I don't bring this up. TS does, over and over. I don't care one way or the other. As far as the IFR, present a study that has been published in a medical journal to support your claim of it being greater than 1 percent. I looked for one, and I posted the one I found. if there's a better one, please find it.

Or, like I suggested, go on twitter and tell the authors of the study that was published three days ago that you think their findings are wrong and that there is irrefutable evidence that the IFR is greater than 1 percent.

In fact, I dare you to do this.

Post their twitter handles please.

Do you really want someone to ask:

“Dear COVIDBROS, can you help me? When you write “Conclusions: Based on a systematic review and meta-analysis of published evidence on COVID-19 until the end of April, 2020” do you mean it’s possible for data to come out in May that can provide more accurate projections? Or do you really mean all future data is invalid until you personally publish a new study saying so?”

How much of a fool do you want him to look like?
What exactly would someone even ask them?

Fyi I’m all for proving TS wrong, especially on this important statistic, but you haven’t displayed any rational thought here.



P.s. it wasn’t published 3 days ago. Just updated. The original came out may 5, which makes sense because it doesn’t include any may data. I don’t know what they updated on the 18th but it was minor, certainly not adding additional data points from may.

Again, if you can find a better study with more recent data, have at it.

As far as twitter, he is more than welcome to ask them anything he wants. But the fact is that data from May is just going to be outdated by June, and that data will be outdated by July.

It's why you can't make any concrete conclusions about IFR right now, or anytime in the future.

TS may be right or he may be wrong. No one knows. That's the point I am trying to make.

But, like I said, I don't care one way or the other. And I am done posting on it. I will ignore all future takes on this and just wait for reliable figure, which should be available in two to three years.

https://papers.ssrn.com/sol3/papers....act_id=3590771
https://www.healthaffairs.org/doi/fu...aff.2020.00455

You can find more.

There are a lot more studies, basically all "preprint" but still more authoritative than forum posters (I think this is obvious). The TLDR is if you ignore unknown asymptomatics, people are getting to just over 1%. There are a lot of methodologies for estimating undetected asymptomatics that really just amount to educated guesses at this point. But they basically all agree there are definitely a lot of undetected asymptomatic out there. Just how much "a lot" is in question.

Bottomline is this, if Chinese, South Korean, and pretty much data everywhere else, suggesting that many people are asymptomatic and undetected, can be relied upon, then it's almost certain the true IFR is sub 1%, and very possibly under 0.6% (where SK is now I think).

That's still terrible for a disease increasingly likely to sweep the entire human population, making the debate of over or below 1% pretty much an academic one.

17% of people in London and 5% in England have had COVID
- British Health Secretary

Ugh I dunno. Watching tooth get schooled here was like watching James Bond die.

Rattles you a bit.

‘How Could the CDC Make That Mistake?’
The government’s disease-fighting agency is conflating viral and antibody tests, compromising a few crucial metrics that governors depend on to reopen their economies. Pennsylvania, Georgia, Texas, and other states are doing the same.

https://twitter.com/alexismadrigal/s...688215552?s=20

https://amp.theatlantic.com/amp/article/611935/

So I went and got an antibody test to see if I had covid (they did a blood draw...needle in the arm, administered by quest), and the result came back negative for antibodies.

I'm super confused by this because to the best of my knowledge the flu only lasts for 1-2 weeks, and I was sick for a SOLID 5.5 weeks earlier this year. I coughed so much I coughed up blood, I spent 3 days in bed, I had a fever, trouble breathing, and my daughter's best friend's family just got back from china and was sick.

But I guess I didn't have covid, wtf? I was at least hoping I had it and would be immune from here on out so I could get "semi" back to normal, but with the way I'm running I'll probably catch covid now too...fml.

When you posted your symptoms the odds were near zero you could have caught it. There weren't enough infected.

