4 ms·
We are freaking lucky that COVID’s death rate is by the lower end of all the estimates. If the actual rate was closer to what Italy was seeing, or even a smalle
by btilly 6y ago
We are freaking lucky that COVID’s death rate is by the lower end of all the estimates. If the actual rate was closer to what Italy was seeing, or even a smaller fraction of thst, Sweden would have had far more dead.
If you think that was luck, then you didn't understand the difference between CFR and IFR.
The Case Fatality Rate (CFR) is the fraction of reported cases that result in death. This is easier to measure. And if you're sick enough to become a reported case, it is the best back of the envelope estimate of how likely you are to die.
The Infection Fatality Rate (IFR) is the fraction of infected people that will die. For estimating how many will wind up dead if you hit herd immunity, that is the relevant figure. It is also harder to produce.
Early on we were seeing CFRs in the 5-10% range. That is the figure that you remember
But the first estimates of the IFR was an analysis of the Diamond Princess. It estimated an IFR of around 0.5%. In late March, Fauci reported that it was about 1%. Both estimates had significant error bars. The figures that Neil Ferguson used in his model was in the same range. All of this happened while we saw CFRs in the 5-10% range.
And therefore there should have been no surprise from the beginning that the CDC estimate of the IFR for the Spring would come out where it did, 0.65%.
- makomk 6y agoPart of the problem is that early on the US media actively misinformed people about this crucial point for partisan reasons, describing the naive CFR as the real WTO-confirmed fatality rate and lower figures as Trumpian lies created to downplay the severity of the disease.
- DanBC 6y ago> And therefore there should have been no surprise from the beginning that the CDC estimate of the IFR for the Spring would come out where it did, 0.65%. Where are you seeing the CDC say that please?
- scottlocklin 6y agoIt's actually considerably lower. 0-19 years: 0.00003 20-49 years: 0.0002 50-69 years: 0.005 70+ years: 0.054 https://www.cdc.gov/coronavirus/2019-ncov/hcp/planning-scenarios.html https://www.cdc.gov/coronavirus/2019-ncov/hcp/planning-scena...
- classichasclass 6y agoThat's a ratio. 0.054 = 5.4%, etc.
- btilly 6y agoNote that 70+ years has an IFR of 5.4%. https://www.census.gov/prod/cen2010/briefs/c2010br-09.pdf https://www.census.gov/prod/cen2010/briefs/c2010br-09.pdf says that in 2010 there were 308,745,538 Americans, 40,267,984 of which were over 65 and 12,435,263 in the 65-70 age bracket. That makes 70+ around 9% of the population, which means that expected deaths from people who are 70+ are an IFR for the general population of 0.49%. That's the bulk of the deaths. Add in the other age groups and, if you do the numbers correctly, you should get close to my figure. (With a small difference because you're quoting updated numbers and I'm quoting figures that are a few months old.)
- DanBC 6y agoFrom your link: > The parameters in the scenarios: > Are estimates intended to support public health preparedness and planning. > Are not predictions of the expected effects of COVID-19.
- btilly 6y agoI found that figure in https://www.statnews.com/2020/08/24/infection-fatality-rate-shows-covid-19-isnt-getting-less-deadly/ https://www.statnews.com/2020/08/24/infection-fatality-rate-... but I am not sure where they got it from.
- qqqwerty 6y agoDiamond Princess has a raw IFR of 1.8%. The studies you were likely referring to were all done in March when the total number of fatalities was 7. That number is now 13. Anyone who claimed to know what the fatality rate was back then is full of it. Even in places that had gotten the spread of the virus under control, there were still significant numbers of unresolved cases. And it will likely be another year or two before we have a good idea what IFR is, as COVID has demonstrated differing levels of severity in different populations, and adequate health care capacity has show to have a large impact, which large portions of the world do not have access to.
- btilly 6y agoThey were done in March. But furthermore they attempted to quantify based on the age breakdown. 0.5% was therefore the best available early estimate, with a large error bar.