5 ms·

It’s hacky and I’m happy to be told how wrong it is. But based on these recent antibody studies in Germany, Finland, and now here in CA, I’ve been assuming an a

by erentz 7y ago

It’s hacky and I’m happy to be told how wrong it is. But based on these recent antibody studies in Germany, Finland, and now here in CA, I’ve been assuming an actual fatality rate of about 0.4, and thus an actual infected rate of 250x our known deaths, which is a much firmer number. This is only a small comfort since it means we may have had about 8.75 million infected and presumably now immune in the USA. Or about 2.6%. There’s still a long way to go in that case.

all_usernames 7y ago
It's known that the vast majority infected do not experience symptoms, or have mild symptoms. I don't think you can draw the conclusion that they had prior immunity.
- erentz 7y ago
  Not concluding prior immunity but concluding that they now have immunity after having been infected. That is in progress towards herd immunity this number of people actually infected being revealed by these antibody studies is what is interesting. And based on my hacky heuristic described above that is currently around 2.6%. So we are still a long way off herd immunity. But at the same time closer than we thought we were.
  - ignoramous 7y ago
    > So we are still a long way off herd immunity. But at the same time closer than we thought we were. Yes and no. A disease is endemic when R0 x S = 1 [0]. R0 for SARS-CoV-2 is estimated to be 2.5 to 3.5 and some models predicting a much higher 5.7 [1]. The herd immunity threshold is given by (1 - S)%, which implies when R0 = 2.5, 60% of the population would need to be immune; similarly 71.42% and 82.45% for R0s, 3.5 and 5.7, respectively. [0] https://en.wikipedia.org/wiki/Endemic_(epidemiology) https://en.wikipedia.org/wiki/Endemic_(epidemiology) [1] https://news.ycombinator.com/item?id=22817942 https://news.ycombinator.com/item?id=22817942
    - fiftyfifty 7y ago
      What percentage of immunity can we expect to see a slowdown in the spread of the disease? While we may need 60% or more to achieve full herd immunity surely somewhere lower than that we can expect to see a noticeable reduction in the spread?
      - ignoramous 7y ago
        > ...noticeable reduction in the spread? I believe the SIR model [0] answers those questions. 3blue1brown's video on the topic is super digestible and informative [1]. In short, the discovery of a prophylactic if not a therapeutic treatment are our best immediate bets, absent which, quarantines and lockdowns will have to persist to curb the spread of a disease as infectious as covid-19 whilst we patiently wait for an effective vaccine against it [2]. [0] https://en.wikipedia.org/wiki/Compartmental_models_in_epidemiology#The_SIR_model https://en.wikipedia.org/wiki/Compartmental_models_in_epidem... [1] https://www.youtube.com/watch?v=gxAaO2rsdIs https://www.youtube.com/watch?v=gxAaO2rsdIs [2] https://en.wikipedia.org/wiki/Vaccine_efficacy https://en.wikipedia.org/wiki/Vaccine_efficacy
- tmandry 7y ago
  “mild” symptoms has been used to mean flu-like, i.e., not something I would call mild and certainly not the same as asympotomatic, but not requiring hospitalization.
  - taeric 7y ago
    I think it is stranger than that. Mild has been anything from a fever for a night, to flu like for a week or so. Really want to get a test on myself and family. We had something that near hospitalized me, but the kids barely registered being sick.
- tom-thistime 7y ago
  Most of what I've seen suggests that at time of positive test results, 40% to 70% of people have not experienced symptoms. Have you seen numbers very different from that? [Edit: 'No symptoms experienced at time of positive test results' is intended to mean the same as 'no symptoms experienced yet at time of positive test results'.]
  - crdrost 7y ago
    That can't be true for the official confirmed numbers. Like, it was for a while impossible to get a test if you are not experiencing multiple symptoms AND can make a case that you might have been exposed via interacting with someone who has traveled. Even now I don't think they're giving the test if you don't have a fever. It's implausible that 50% of the people tested by the official pathway had no symptoms because the official pathway is not available when you have no symptoms. But if you mean that you've seen a 50% "asymptomatic transmission" rate, in other words studies like this that purport to test a bunch of people at random and see how many of them have COVID-19, note that this definition of "asymptomatic" may include many with symptoms not severe enough to be hospitalized -- in other words they might have been feverish and coughing but not moreso than a typical cold. I myself have had a really nasty cough but it is not "dry" but "productive" and it has not come with a fever -- I would really love to be tested but right now that does not seem to be possible! Maybe I am in this above grouping of "asymptomatic" folks. The OP article suggests that as many as 98-99% of cases might not involve hospitalization and therefore might be cases like mine (assuming I do indeed have COVID-19, maybe I don't). Now it is likely that there is some sort of selection bias in how they recruited candidates, but still it tends to push that number above your high end of 70%, suggesting maybe it's closer to 80 or 90%. On the other hand while the paper says that it did its due diligence subtracting out the test's false-positive rate, a rather small error here can have a big impact on