Your experience makes me wonder how reliable the antibody tests are and even showing >1% IFR, if they're not still underestimating. I realize they're supposed to be random but there's still a selection bias. If they choose random people but they have to say yes, people who's had flu like symptoms will say it more than people who's never had any and therefore cannot have it. And even without this bias it's sampling people more frequently out in public. For example in New York where they asked for shoppers to test, if you go out shopping 10 times a weeks, vs a cautious person one time a week, you'll be 10x more likely to be picked and much more likely to have gotten corona (because you're out and about a lot). Add that to the bias of saying yes. Where it's based on self-choice drive up testing sites this is even worse. If it's based on lab-taken blood, this selects for people in medical environments/hospitals where covid tends to spread. No way blood donors aren't more likely to get it for example.

I came across that meta-study that jsb posted that estimates the IFR at 0.74% (but didn't post it because it seemed too flimsy) because its author shared it as a counterpoint to a new study from the Stanford professor who was behind the Santa Clara study, and who continues beating the drum that the true IFR is 0.1% or some such preposterous number. It's kind of funny that jsb got ridiculed for posting the Stanford guy's study, and then—for the same reason—is getting ridiculed for posting a study from a guy refuting the Stanford study.

Link to antibody data on South Korea?

Here's Iceland's data. Their disease burden isn't really from young travelers. The median age of infected is well over the median age of the country. It's convenient to ignore the Iceland data when there's so much other data pointing to a higher death rate, but I think that's a mistake. In Iceland, 869 people between 40 and 69 had the virus, including 212 over age 60, and only two people died.

What's annoying about these IFR discussions is that it's super difficult to prove anyone wrong unless they're off by an order of magnitude. If jsb wanted to be a dick, he could continue trotting out Iceland, or Qatar, and proclaim IFR is well under 1%. It obviously changes by population, but what I don't see mentioned hardly ever here is that it changes over time. Not just in terms of who has died so far—have younger people been disproportionately infected, and seniors are an IFR time bomb, or is NYC's IFR so high because Cuomo insisted COVID-positive seniors be sent back into nursing homes and dying seniors skewed the IFR upward—but in the fact that treatment gets better over time. A huge number of deaths are from when we were venting patients right away, thinking this was all about ARDS, but now that has shifted dramatically and we're treating the systemic endothelial damage much better. It wouldn't surprise me if the IFR were cut in half just from the change in treatment protocols over the past month or two. That was much of why when TS and I talked about the IFR several pages ago, I said I would take the under on 1% by the end of the year. So I think discussions on a precise IFR even for a particular country or time period are pretty moot. The value in IFR estimates is knowing whether herd immunity is viable, or whether a vaccine is worth administering given its side effects. Even if you think IFR is 0.5%, which may be true in Iceland, going back to a pre-COVID lifestyle with no contact tracing or mask wearing results in too many deaths.

Yeah man, 1800 people infected in a young country - most of them healthy travelers before it was nipped in the bud - disprove >200,000 data points.

You're not being serious. IFR is 0.5% even in this population, and a single nursing home infection would have made it 1.5%. So jsb could trot out Iceland, but he'd be correctly mocked there as well.

Cuomo's nursing homes don't matter because they're a small percentage of all very-high-death old people. They're more than offset in an IFR calculation by the fact that old people in population level samples have 1/3 the antibodies of the middle aged.

the data is completely overwhelming that the IFR is greater than 1%. It's amazing to see people argue it, actually.

So that another 1% IFR data point (without hospital overwhelming and with death undercounting).

The antibody tests are population level samples. Your only out is that the infected asymptomatic aren't making antibodies, which the data proves false.

Most likely is that the "asymptomatic" are false positives from non-infected very low doses (in the nose but not the lungs, could easily catch it with proper dosing) or actual false positives from the tests (some earlier tests had terrible reliability).
South Korea is at 2.37%, and they aggressively contact traced and tested everyone they found infected, early and fast (before the first death and sufficient to halt the infection), so the odds that they caught the bulk are very high. Where on earth are you getting 0.6%? South Korea only just announced antibody testing will be undertaken and I'm unaware of any results yet.

