Hacker News Reader

Comment by dirkt | Hacker News Reader

Until a vaccine is found, 60-70% of the population will need to get infected for "herd immunity". While at the same time making sure infection doesn't grow exponentially, and overwhelms the health system.

Loosening the lockdown, and being prepared to row back, will achieve that if done carefully, while at the same time getting the economy slowly back to normal.

And that probably are a lot more people are already infected without symptoms will help with that.

Trying to somehow contain the infection won't work, because it will just get carried in from the outside again, eventually.

View on HN

Unfortunately herd immunity does not work. Germany has ~83m pple x 70% = 58m need to be infected for herd immunity The ages 65+ you want to protect and don't infect ~17m That reduces the people which naturally get herd immunity to (83m- 17m) x 70% = 46m

How do you want to infect the remaining 12m pple? By force? If you go by likelyhood of death by age, basically everyone below the age of 45 has to be infected to protect the elderly.

We can roughly estimate how long that will take if we stay at R=1. That's maybe 100k new infections monthly in germany. Meaning it takes ~40 years to reach 60%. (83000k*0.6/100k/12)

Maybe it's less if you can get more aggressive with more immunity, but even if I'm wrong by a factor of 10 it's still years and the vaccine is the more realistic bet.

We can roughly estimate precisely nothing until we know how many people have been infected, how many are immune and how long immunity lasts. We currently know none of those things.

I took the best case scenario: all tested become immune for ever. If you want add a 3x factor for unreported cases and you're still at 10 years waiting time.

Estimates make sense, especially if they proof your point even if you are off by a factor of 10.

One way to shorten the time is to constantly increase hospital capacity, so that in each month lets you infect more people. Especially if plasma donations turn out to work to reduce severe cases.

Another way is to separate young people from old people more effectively. Then you might be able to permit 1m young people to be infected each month with the same hospital budget we have today

That's assuming that the only two outcomes of a SARS-2 infection are "dead" and "completely recovered". There are case reports of young and otherwise healthy adults who survived SARS-2, but where a CT shows the lung tissue mostly obliterated. Sure, they survive, but running a marathon or going on that diving vacation might be a sudden death sentence for them.

I hope we can get actual studies soon because the anecdata is scary to me. I hope these stories only make news because they are rare not just because it happens to be a celebrity:

https://www.nbcnewyork.com/entertainment/entertainment-news/...

And that is also one possible way into desaster. By the time anyone gets close to herd immunity, the virus has mutated over and over again.

60% to 70% refers to an R0 of about 3. But the R0 is more likely to be around 6. That means you need 5/6th of a population to be immune - that's about 80% to 90%.

UPDATE:

"Results show that the doubling time early in the epidemic in Wuhan was 2.3–3.3 days. Assuming a serial interval of 6–9 days, we calculated a median R0 value of 5.7"

https://wwwnc.cdc.gov/eid/article/26/7/20-0282_article

A herd immunity strategy involving fast immunization by continuing business as usual would then involve an R0 > 5.

You're making up numbers. We don't know the R0 and likely won't for some time.

"Results show that the doubling time early in the epidemic in Wuhan was 2.3–3.3 days. Assuming a serial interval of 6–9 days, we calculated a median R0 value of 5.7"

https://wwwnc.cdc.gov/eid/article/26/7/20-0282_article

A herd immunity strategy involving fast immunization by continuing business as usual would then involve an R0 > 5.

There have been several studies that have come up with different numbers. I guess we can calculate some interesting stuff based on the range, but it's a really big range (I've seen somewhere in the range of 1.5 to 7 in various articles/studies).

R0 is environment dependent. It is probably in the 2-3 range for say California using hospitalization data.

There is not going to be a vaccine and there is not going to be herd immunity. You can't count on either of those outcomes any more than you can count on there being a vaccine and herd immunity for the common cold.

However you can make it mandatory for anyone with cold/flu symptoms to self-isolate rather than do the manly thing of bring it into an office and work through it. You can also make it so that people have to go into quarantine after international travel. You can also do some testing.