that number because there are so few true positives right now. In some ways the paper's claim is a bit of good news, it means that this disease has much lower mortality than we were originally told. (There may be lower bounds on mortality -- I have heard of one town in Italy where 1% of its population is gone due to COVID-19 in which case that would seem to be a good lower bound.) In other ways it is bad news, it means that this disease, still with nontrivial mortality, is going to be much less affected by quarantine countermeasures and the social distancing stuff is really the only reason it is spreading as slowly as it is, so that we need to endure the pain of isolation for much longer -- potentially until testing becomes widespread and cheap. That may not be too much longer, as mentioned in a previous HN comment/story [1] that if you can get testing to work with the old Sanger sequencers used in the Human Genome Project you can maybe ramp up to 200,000 samples/day at first with possibilities to go up to 1M/day once you solve additional logistical problems. Before that the production of kits and sourcing of reagents may be your limit. [1] https://news.ycombinator.com/item?id=22808208 https://news.ycombinator.com/item?id=22808208
jdc 7y ago
But by immune, do you mean asymptomatic and contagious?
- erentz 7y ago
  Only talking about where we are in terms of progress to herd immunity. Edited it to say “presumably [now] immune” which hopefully is clearer. See my other comment below for elaboration.
MiguelVieira 7y ago
I've been using 0.5% for mental math, rounding down from the lower bound from this study in The Lancet: https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(20)30243-7/fulltext https://www.thelancet.com/journals/laninf/article/PIIS1473-3... I would be surprised if, after this current wave of infections, the percentage of people with antibodies in the US is higher than the low single digits. Herd immunity without a vaccine is a pipe dream. Our best bet is to massively ramp up testing and contact tracing and really start pushing the number of infections down to a point where parts of society can start functioning again.
- joe_the_user 7y ago
  So a 0.5% fatality rate sounds like "only" five times the rate of the flu. EXCEPT that I would bet that if you thoroughly traced all the asymptomatic but positive for the flu, it's fatality rate would decline also and it still causes significant deaths. Total Covid deaths are already at a "mild flu season" level already after a month and continuing some level of exponential growth after social distancing. It's sad and awful but it seems like with the current administration approach, the response of ramped up contract tracing has pretty much flown away and with these figures, we're looking 500K deaths, 250K over the next two month and another 250K over a longer period. Hide your parents...
  - irq11 7y ago
    Except that a huge portion of those deaths are in places that may have already reached herd immunity, if the ratio of comfirmed / actual cases is as small as this study suggests. NYC, for example, has 123k confirmed cases. Multiply by 50, and you have 6.1M, which is close to 70% of the total population.
    - JamesBarney 7y ago
      Where are you getting the assumption that only 1 in 50 infected are confirmed? That seems like an especially low confirmed case to infection ratio.
Uhhrrr 7y ago
This is what I've been assuming - 200 to 250x. But 250xdeaths gives you the infected number from two weeks ago. To get current infected you then have to multiply it by communication^(days to death), something like 1.1^14, with the first term varying depending on population density and lockdown measures, and the second on how on the ball that locality is for detecting cases and keeping people alive.
Animats 7y ago
There’s still a long way to go in that case. Yes. Stanford's recent study says 2-4% of the population of Santa Clara County has been infected. Those are probably reasonable bounds. Herd immunity is around 80% for this. So about 20 to 40 times as many people need to die before it's over. That test needs to be repeated weekly to get the growth rate. (Preferably not with the scheme from Verity, which requires that you sign up for a Google account.[1]) Now that California hospital admissions and deaths for COVID-19 have flattened, growth should be linear for a while. But the rate of increase is still not really known. Still, we're probably looking at a year and a million deaths in the US, within a factor of 2 either way. [1] https://verily.com/stories/the-project-baseline-covid-19-program-responding-to-a-health-crisis/ https://verily.com/stories/the-project-baseline-covid-19-pro...
- danieltillett 7y ago
  The problem with herd immunity to this virus is how short the immunity to corona virus is - around a year. If we wait for herd immunity to build from low level natural spread then we will never get there. There is a very major risk with extrapolating the Santa Clara data in the motivation of those being tested has not been controlled. Who is more likely to want to know they have been infected - those who have had classic COVID-19 symptoms or those who haven’t.
  - Animats 7y ago
    How short the immunity to corona virus is - around a year. Source? Right now, that seems unknown - too soon to tell. A few months of measuring antibody levels and we'll know more.
    - danieltillett 7y ago
      This is what we know from other human corona viruses. SARS-CoV-2 could be different, but without evidence I would not want to bet on long immunity.

