other basic measures) will look like (compare to a reasonable degree) the typical death rate for influenza, had there been a 0% rate of vaccination. What society is doing is basically trying to adapt (with various levels of 'success') to this new and evolving reality. @LC Thank you for supplying the link for the European monitoring of mortality with seasonal variations. Apologies. To my own embarassment upon re-reading this I made a primary school calculation error here (my only defence: it was early). 3100 out of 500 000 obviously isn't 0.06%, but 0.6%, which makes quite a bit of difference here, Still not quite the 3% some are saying, but definitely not flu percentages either.

I found this interesting: Dutch bloodbank reports in preliminary discussion antibodies found in ~ 3% of 4208 blood donors they have tested. All blood donors should be symptom free and not have had contacts with anyone with symptoms for 14 days before they are allowed to donate. I'm sure you could argue the people who donate blood by definition are not a perfect representation of a population, but otherwise this seems pretty random to me (and probably underreporting as everyone tested is by definition asymptomatic). Netherlands has a 17 million population, with right now 28K registered cases and 3100 deaths. 3% of 17 million is 500K. So suggests real number of infected patients compared to patients tested positive could be 18x as large. 3100 deaths is surely underreported as it is everywhere (nursing home deaths etc), but let's make it 5000-10000, you get a mortality rate of 0.1%-0.2%. So, based on this report, herd immunity seems far away, except maybe in regions that have been hit particularly hard. On the other hand, death rate does seem much, much lower than initial stats (and some people) were saying . And number of asymptomatic patients are many times the number of patients that get sick (as more and more research is now suggesting).

No, they were appallingly bad, to the extent that I'd lose a some degree of respect for all doctors if it weren't for Dalal's sensible posts on this thread to pull me back off that ledge. Frankly, it terrifies me that doctors exist who ignore evidence for gut feel, make wild speculations not even supported by common sense let alone facts, and then when the evidence proves them wrong, continue to insist that they were right. (What the heck is one to do if one seriously needs a doctor, and this is the doctor one gets? Just roll over and die?) That seems a bit harsh to me. I've been in this thread a while now and I know for sure who I'd like my doctor to be (if anyone here is a doctor at all; or maybe we all are, who knows); anyway, I'm pretty sure we all have a conscious or subconscious preference to whatever or whomever fits our pre-existing bias best, so let's not get too personal. https://www.bloomberg.com/news/articles/2020-04-15/china-s-data-on-symptom-free-cases-reveals-most-never-get-sick. Coincedentally this source reports a similar percentage of asymptomatic patients as Boston's homeless study did (you know, the study nobody said was random). Some of you think it's all China's fault and all China does is lie, bla bla bla, but still.

And now this: "Whoa, 147 (36%) out of 408 people tested positive for the #coronavirus at a large homeless shelter in Boston https://medrxiv.org/content/10.1101/2020.04.12.20059618v1. More interestingly, only ~1/6 showed symptoms among those tested positive, i.e. 1:5 for symptomatic vs asymptomatic. #COVID19" So...um...from the BMJ study: Does this sound like a random sample? If I tested residents of a nursing home with a large cluster of covid cases, is that result going to give me the population incidence? Or is it going to vastly overestimate it? Uhm, absolutely nobody said this was a random sample. Even the tweet (which, for clarity, isn't mine) implictly says this by saying "More interestingly ...".

And now this: "Whoa, 147 (36%) out of 408 people tested positive for the #coronavirus at a large homeless shelter in Boston https://medrxiv.org/content/10.1101/2020.04.12.20059618v1. More interestingly, only ~1/6 showed symptoms among those tested positive, i.e. 1:5 for symptomatic vs asymptomatic. #COVID19"

In all the talk about the contact tracing apps nobody seems to want to get into how it will work in practice. If I go out running, and run past someone who is infected, how will the system evaluate risk. How will the system know if and in what way I interacted with this person? If you set the filter too loose, it won't work. If you set it too tight, your whole society will be in quarantine in days and everybody will de-install it within a week. Also, if I sit in locked up my apartment all day, and my infected neighbour will sit 1 meter away for hours in his apartment, but with a wall between us, will the app know there's a wall? I can think of countless more questions like this.