4 ms·

The idea is that in the long run everybody will get it, so might as well "take it on the chin" now to gain herd immunity, and minimize damage to the economy. Th

by knzhou 7y ago

The idea is that in the long run everybody will get it, so might as well "take it on the chin" now to gain herd immunity, and minimize damage to the economy. There's some rumblings of protecting the old during the summer when this is supposed to happen, though I'm not sure how that's supposed to work when there are millions of active cases at once. I suspect the real reason, as some government officials have outright said, is that it would free up money for the NHS by killing off "bed blockers". Their reasoning is that if the only things you can do are China-style lockdown or nothing, then the disease will just resume exponentially growing the second lockdown ends, so it's pointless to do it at all. While that may be technically right (under some optimistic, unverified assumptions, e.g. no reinfection), I think it's a false dichotomy. Singapore, Hong Kong, and South Korea aren't under lockdown, they're stopping the spread by just having good testing and case tracking capacity, and good hygiene. And the longer things go on, the better they're getting at doing that. If things are getting out of hand, a temporary lockdown is effective (as it was in China), to get the numbers down while building up this kind of capacity. But the UK seems to have just dismissed this option out of hand... which means they seem to be heading blindly to disaster.

dnautics 7y ago
> While [all or nothing] is technically right No, it's not. Have you done the math?
- knzhou 7y ago
  Well, I've done the same back of the envelope estimates they're claiming, but what do you have in mind? If you're saying that their plan is doomed and going to kill millions, I agree.
  - dnautics 7y ago
    I don't think your claim that you need to go all Hubei to slow or curtail the spread of an epidemic is correct. "Herd immunity" is a very well studied concept, but it depends on lots of parameters, like transmission likelihood, social exchange rate, transmissibility window, and we (as in humanity), much less you, don't have a very good idea of what those numbers are (and honestly, orgs like WHO and CDC don't either). Depending on what those numbers are, it could (making some numbers up here) be that a 20% decrease in social contact is sufficient (and the US was never doomed from the start relative to china) or maybe a 40% decrease is sufficient (and the US is probably fine now) or maybe a 90% decrease is necessary (and indeed going all Hubei is required). But given that there haven't been major outbreaks in "china outside of Hubei", I think we have good reason to suspect that maybe full authoritarian clampdown isn't necessary, and voluntary self-quarantine is likely to be effective. But who knows? Don't take my word for it. I'm just a no-longer-in-the-industry chemistry PhD that has studied some autocatalysis differential equations. It's also possible that there are genetic components to these parameters, so what works (or doesn't) in italy might not be the same in thailand.
dvdhnt 7y ago
Very informative. Thank you
mytailorisrich 7y ago
The problem, if you do the maths, is that this must be done over several years and, thus, the same drastic measures as other countries must be taken now. Having 40+ million cases in the UK this year alone does not seem even remotely manageable for the health services. I don't even see extra hospitals being built or anything at the moment, just threats to schools if they close.
- knzhou 7y ago
  That's exactly right. The NHS is going to completely collapse before their "plan" is even 1% finished. I suspect that once this happens, they're just going to fold and adopt the same plan as every other country -- but after thousands of unnecessary deaths.
melling 7y ago
Strange, I asked about this a few days ago and people thought it was a bad idea: https://news.ycombinator.com/item?id=22542494 https://news.ycombinator.com/item?id=22542494 “ If you’re under 30 without any other health issues, you should be fine? Might not realize you have it? I wonder if it would be easier to create a low dosage vaccine for younger people, thus creating a bit of a firewall so it doesn’t spread as rapidly. I know nothing about medicine so this may be silly.”
- Johnny555 7y ago
  Maybe people don't understand what a "low dosage vaccine" means? If any vaccine at all can be created in time to help, why not have it be a "full dosage" vaccine and give it to everyone? But starting with young healthy people seems counterproductive, start with those at high risk of catching it (front line healthcare workers) and those at high risk if dying from it.
masklinn 7y ago
TBF it's probably too late for most countries to stop transmission entirely (SK certainly doesn't though that's in part due to Patient 31). However that doesn't mean you should let the virus run unchecked, because doing that overwhelms the healthcare system which seems to be what ended up happening in Italy: there are too many case requiring intensive care (especially as the system is not really prepared yet), which means not everybody can be treated properly, which means more people die that necessary, from both coronavirus and other IC-requiring issues (e.g. strokes or accidents and whatnot). The nominal death rate of covid-19 seems to be about 0.5% at worst, possibly lower because lots of asymptomatic cases (so the numbers we have right now are likely over-estimates). By comparison the flu is about 0.1% (thought what's also unclear is how much of the population it can spread to before it basically burns itself out, IIRC flu is about 10% but for covid I've seen speculations from 20 to 70%). However if intensive care gets overloaded, it shoots up to 5~7%, possibly even higher. That's a lot of extra and unnecessary deaths. That's going from a single "unavoidable" death to 10 or 15. That's what "flattening the curve" is about, all the social distancing, hygiene, etc…[0] it's about slowing down the spread from its (again nominal) 33% per day or something, to give the healthcare system the time to ramp up to "full capacity" (and habituation) and keep the number of new cases below its capacity. That, incidentally, is also why universal testing is not that useful: people with fragilities or comorbidities and medical personnel need to be checked quickly because that impacts either their care or whether / how they can care for others. For the vast majority, if you think you might have COVID you stay home so you're not a spreader, which is what you'd do if you were a confirmed case. That's about it. Sure knowing would be nice, but even South Korea simply can not test everybody, they can test 10000 people a day out of a population of 51 million, this means it'd take them 14 years to test everybody. That's obviously not useful. It's also why you should absolutely not call (to say nothing of going to) ER for covid. Hopefully your healthcare system has a contact point for covid, look up that number and if you think you need to call these people, not 911 or 112. Also note: outside is safe, you can go out and watch trees or run or bike. In fact I really should. You should just avoid direct contact, stay out of droplets (coughing) range, try to not cough on other people (cough in your elbow or a hankie), wash your hands when you can (especially when you get home). Also treat all doors like they're the worst toilet doors you've ever seen e.g. if you can open them without your hands do so, if you can't then try very hard not to touch yourself until you've sanitised or washed your hands. [0] also masks if the supply is good enough which is not the case in most countries, which is why it's recommended to leave them to medical professionals who can't just stay home because they have to go out and treat cases