by erentz 7y ago

all_usernames 7y ago

It's known that the vast majority infected do not experience symptoms, or have mild symptoms. I don't think you can draw the conclusion that they had prior immunity.

erentz 7y ago
Not concluding prior immunity but concluding that they now have immunity after having been infected. That is in progress towards herd immunity this number of people actually infected being revealed by these antibody studies is what is interesting. And based on my hacky heuristic described above that is currently around 2.6%. So we are still a long way off herd immunity. But at the same time closer than we thought we were.
- ignoramous 7y ago
  > So we are still a long way off herd immunity. But at the same time closer than we thought we were. Yes and no. A disease is endemic when R0 x S = 1 [0]. R0 for SARS-CoV-2 is estimated to be 2.5 to 3.5 and some models predicting a much higher 5.7 [1]. The herd immunity threshold is given by (1 - S)%, which implies when R0 = 2.5, 60% of the population would need to be immune; similarly 71.42% and 82.45% for R0s, 3.5 and 5.7, respectively. [0] https://en.wikipedia.org/wiki/Endemic_(epidemiology) https://en.wikipedia.org/wiki/Endemic_(epidemiology) [1] https://news.ycombinator.com/item?id=22817942 https://news.ycombinator.com/item?id=22817942
  - fiftyfifty 7y ago
    What percentage of immunity can we expect to see a slowdown in the spread of the disease? While we may need 60% or more to achieve full herd immunity surely somewhere lower than that we can expect to see a noticeable reduction in the spread?
    - ignoramous 7y ago
      > ...noticeable reduction in the spread? I believe the SIR model [0] answers those questions. 3blue1brown's video on the topic is super digestible and informative [1]. In short, the discovery of a prophylactic if not a therapeutic treatment are our best immediate bets, absent which, quarantines and lockdowns will have to persist to curb the spread of a disease as infectious as covid-19 whilst we patiently wait for an effective vaccine against it [2]. [0] https://en.wikipedia.org/wiki/Compartmental_models_in_epidemiology#The_SIR_model https://en.wikipedia.org/wiki/Compartmental_models_in_epidem... [1] https://www.youtube.com/watch?v=gxAaO2rsdIs https://www.youtube.com/watch?v=gxAaO2rsdIs [2] https://en.wikipedia.org/wiki/Vaccine_efficacy https://en.wikipedia.org/wiki/Vaccine_efficacy
tmandry 7y ago
“mild” symptoms has been used to mean flu-like, i.e., not something I would call mild and certainly not the same as asympotomatic, but not requiring hospitalization.
- taeric 7y ago
  I think it is stranger than that. Mild has been anything from a fever for a night, to flu like for a week or so. Really want to get a test on myself and family. We had something that near hospitalized me, but the kids barely registered being sick.
tom-thistime 7y ago
Most of what I've seen suggests that at time of positive test results, 40% to 70% of people have not experienced symptoms. Have you seen numbers very different from that? [Edit: 'No symptoms experienced at time of positive test results' is intended to mean the same as 'no symptoms experienced yet at time of positive test results'.]
- crdrost 7y ago
  That can't be true for the official confirmed numbers. Like, it was for a while impossible to get a test if you are not experiencing multiple symptoms AND can make a case that you might have been exposed via interacting with someone who has traveled. Even now I don't think they're giving the test if you don't have a fever. It's implausible that 50% of the people tested by the official pathway had no symptoms because the official pathway is not available when you have no symptoms. But if you mean that you've seen a 50% "asymptomatic transmission" rate, in other words studies like this that purport to test a bunch of people at random and see how many of them have COVID-19, note that this definition of "asymptomatic" may include many with symptoms not severe enough to be hospitalized -- in other words they might have been feverish and coughing but not moreso than a typical cold. I myself have had a really nasty cough but it is not "dry" but "productive" and it has not come with a fever -- I would really love to be tested but right now that does not seem to be possible! Maybe I am in this above grouping of "asymptomatic" folks. The OP article suggests that as many as 98-99% of cases might not involve hospitalization and therefore might be cases like mine (assuming I do indeed have COVID-19, maybe I don't). Now it is likely that there is some sort of selection bias in how they recruited candidates, but still it tends to push that number above your high end of 70%, suggesting maybe it's closer to 80 or 90%. On the other hand while the paper says that it did its due diligence subtracting out the test's false-positive rate, a rather small error here can have a big impact on that number because there are so few true positives right now. In some ways the paper's claim is a bit of good news, it means that this disease has much lower mortality than we were originally told. (There may be lower bounds on mortality -- I have heard of one town in Italy where 1% of its population is gone due to COVID-19 in which case that would seem to be a good lower bound.) In other ways it is bad news, it means that this disease, still with nontrivial mortality, is going to be much less affected by quarantine countermeasures and the social distancing stuff is really the only reason it is spreading as slowly as it is, so that we need to endure the pain of isolation for much longer -- potentially until testing becomes widespread and cheap. That may not be too much longer, as mentioned in a previous HN comment/story [1] that if you can get testing to work with the old Sanger sequencers used in the Human Genome Project you can maybe ramp up to 200,000 samples/day at first with possibilities to go up to 1M/day once you solve additional logistical problems. Before that the production of kits and sourcing of reagents may be your limit. [1] https://news.ycombinator.com/item?id=22808208 https://news.ycombinator.com/item?id=22808208