Rabies has been kept outside of the UK for decades by the quarantine approach. Pets can go one way from the UK to the rest of the world quite freely but on the way back they have to be quarantined. With rabies free Ireland an exception can be made.

With this approach we can avoid the indefinite lockdown required until the holy vaccine is found.

I think a vaccine is quite likely to be developed within the next couple years, the virus isn't as complex or difficult as influenza or hiv to target and we have some well defined targets to go after, as well as numerous vaccines already in clinical trials.

Sorry, I have no knowledge on virology topics. But this “herd immunity” thing is a complete nonsense. Politicians may repeat it as long as they want, but it still makes no sense. 50 million people must be infected, 1,5 million will die. It will take many years. While immunity very probably(????) disappear after months. Asian countries had many different viruses in the past, all of the contained. Why is it so difficult to follow their path?

Asian countries have not contained all viral outbreaks in the past. In both 1957 and 1968, Asian-origin flu epidemics killed millions. It's possible that containment could have been achieved if every country had swiftly followed Korea's example in early February, but there's no precedent for containing a human-to-human transmissible disease with millions of cases across the world without a vaccine.

Thank you very much. I am too young to remember these events. But I am also confident, that current technology is better than 50 years ago.

How do you know it's complete nonsense if you have no knowledge on the subject?

How do you start? Write invitation letters with date and time to first batch to be infected? Nonsense it is because it’s barely doable. The people will simply leave country unless one puts military on the borders.

Well, unless you have a vaccine, the virus does all of the work for you. Also, you can't really leave a country now anyway, and even if you could, the virus would be at the destination.

The herd immunity is not really something that you do but more of an unavoidable side effect. I'm not sure that we have much of a choice now that we are well beyond the point of containment.

I think the analogy says more about the person using it than the actual topic. If it was appropriate to compare people with... well... cows? Why not slaughter the lot and start over? Or take the stupid analogy to the next level and talk crop engineering? There! The solution has been staring us in the face all along.... or wait...

That's just the terminology used in epidemiology. Don't read anything into it.

Oh but I do! Group immunity or collective immunity will do just fine. I really don't care how popular this zoology terminology is. For me taxonomy works like: A pack of wolves, a flock of birds and a herd of cattle.

Beware when they start calling you [say] a consumer.

Herd immunity is a real phenomenon -- it's the strategic goal behind widespread vaccination, and the reason that vaccination controls disease even if less than 100% of the population takes the vaccine or gets immunity from it.

That said, I think you're right to be skeptical of the "let everyone get it and we'll have natural herd immunity" idea.

You don't know how many people are currently immune. Or how immunity even works for this virus. No one does. Zika was rendered mostly harmless by herd immunity before experts even knew what happened.

I don't see what is gained by making armchair proclamations as if you know something the rest of us don't. You don't even seem to know what you don't know.

The death rate is certainly lower than 3%.

Worldwide at this moment total (cumulative) deaths over total (cumulative) cases is nearly 7%, with considerably over half the total cumulative cases still active. Granted that we know that there are many undiagnosed cases, but we don't have a clear picture how many. There's no basis right now for your assertion.

https://www.worldometers.info/coronavirus/

The basis is studies of closed communities. Here in Germany there was a study of a whole village (testing everyone), that found a death rate of 0.37%. Other interesting data sets may be Diamond Princess cruise ship, Theodore Roosevelt carrier, or Health Care Workers.

So saying there is "no basis" is a bit rich. There may still be uncertainty.

Edit: this (just googled) article cites Diamong Princess to lead to an estimate of 0.5% death rate https://www.sciencenews.org/article/coronavirus-outbreak-dia... (the percentage of deaths on the DP was higher, but the passengers were also comparatively old on average).

Edit 2: lol for the downvote. What the fuck?

Edit 3: about the study in Germany https://www.theblaze.com/news/german-study-shows-coronavirus... (don't know "The Blaze", but if you care, you can google for better sources - google for "Streeck").

Edit 4: Theodore Roosevelt carrier https://en.wikipedia.org/wiki/2020_coronavirus_pandemic_on_U...

