4 ms·

> (There are other ways we can then cross validate that.) Tell me more, I'm interested. > by positing that the virus became more deadly. I posit that inappro

> (There are other ways we can then cross validate that.) Tell me more, I'm interested. > by positing that the virus became more deadly. I posit that inappropriate treatments make the virus more deadly than it has to be.

Robin_Message 6y ago
Inappropriate (or rather, best-we-can-think-of-at-the-time-which-we-later-improved) treatment still saved many lives compared to the toughing it out at home approach. Doctors already know ventilation is bad for people; they were only doing it for people who would otherwise die. In the UK at least, we're seeing a fatality rate of perhaps 50% what it was in March due to some drug treatments and better use of ventilation (especially prone). But without treatment, the virus would be perhaps 2 or 3 times more deadly (as was seen in Wuhan when hospitals were overwhelmed).
jsnell 6y ago
Sure, here's a few ways to cross-validate: If we had a high prevalence of Covid before there was any testing, we'd expect that the moment testing started, we'd see a high prevalence everywhere. We didn't. Most places saw only a small proportion of tests come back positive to start with, which then increased over time. (When looking at this kind of testing that's been done using antibody tests, it's good to remember that even a 99% specific test will mostly be returning false positives when the prevalence is very low). We can re-test historical samples of various kinds for Covid. That has been done, and while they can find individual cases from a month or two before the first confirmed "live" case, there's no sign of it being widely spread. We can also deduce the transmission patterns of the virus from the phylogenetic tree. If the virus had been widely circulating outside of China in 2019, sequencing would have show variants that are common in the west branch off from the Chinese ones earlier than they did.

Robin_Message 6y ago

Inappropriate (or rather, best-we-can-think-of-at-the-time-which-we-later-improved) treatment still saved many lives compared to the toughing it out at home approach. Doctors already know ventilation is bad for people; they were only doing it for people who would otherwise die. In the UK at least, we're seeing a fatality rate of perhaps 50% what it was in March due to some drug treatments and better use of ventilation (especially prone). But without treatment, the virus would be perhaps 2 or 3 times more deadly (as was seen in Wuhan when hospitals were overwhelmed).

jsnell 6y ago

Sure, here's a few ways to cross-validate: If we had a high prevalence of Covid before there was any testing, we'd expect that the moment testing started, we'd see a high prevalence everywhere. We didn't. Most places saw only a small proportion of tests come back positive to start with, which then increased over time. (When looking at this kind of testing that's been done using antibody tests, it's good to remember that even a 99% specific test will mostly be returning false positives when the prevalence is very low). We can re-test historical samples of various kinds for Covid. That has been done, and while they can find individual cases from a month or two before the first confirmed "live" case, there's no sign of it being widely spread. We can also deduce the transmission patterns of the virus from the phylogenetic tree. If the virus had been widely circulating outside of China in 2019, sequencing would have show variants that are common in the west branch off from the Chinese ones earlier than they did.