jdc 7y ago

But by immune, do you mean asymptomatic and contagious?

erentz 7y ago
Only talking about where we are in terms of progress to herd immunity. Edited it to say “presumably [now] immune” which hopefully is clearer. See my other comment below for elaboration.

MiguelVieira 7y ago

I've been using 0.5% for mental math, rounding down from the lower bound from this study in The Lancet: https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(20)30243-7/fulltext https://www.thelancet.com/journals/laninf/article/PIIS1473-3... I would be surprised if, after this current wave of infections, the percentage of people with antibodies in the US is higher than the low single digits. Herd immunity without a vaccine is a pipe dream. Our best bet is to massively ramp up testing and contact tracing and really start pushing the number of infections down to a point where parts of society can start functioning again.

joe_the_user 7y ago
So a 0.5% fatality rate sounds like "only" five times the rate of the flu. EXCEPT that I would bet that if you thoroughly traced all the asymptomatic but positive for the flu, it's fatality rate would decline also and it still causes significant deaths. Total Covid deaths are already at a "mild flu season" level already after a month and continuing some level of exponential growth after social distancing. It's sad and awful but it seems like with the current administration approach, the response of ramped up contract tracing has pretty much flown away and with these figures, we're looking 500K deaths, 250K over the next two month and another 250K over a longer period. Hide your parents...
- irq11 7y ago
  Except that a huge portion of those deaths are in places that may have already reached herd immunity, if the ratio of comfirmed / actual cases is as small as this study suggests. NYC, for example, has 123k confirmed cases. Multiply by 50, and you have 6.1M, which is close to 70% of the total population.
  - JamesBarney 7y ago
    Where are you getting the assumption that only 1 in 50 infected are confirmed? That seems like an especially low confirmed case to infection ratio.

Uhhrrr 7y ago

This is what I've been assuming - 200 to 250x. But 250xdeaths gives you the infected number from two weeks ago. To get current infected you then have to multiply it by communication^(days to death), something like 1.1^14, with the first term varying depending on population density and lockdown measures, and the second on how on the ball that locality is for detecting cases and keeping people alive.

Animats 7y ago

There’s still a long way to go in that case. Yes. Stanford's recent study says 2-4% of the population of Santa Clara County has been infected. Those are probably reasonable bounds. Herd immunity is around 80% for this. So about 20 to 40 times as many people need to die before it's over. That test needs to be repeated weekly to get the growth rate. (Preferably not with the scheme from Verity, which requires that you sign up for a Google account.[1]) Now that California hospital admissions and deaths for COVID-19 have flattened, growth should be linear for a while. But the rate of increase is still not really known. Still, we're probably looking at a year and a million deaths in the US, within a factor of 2 either way. [1] https://verily.com/stories/the-project-baseline-covid-19-program-responding-to-a-health-crisis/ https://verily.com/stories/the-project-baseline-covid-19-pro...

danieltillett 7y ago
The problem with herd immunity to this virus is how short the immunity to corona virus is - around a year. If we wait for herd immunity to build from low level natural spread then we will never get there. There is a very major risk with extrapolating the Santa Clara data in the motivation of those being tested has not been controlled. Who is more likely to want to know they have been infected - those who have had classic COVID-19 symptoms or those who haven’t.
- Animats 7y ago
  How short the immunity to corona virus is - around a year. Source? Right now, that seems unknown - too soon to tell. A few months of measuring antibody levels and we'll know more.
  - danieltillett 7y ago
    This is what we know from other human corona viruses. SARS-CoV-2 could be different, but without evidence I would not want to bet on long immunity.