by knzhou 7y ago

dnautics 7y ago

> While [all or nothing] is technically right No, it's not. Have you done the math?

knzhou 7y ago
Well, I've done the same back of the envelope estimates they're claiming, but what do you have in mind? If you're saying that their plan is doomed and going to kill millions, I agree.
- dnautics 7y ago
  I don't think your claim that you need to go all Hubei to slow or curtail the spread of an epidemic is correct. "Herd immunity" is a very well studied concept, but it depends on lots of parameters, like transmission likelihood, social exchange rate, transmissibility window, and we (as in humanity), much less you, don't have a very good idea of what those numbers are (and honestly, orgs like WHO and CDC don't either). Depending on what those numbers are, it could (making some numbers up here) be that a 20% decrease in social contact is sufficient (and the US was never doomed from the start relative to china) or maybe a 40% decrease is sufficient (and the US is probably fine now) or maybe a 90% decrease is necessary (and indeed going all Hubei is required). But given that there haven't been major outbreaks in "china outside of Hubei", I think we have good reason to suspect that maybe full authoritarian clampdown isn't necessary, and voluntary self-quarantine is likely to be effective. But who knows? Don't take my word for it. I'm just a no-longer-in-the-industry chemistry PhD that has studied some autocatalysis differential equations. It's also possible that there are genetic components to these parameters, so what works (or doesn't) in italy might not be the same in thailand.