There are a lot of people on HN (and the internet at large) who view any statement adjusting-downward death rates as "COVID denialism", and will reflexively downvote attempts to correct in that direction.

In the most charitable reading, you could view them akin to the CDC and masks, trying to keep the public slightly misinformed for the greater good (preserve mask supplies for hospitals). In the least charitable reading, there are a lot of depressed and angry people on the internet, who want to believe that this disease is truly civilization-ending, and attack any opposing narrative.

Either way, I would not take it personally, or as a judgement on your actual quoted facts.

Another interpretation is that a lot of people on the internet (HN, reddit) have actually seen net improvements to their life from the lockdown.

For the introverted tech worker that prefers to work from home, the quarantine hasn't really negatively impacted their routine that much, and it might have actually lifted any peer pressure to go out and socialize with friends/co-workers/family.

For someone in that situation, there's not much of a benefit to advocate for reopening, and the deadlier the virus seems, the longer the lockdown will continue.

I really hope you're wrong, because that would be implying that HNers are selfish assholes who don't care a dime about the rest of their community.

While I personally am comfortable working from home and my company is doing well, I'm sad that I couldn't visit my parents and my grandmother for Easter. And it particularly frustrates me that my grandmother has to sit at home and practice social distancing, where she cannot do anything much besides sitting in her chair and watching TV all day. (She doesn't have or want a smart device, and her chronic back pain keeps her from engaging in most other home activities.) And that's before I consider the people around me whose lifelihoods have been impacted or obliterated by the shutdown, or the essential workers who often have to work without adequate PPE.

I don't mean to imply that it's the majority of people here. People are complex and HNers aren't a monolithic group. But I do think the general pro-lockdown sentiment wouldn't be nearly as strong if the majority of posters were in the service or retail industry.

That's not to say people here are selfish, only that it's easier to advocate for safety when you yourself are in a position of relative stability.

Again, the current worldwide figure is 7%, with two thirds of the known cases still pending.

Even if absolutely 0 of the pending cases turn into deaths, and the true percentage of cases that are undetected is the full 50% that has been mentioned recently, that still puts us well above 3%.

Arithmetic: 7 in 100 right now. Add 0 to deaths, resolve all open cases and discover double the known case count: 7 in 200, 3.5%

---

For the surveys you mentioned. The Roosevelt still has active cases. The Princess likewise did at the time that article was written. You're citing that 0.5% figure incorrectly: that's a projection from the Diamond Princess onto China. I found a different source for the German survey -- that link isn't working for some reason. The same applies there.

Most importantly, these surveys are not representative of the global population. See above.

The recent results that have shown 50% asymptomatic cases are not conclusive yet, because it's too soon to say that all of those people will never have symptoms.

I would be ecstatic if I was wrong; I would love it if the upper estimate for asymptomatic cases was correct, and we had far fewer deaths than anyone's expecting. But it's just not in the numbers right now.

"Again, the current worldwide figure is 7%, with two thirds of the known cases still pending."

No. You can not simply calculate the death rate by dividing known deaths by known cases. Those numbers depend on who got tested and on how deaths are being counted, which also varies greatly from country to country. For starters, people who have severe symptoms are more likely to show up at hospitals and get tested. That introduces a huge bias already.

"You're citing that 0.5% figure incorrectly"

No I didn't, read what I wrote. And ALL numbers are projections. Even your calculation is a projection. Your calculation is an extraordinarily simplistic projection. Scientists hopefully make better projections, because they account for more influencing factors.

"The recent results that have shown 50% asymptomatic cases are not conclusive yet, because it's too soon to say that all of those people will never have symptoms"

What makes you so sure? It seems possible to make an estimate of when somebody got infected, especially in those studies of closed communities. After two weeks or so (scientists would determine a more prices number), it would be reasonable to assume they won't develop symptoms.

Death rates change based on how well the healthcare system can keep up. The Diamond Princess was best case in that sense, so the floor of the death rate is about 0.5% if your somewhere with great medical care and f cases. As for the German study:

> About 1,000 people participated in the study by way of questionnaires, throat swabs, and blood tests to detect the presence of COVID-19 antibodies. These interim results are based on data from about 500 people.

