3 ms·

> And the paper is still much closer than the 1-3% fatality rates often touted at the time. The paper you're citing shows that at the time (mid-2020) that Ioan

> And the paper is still much closer than the 1-3% fatality rates often touted at the time. The paper you're citing shows that at the time (mid-2020) that Ioannidis wrote his paper estimating a 0.2% IFR, there were many regions in the world where a 1% fatality rate was well within the 95% confidence range. > Later research has put the IFR rate at 0.47% to 0.31% To be more precise, an IFR dropping from 0.47% in March 2020 to 0.31% in January 2021, mostly due to improved treatments. > [herd immunity would protect the population] is correct. Reinfections are less likely to result in severe symptoms. Yes, but this means you cannot claim 400 million infections as a denominator for a constant 0.2% IFR, because many of these are reinfections, so should have a lower IFR than whatever was estimated in mid-2020. As an aside, before Ioannidis started engaging in COVID-19 research, he estimated that the US would suffer 3.3 million infections and 10,000 deaths: https://www.statnews.com/2020/03/17/a-fiasco-in-the-making-as-the-coronavirus-pandemic-takes-hold-we-are-making-decisions-without-reliable-data/ https://www.statnews.com/2020/03/17/a-fiasco-in-the-making-a... His subsequent research seems to have been designed to justify the policy preference, rather than the other way around.

Manuel_D 4y ago
> Yes, but this means you cannot claim 400 million infections as a denominator for a constant 0.2% IFR, because many of these are reinfections, so should have a lower IFR than whatever was estimated in mid-2020. So you agree with him: as herd immunity builds, the IFR goes down as people get more resistant. In fact you pointed this out yourself in a previous comment. > And the whole assumption of the Great Barrington declaration, of which Bhattacharya was an author and Ioannidis a sympathizer, was that the rate would drop considerably on reinfection, because herd immunity would protect the population. And this is exactly what happened. COVID became less serious as the pandemic wore on, and dropped close to his estimate even before the release of the vaccines. Infection fatality rates were vastly overestimated. I got lots of flak on HN for pointing to the German random testing data estimating an IFR of 0.37%. people were calling me a moron telling me this was an order of magnitude off. Jay's claim was less conservative, but still much closer to reality than the consensus at the time. Heck, even if you use the figure of 0.47%, Jay was still closer than most expert opinions.

Manuel_D 4y ago

> Yes, but this means you cannot claim 400 million infections as a denominator for a constant 0.2% IFR, because many of these are reinfections, so should have a lower IFR than whatever was estimated in mid-2020. So you agree with him: as herd immunity builds, the IFR goes down as people get more resistant. In fact you pointed this out yourself in a previous comment. > And the whole assumption of the Great Barrington declaration, of which Bhattacharya was an author and Ioannidis a sympathizer, was that the rate would drop considerably on reinfection, because herd immunity would protect the population. And this is exactly what happened. COVID became less serious as the pandemic wore on, and dropped close to his estimate even before the release of the vaccines. Infection fatality rates were vastly overestimated. I got lots of flak on HN for pointing to the German random testing data estimating an IFR of 0.37%. people were calling me a moron telling me this was an order of magnitude off. Jay's claim was less conservative, but still much closer to reality than the consensus at the time. Heck, even if you use the figure of 0.47%, Jay was still closer than most expert opinions.