That is funny, and it’s also a good thing - shows a consistent interpretation of data.

That study gave a range- .75% was the high end. I’d say that’s right around the low range of the true IFR given current data and projecting optimistically that improved treatment this year will significantly improve the IFR trajectory.


Most recently, the bigger issue was his blind rejection of May data because the most recent study he could find on google didn’t reference it. Just a bad look.

Approximately 1.5%?

England (approximate):
36,000,000 people
5% infected = 1,800,000 people
27,000 deaths
IFR = 1.5%

Assuming the antibody tests accurately represent actual infections.

I didn't say the Iceland data disproved any other data. I didn't argue the IFR was 0.5% anywhere other than Iceland. I was responding to Kelhus who brought up Iceland, and I said that the Iceland data counts. If you want to weight it in your IFR calcs as 1,800/200,000, fine, but it isn't 0/200,000. And my point to Kelhus was that it's not a sample indicative of young travelers. The median age of those infect is 42. That's a higher median age than the US population. If you want to say that the Iceland data is inapplicable to the US or to France (whose median age is 41.4) because Iceland just got lucky with healthy infectees, or a low obesity rate, or that a diet of puffins and a culture of powerlifting lowers IFR, fair enough. But the young traveler theory doesn't match with a median infectee age of 42.

And I only mentioned Iceland in response to Kelhus. I think it's a very small data point, as I've said before.


Most of my post was conveying that IFR is variable not just across populations but across time. For some reason people seem to intuit this for R0, but obsess over a few percentage points for IFR. The IFR in San Francisco is lower than it is in New Orleans (people are healthier and hospitals are better). And the IFR in both places is lower now than it was two months ago (because treatment has improved).

population is closer to 55 million

Glaring error. Thanks.

Population of England: 55 million people
5% infected = 2.75 million people (according to UK Health Secretary)
27,000 deaths
IFR = 0.98%

(not confident the # of deaths is right or up to date)

We have had a care home nightmare in the UK and its the UK, home of the omnishambles.

We will trend higher than international averages on IFR, well at least against other developed countries.

Not just care homes, 10%+ of the deaths are people who caught it in hospital. They will have also have a higher IFR than the general population.

Regarding Moderna, Fauci - a vaccine expert if there ever was one - has exactly the same take as me:
That's the reason the line moved from 70% to 90% on all vaccine candidates on this Moderna news: because it crosses an important hurdle of proving a reliable human immune response without major side effects. It's not about Moderna, it's the proof of concept that matters.

How are self-claimed experts in this thread getting it wrong while I (a self proclaimed "know jack **** about vaccines") am getting it right? Just amazing.

Cautiously optimistic (Fauci) and 90% of the way to a vaccine (you) are not the same thing.

I'm not an expert - but I did ask two experts and they both said your take was way off.

EDIT: your earlier post says Moderna specifically was 90% of the way there. Now you're dialing back to ALL vaccine candidates.

In addition to the UK government, the NHS will have some tough questions to answer at the end of this.

The proof of concept is for their RNA platform only.

Fauci clearly doesn't think 90%, but I wasn't quoting him for that, and others do. Pre Moderna data, The Oxford group lead (a vaccine expert if there ever was one) thinks 80% chance of their own single vaccine proven working by September and in production. If we assume she's not lying/crazy with 80% for a single specific vaccine proven and well into production by September, then 90% for all 100+ vaccine candidates by November is standard.

Do these experts also believe the head of the Oxford Group is way off? Experts are slow and cautious and only good for analyzing stuff that fits into previous molds of accepted belief and outcomes. There are 100+ groups being funded billions of dollars working on this, there are solid proofs of efficacy/low side effects in related vaccines (MERS), and now we have the full stack of proof for this one (monkey prevention of infection, reliable human antibody creation with apparent low side effects).
The post doesn't say that at all. The evidence is stacking up that we can have a reliable, low side effects, antibody producing immune response from covid signature proteins. How they are delivered is irrelevant, one of 100+ ways will get there.