So only a sample size of 500 plus not accounting for other local factors there, like being a small village.

"Death rates change based on how well the healthcare system can keep up."

Not really, actually. At least I haven't seen a number of "people who died because they couldn't receive care" - do you have one? It seems possible that once health care systems are overwhelmed, more people die. But how often, if ever in this pandemic, has that point been reached? I don't think that often - there is no good treatment anyway.

"only a sample size of 500"

It's science. 500 can be a lot, or few. It was a sound and thorough study by acknowledged scientists.

> At least I haven't seen a number of "people who died because they couldn't receive care" - do you have one?

No, that's impossible to provide between people dying at home because hospitals are full and the general chaos of those situations. And then there's the big difference between quality of care in various countries, death rates of just about everything will be much higher in Mozambique than they will be in Sweden, it's nonsensical to think that quality of medical care will not have an effect.

> It seems possible that once health care systems are overwhelmed, more people die. But how often, if ever in this pandemic, has that point been reached?

Many times in many countries, look at the national totals somewhere like here (https://covid19info.live/), the trend is nearly always that more cases at once translate to much higher death rates. Partly due to not knowing the true number of infections, but mostly due to the availability of medical care.

"No, that's impossible to provide between people dying at home because hospitals are full and the general chaos of those situations."

That's another claim you make, that you should be able to support. Have people been denied hospital care and therefore died at home?

"the trend is nearly always that more cases at once translate to much higher death rates"

Could have any number of explanations, like maybe infected old people homes.

"it's nonsensical to think that quality of medical care will not have an effect."

Again - are there cases of people dying because they couldn't receive care in the US, and how many?

"but mostly due to the availability of medical care."

Again, can you give any examples of people dying because they couldn't receive care? Especially in the US. And show that it is a significant number of people?

Herd immunity is not just a buzzword. We have two options to reduce the harm of infectious diseases: eradicate them or build herd immunity. The first option has become impossible with this virus that's now truly global and likely much more spread than is known. Therefore herd immunity is the ONLY option.

The question then is how to get as many people as possible immune - you'll either need a vaccine or many need to get the infection and survive. A vaccine is not on the horizon for another 6-12 months, so the best course of action is to build the number of survivors.

What brings death? Mainly the sudden onslaught on the health system. If there is no bed, respirator, blood supply, ... Left then you're much more likely to have patients die than if you have a hospital at regular capacity. There are some that WILL for but by slowly building immunity (i.e. allowing some parts of the population to get exposed) you avoid the high death toll of an unmanaged outbreak.

So if total eradication is impossible, slow infection rates are better than none. Why? Because if you don't build the immunity you will again have violent outbreaks killing many. And you simply can't lock everyone at home for another 6 months.

> If there is no bed, respirator...

Most of the people who go on respirators die. If the health system is going to be our weapon here until the vaccine is developed, we need to find ways to manage the disease before it gets to the point where we intubate people. Remdesivir might be that drug, but we're unlikely to have the kind of supply needed for what you propose until after the first vaccines potentially may be available.

There is no reason you can't keep a disease at a very low level (to the point any given person is unlikely to get the disease over their lifetime) via testing, contact tracing and quarantine. Iceland and South Korea both prove you can in fact.

Iceland is tiny and it is a little early to proclaim SK an unqualified success. If the only way to keep the disease at a very low level is to have a new normal of strong quarantine indefinitely, then you are asking for something unrealistic. Even more so in western democracies where people are accustomed to a great deal of personal freedom.

Eradication is not impossible but takes will and choices.

If the virus had certain death rates and longer periods where it remained dormant. We would have to make choices to eradicate that. We're in a different situation where we have different choices and outcomes.

Immunity does not disappear. The virus could mutate, but not likely and surely would become less deadly. Herd immunity is a fact. It is natural. And we’ll probably reach it sooner than you think. I hope like everybody for a vaccine but I personally don’t think we’ll make it on time enough.

Reply on news.ycombinator.com