5 ms·

if the US has 33K deaths and the death rate is 0.6% then the US must have at least 5 million cases. If the death rate is lower (most likely around 0.3%) then i

if the US has 33K deaths and the death rate is 0.6% then the US must have at least 5 million cases. If the death rate is lower (most likely around 0.3%) then it means 10 million infections. Also a good news actually.

cm2187 7y ago
Agree, though still a long way from herd immunity, which I understand starts at 50%-ish.
- btilly 7y ago
  Herd immunity varies by disease. A lot. For each disease we estimate a number called R0, which is how many people the average person will transmit it to if no precautions are taken. Herd immunity sets in when only 1/R0 people have not managed to get it. With the flu, R0 is about 1.3. Meaning that herd immunity to a particular strain starts slowing the spread after about 1/4 of us are immune. (It can continue spreading for some time after that point.) I calculated that from the fact that 1 - 1/1.3 = 0.23076923076923076... R0 for COVID-19 is likely around 5.6 +- 0.9 per https://www.medrxiv.org/content/10.1101/2020.02.07.20021154v1.If https://www.medrxiv.org/content/10.1101/2020.02.07.20021154v... we take the middle of that range, over 82% of us have to get it before herd immunity starts to work in our favor.
  - dmoy 7y ago
    This is also why people refusing the MMR vaccine is really, really hurtful to vulnerable populations. Measles R0 is like 15 or something - it's crazy infectious. So one kid in a daycare or class of 20 with parents who don't believe in vaccines is enough to push you under herd immunity. Bake in the fact that MMR vaccine is only ~97% good against measles, and you end up with a bunch of dead kids. We got lucky last time around in Seattle area, in that it only cost us $1m and 0 deaths to get it under control.
  - networkimprov 7y ago
    Herd immunity is derived from R0 and "susceptibility" percentage. The value for the latter also varies with disease and strain. If you look at the Italian data, their % positive level never crossed ~30%, and overall averaged <20%. That might indicate that susceptibility for this agent is well below 50%. https://en.wikipedia.org/wiki/2020_coronavirus_pandemic_in_Italy https://en.wikipedia.org/wiki/2020_coronavirus_pandemic_in_I...
    - btilly 7y ago
      If you look at the Italian data, their % positive level never crossed ~30%, and overall averaged <20%. That might indicate that susceptibility for this agent is well below 50%. The more likely explanation is that social distancing was implemented early enough, and tests were given to people who might or might not have had it. There is no reason to believe that we burned through most of the vulnerable people in the population.
  - Mvandenbergh 7y ago
    Could you post a working version of that link? Most R0 estimates I have seen are between 2.5 and 3.5. 5.6 seems really high to me. It is worth considering that even if the % immune is lower than the herd immunity level, it still slows the outbreak and reduces Re. That gives us a much slower ramp up and gives other measures like less restrictive social distancing, contact tracing, etc. a chance to work. Even if all those together only get Re down to between 1 and 2, we still benefit since it means we might only need another 1 or 2 lockdown periods before we get an effective vaccine.
    - DanBC 7y ago
      I think it's this link: https://www.medrxiv.org/content/10.1101/2020.02.07.20021154v1 https://www.medrxiv.org/content/10.1101/2020.02.07.20021154v...
pesfandiar 7y ago
Those death rates in US are optimistic in my opinion. What if the actual rate is higher because of worse health in impoverished demographics, or the simple fact people might avoid healthcare because of its prohibitive costs?
- glofish 7y ago
  these death rates come from published research if anything the death rates over across all age groups are even lower - this is a very common observation as initially most non-symptomatic are missed. there is a very strong age dependence with this disease, and probably many other risk factors - predisposition to pneumonia etc.
  - pesfandiar 7y ago
    If I understand correctly, research about the death rate is based on people who were tested positive (any source that's based on the general population?). This, depending on the region, could mean people with severe symptoms or non-randomly-selected test takers only. This adds much uncertainty to the estimated overall infection numbers.
- three_seagrass 7y ago
  Lots of people in NYC are getting dragged out of their homes in body bags that will never get tested. https://www.yahoo.com/news/teams-retrieving-280-bodies-day-170100636.html https://www.yahoo.com/news/teams-retrieving-280-bodies-day-1...
  - radioactivist 7y ago
    They are now including many of the suspected (but untested) cases in their reported numbers now. It increased the total from New York by a significant amount, see [1]. The increase is roughly in line with similar estimates from excess mortality [2], so I would guess they are catching most of the deaths with the new reporting scheme. [1] https://www1.nyc.gov/assets/doh/downloads/pdf/imm/covid-19-deaths-confirmed-probable-daily-04142020.pdf https://www1.nyc.gov/assets/doh/downloads/pdf/imm/covid-19-d... [2] https://www.nytimes.com/interactive/2020/04/10/upshot/coronavirus-deaths-new-york-city.html https://www.nytimes.com/interactive/2020/04/10/upshot/corona...
IAmEveryone 7y ago
That’s 3% of population. Meaning it would 25x times as many deaths to get to “herd immunity”. That’s 750,000 deaths. Good news?
- glofish 7y ago
  not sure what you are getting at, having 10x more infections yet the situation is still manageable is actually pretty good news

cm2187 7y ago

Agree, though still a long way from herd immunity, which I understand starts at 50%-ish.