dvdhnt 7y ago

Very informative. Thank you

mytailorisrich 7y ago

The problem, if you do the maths, is that this must be done over several years and, thus, the same drastic measures as other countries must be taken now. Having 40+ million cases in the UK this year alone does not seem even remotely manageable for the health services. I don't even see extra hospitals being built or anything at the moment, just threats to schools if they close.

knzhou 7y ago
That's exactly right. The NHS is going to completely collapse before their "plan" is even 1% finished. I suspect that once this happens, they're just going to fold and adopt the same plan as every other country -- but after thousands of unnecessary deaths.

melling 7y ago

Strange, I asked about this a few days ago and people thought it was a bad idea: https://news.ycombinator.com/item?id=22542494 https://news.ycombinator.com/item?id=22542494 “ If you’re under 30 without any other health issues, you should be fine? Might not realize you have it? I wonder if it would be easier to create a low dosage vaccine for younger people, thus creating a bit of a firewall so it doesn’t spread as rapidly. I know nothing about medicine so this may be silly.”

Johnny555 7y ago
Maybe people don't understand what a "low dosage vaccine" means? If any vaccine at all can be created in time to help, why not have it be a "full dosage" vaccine and give it to everyone? But starting with young healthy people seems counterproductive, start with those at high risk of catching it (front line healthcare workers) and those at high risk if dying from it.

masklinn 7y ago

TBF it's probably too late for most countries to stop transmission entirely (SK certainly doesn't though that's in part due to Patient 31). However that doesn't mean you should let the virus run unchecked, because doing that overwhelms the healthcare system which seems to be what ended up happening in Italy: there are too many case requiring intensive care (especially as the system is not really prepared yet), which means not everybody can be treated properly, which means more people die that necessary, from both coronavirus and other IC-requiring issues (e.g. strokes or accidents and whatnot). The nominal death rate of covid-19 seems to be about 0.5% at worst, possibly lower because lots of asymptomatic cases (so the numbers we have right now are likely over-estimates). By comparison the flu is about 0.1% (thought what's also unclear is how much of the population it can spread to before it basically burns itself out, IIRC flu is about 10% but for covid I've seen speculations from 20 to 70%). However if intensive care gets overloaded, it shoots up to 5~7%, possibly even higher. That's a lot of extra and unnecessary deaths. That's going from a single "unavoidable" death to 10 or 15. That's what "flattening the curve" is about, all the social distancing, hygiene, etc…[0] it's about slowing down the spread from its (again nominal) 33% per day or something, to give the healthcare system the time to ramp up to "full capacity" (and habituation) and keep the number of new cases below its capacity. That, incidentally, is also why universal testing is not that useful: people with fragilities or comorbidities and medical personnel need to be checked quickly because that impacts either their care or whether / how they can care for others. For the vast majority, if you think you might have COVID you stay home so you're not a spreader, which is what you'd do if you were a confirmed case. That's about it. Sure knowing would be nice, but even South Korea simply can not test everybody, they can test 10000 people a day out of a population of 51 million, this means it'd take them 14 years to test everybody. That's obviously not useful. It's also why you should absolutely not call (to say nothing of going to) ER for covid. Hopefully your healthcare system has a contact point for covid, look up that number and if you think you need to call these people, not 911 or 112. Also note: outside is safe, you can go out and watch trees or run or bike. In fact I really should. You should just avoid direct contact, stay out of droplets (coughing) range, try to not cough on other people (cough in your elbow or a hankie), wash your hands when you can (especially when you get home). Also treat all doors like they're the worst toilet doors you've ever seen e.g. if you can open them without your hands do so, if you can't then try very hard not to touch yourself until you've sanitised or washed your hands. [0] also masks if the supply is good enough which is not the case in most countries, which is why it's recommended to leave them to medical professionals who can't just stay home because they have to go out and treat cases