Read on Twitter
Tony Fauci says the US death toll is ‘going to be very disturbing’, even when you consider the 120k+ lives lost already. He’s right. I would optimistically estimate that the final number will be closer to 1 million, if we are lucky 
https://abs.twimg.com/emoji/v2/... draggable="false" alt="🧵" title="Thread" aria-label="Emoji: Thread">1/n https://www.theguardian.com/world/2020/jun/30/coronavirus-fauci-senate-states-reverse-reopening">https://www.theguardian.com/world/202...
To get a hold on the problem all we need to do is these CDC stats. If 20M infections produce 120K deaths, then how many more would we expect in order to generate enough infection and immunity to slow the pandemic?
2/n https://www.npr.org/sections/coronavirus-live-updates/2020/06/25/883520249/cdc-at-least-20-million-americans-have-had-coronavirus-heres-who-s-at-highest-ri">https://www.npr.org/sections/...
2/n https://www.npr.org/sections/coronavirus-live-updates/2020/06/25/883520249/cdc-at-least-20-million-americans-have-had-coronavirus-heres-who-s-at-highest-ri">https://www.npr.org/sections/...
Well the ‘herd immunity; threshold beyond which immunity is expected to reduce transmission is about 60%. Let’s be generous and reduce it to 50%
OK (quick google) that would be about 170M people infected. Let& #39;s do some arithmetic 3/n
OK (quick google) that would be about 170M people infected. Let& #39;s do some arithmetic 3/n
If 20M infections have already led to 125K deaths, we expect 170M to result in more than one million deaths over the pandemic in the US. This is an optimistic assessment. Reasons and caveats follow 4/n
While this calculation is crude. For eg it takes no account of the fact that risk of death climbs spectacularly with age and comorbidities. Nowhere has insulated those groups at risk from groups not at risk. So that doesn’t work 5/n
And the pandemic so far has been shockingly concentrated in communities less able to protect themselves. But there are both many more such communities, and many others that have yet to realize the virus matters for them too 6/n
The numbers we need to get infected to obtain population immunity might be lower because of host heterogeneity in transmission rates (a reasonable argument). I hear you. I reduced the herd immunity threshold from 65% to 50% which is generous. Happy? 7/n
Moreover 50% herd immunity neglects to mention that this only marks the point that, in an unmitigated outbreak, transmission starts to decline due to population immunity. Transmission will not completely stop at this point. It will keep on, declining until it burns out 8/n
So Infection will not stop once transmission declines because of immunity. It will carry on. 50% of Americans infected is a minimum. I’m assuming the best immunity, and no ‘overshoot’. It’s spectacularly optimistic. The reality could be much larger. 9/n
And then we should not forget the part of the picture that includes milder manifestations with long term consequences. About 10% of younger people with symptoms will be living with this a long long time https://www.theguardian.com/commentisfree/2020/jun/28/coronavirus-long-haulers-infectious-disease-testing">https://www.theguardian.com/commentis... 10/n
How can I put this? Someone you know is likely to die from this (I can already think of multiple examples). Going forward, how do we stop more of that happening? Easy. Stop transmission. Stop blowing on the embers. It& #39;s epidemiology, not rocket science 11/end