btilly 7y ago
Herd immunity varies by disease. A lot. For each disease we estimate a number called R0, which is how many people the average person will transmit it to if no precautions are taken. Herd immunity sets in when only 1/R0 people have not managed to get it. With the flu, R0 is about 1.3. Meaning that herd immunity to a particular strain starts slowing the spread after about 1/4 of us are immune. (It can continue spreading for some time after that point.) I calculated that from the fact that 1 - 1/1.3 = 0.23076923076923076... R0 for COVID-19 is likely around 5.6 +- 0.9 per https://www.medrxiv.org/content/10.1101/2020.02.07.20021154v1.If https://www.medrxiv.org/content/10.1101/2020.02.07.20021154v... we take the middle of that range, over 82% of us have to get it before herd immunity starts to work in our favor.
- dmoy 7y ago
  This is also why people refusing the MMR vaccine is really, really hurtful to vulnerable populations. Measles R0 is like 15 or something - it's crazy infectious. So one kid in a daycare or class of 20 with parents who don't believe in vaccines is enough to push you under herd immunity. Bake in the fact that MMR vaccine is only ~97% good against measles, and you end up with a bunch of dead kids. We got lucky last time around in Seattle area, in that it only cost us $1m and 0 deaths to get it under control.
- networkimprov 7y ago
  Herd immunity is derived from R0 and "susceptibility" percentage. The value for the latter also varies with disease and strain. If you look at the Italian data, their % positive level never crossed ~30%, and overall averaged <20%. That might indicate that susceptibility for this agent is well below 50%. https://en.wikipedia.org/wiki/2020_coronavirus_pandemic_in_Italy https://en.wikipedia.org/wiki/2020_coronavirus_pandemic_in_I...
  - btilly 7y ago
    If you look at the Italian data, their % positive level never crossed ~30%, and overall averaged <20%. That might indicate that susceptibility for this agent is well below 50%. The more likely explanation is that social distancing was implemented early enough, and tests were given to people who might or might not have had it. There is no reason to believe that we burned through most of the vulnerable people in the population.
- Mvandenbergh 7y ago
  Could you post a working version of that link? Most R0 estimates I have seen are between 2.5 and 3.5. 5.6 seems really high to me. It is worth considering that even if the % immune is lower than the herd immunity level, it still slows the outbreak and reduces Re. That gives us a much slower ramp up and gives other measures like less restrictive social distancing, contact tracing, etc. a chance to work. Even if all those together only get Re down to between 1 and 2, we still benefit since it means we might only need another 1 or 2 lockdown periods before we get an effective vaccine.
  - DanBC 7y ago
    I think it's this link: https://www.medrxiv.org/content/10.1101/2020.02.07.20021154v1 https://www.medrxiv.org/content/10.1101/2020.02.07.20021154v...

pesfandiar 7y ago

Those death rates in US are optimistic in my opinion. What if the actual rate is higher because of worse health in impoverished demographics, or the simple fact people might avoid healthcare because of its prohibitive costs?

glofish 7y ago
these death rates come from published research if anything the death rates over across all age groups are even lower - this is a very common observation as initially most non-symptomatic are missed. there is a very strong age dependence with this disease, and probably many other risk factors - predisposition to pneumonia etc.
- pesfandiar 7y ago
  If I understand correctly, research about the death rate is based on people who were tested positive (any source that's based on the general population?). This, depending on the region, could mean people with severe symptoms or non-randomly-selected test takers only. This adds much uncertainty to the estimated overall infection numbers.
three_seagrass 7y ago
Lots of people in NYC are getting dragged out of their homes in body bags that will never get tested. https://www.yahoo.com/news/teams-retrieving-280-bodies-day-170100636.html https://www.yahoo.com/news/teams-retrieving-280-bodies-day-1...
- radioactivist 7y ago
  They are now including many of the suspected (but untested) cases in their reported numbers now. It increased the total from New York by a significant amount, see [1]. The increase is roughly in line with similar estimates from excess mortality [2], so I would guess they are catching most of the deaths with the new reporting scheme. [1] https://www1.nyc.gov/assets/doh/downloads/pdf/imm/covid-19-deaths-confirmed-probable-daily-04142020.pdf https://www1.nyc.gov/assets/doh/downloads/pdf/imm/covid-19-d... [2] https://www.nytimes.com/interactive/2020/04/10/upshot/coronavirus-deaths-new-york-city.html https://www.nytimes.com/interactive/2020/04/10/upshot/corona...

IAmEveryone 7y ago

That’s 3% of population. Meaning it would 25x times as many deaths to get to “herd immunity”. That’s 750,000 deaths. Good news?

glofish 7y ago
not sure what you are getting at, having 10x more infections yet the situation is still manageable is actually pretty good news