6 ms·

> That means doing your part to not be a vector of infection. I believe I do my part for this and every flu season which literally kills 50,000 Americans each

> That means doing your part to not be a vector of infection. I believe I do my part for this and every flu season which literally kills 50,000 Americans each year. The extra dozen deaths from this new virus really don't change my existing habits: I wash my hands frequently, thoroughly wash fruits and veggies, I stay home if I'm sick, and I get my flu shot each year. I also get regular check ups. I also try to eat a rich, varied diet, get good sleep, do some exercise, don't smoke and avoid over drinkings or overindulging in sweets/fats/salts. I also, and this is real important for me at least, try to avoid getting stressed over things I have no control over. Again, this is just me.

lern_too_spel 7y ago
Without containment, COVID-19 will directly kill at least 10x more people in the US this year. Worse, it will overwhelm hospitals and suffocate medical supply chains, resulting in more indirect deaths as well. You're comparing the deaths of multiple already pandemic flus to a single disease which has not yet spread throughout the United States.
- NicolasGorden 7y ago
  > kill at least 10x more people in the US this year Since the Corona virus has a 3% mortality rate, your estimate of 'at least 10x' would mean 500,000 american dead by the end of year, which would require 150,000,000 to be infected this year, or basically, 48%~ of the population getting infected this year. I'm not saying your wrong, but can you provide a link to a reputable site such as CDC/WHO that supports your claim?
  - lern_too_spel 7y ago
    Unlike the flu, nobody has prior immunity to the disease, so everyone is susceptible to getting the disease. Without containment, nearly everyone will get the disease. You have also undercounted the US population by 10%.
    - NicolasGorden 7y ago
      > Without containment, nearly everyone will get the disease You might be right in your claims. But can you provide me with any reputable sources to verify your claim?
      - lern_too_spel 7y ago
        This is pure logic. Nobody has to say it because the conclusion follows trivially. Which part of the deduction were you not able to follow?
        NicolasGorden 7y ago
        You are absolutely right to ask if I'm not following. Sometimes I fail to follow. In terms of Pure logic, I'd love to learn about that one day. Anyway, deducting a conclusion from reality can be possible, though I'm not sure I agree trivial. I will say I think using statistical past experiences to model future outcomes without understanding underlying mechanisms of action would normally be consider inductive reasoning, not deductive. But even if you arrived at your conclusion using deductive reasoning you could still create a logically valid statement that is unsound. Which is why I ask, what were the original observations that allowed you to deduct that the infection rate will be near 100%.
        lern_too_spel 7y ago
        I didn't capitalize pure logic. You did. Don't put words in my mouth. Clearly, I am using pure to mean "only" or "100%." We can deduce that the infection rate will be nearly 100% because there is no significant subpopulation that is isolated from the rest of the population. Without isolation, the infection spreads. See also: rhinovirus-based colds, which have infected 100% of the US population despite different people having prior immunity to different strains, slowing its spread.
        NicolasGorden 7y ago
        Thanks for your reference. I looked up Rhinovirus. According to wikipedia, it seems that: > 99 serotypes of human rhinoviruses affecting humans have been sequenced. Of which, one seems to be the most contagious: > The human Rhinovirus Strain 1A (which was at the time held responsible for at least 30 percent of the common cold infections) Since the 'cold' is something that about 10% of the population get, and since one strain of Rhinovirus once managed to account for 30% of common cold infections, it would stand to reason that the most infectious Rhinovirus managed to infect 3% of the population in a year. Please help me understand how that allows you to 'deduct' that COVID-19 will have a near 100% infection rate?
  - Simon_says 7y ago
    Half the population becoming infected this year is not that unlikely if no preventive measures are taken. The CDC and WHO stopped being reputable sites about a month ago.
    - NicolasGorden 7y ago
      Ok, so then where did the stats come from? What is your reputable source of information if you think the CDC and WHO aren't reputable. I'm open to learning, maybe you can teach me.
      - claudeganon 7y ago
        The current rate of estimated infection doubling from genomic analysis is 7 days, absent containment measures and other interventions. Assume there are 1000 cases today and do the math. You get there pretty quick: https://bedford.io/blog/ncov-cryptic-transmission/ https://bedford.io/blog/ncov-cryptic-transmission/
        NicolasGorden 7y ago
        > Half the population becoming infected this year is not that unlikely > where did the stats come from? Maybe I misread it, but your article does not corroborate the 50% infection of the US population.
        lern_too_spel 7y ago
        Assuming a mortality rate of 1%, only 55 million people, or about 1/6 of the US population will need to be infected to reach 550k deaths directly from COVID-19. The 1/6 and 1% numbers are both conservative. Hence, at least 10x the flu. The indirect deaths will be larger still because 10x more will be hospitalized, which the medical infrastructure is not prepared to handle.
        claudeganon 7y ago
        Apologies, I linked the wrong thing. This can be extrapolated from what the American Hospital Association is projecting: https://www.businessinsider.com/presentation-us-hospitals-preparing-for-millions-of-hospitalizations-2020-3 https://www.businessinsider.com/presentation-us-hospitals-pr... As the article also references, Marc Lipsitch at Harvard projects the same.
  - pcmonk 7y ago
    You're off by an order of magnitude. 3% of 150m is 4.5m.
    - NicolasGorden 7y ago
      Thanks. You are 100% right.
  - Bombthecat 7y ago
    Wasnt someone from who or was it a German scientist? I forgot, but he basically excepts that 60 to 70 percent will be infected before it slows down.
    - NicolasGorden 7y ago
      IF the WHO estimated 80% infection rates I would be very concerned and would not have written any of this. Please, if you have a WHO link that predicts what you mention, please share it with me.
      - Bombthecat 7y ago
        A yeah, was a specialist from Germany, Berlin hospital charité: https://www.google.com/amp/s/www.sueddeutsche.de/gesundheit/krankheiten-experte-erwartet-60-bis-70-prozent-infizierte-in-deutschland-dpa.urn-newsml-dpa-com-20090101-200228-99-108884!amp https://www.google.com/amp/s/www.sueddeutsche.de/gesundheit/... Google translate should give you the jist.
natalyarostova 7y ago
> The extra dozen deaths from this new virus really don't change my existing habits This is either deliberately misleading, or you are not properly able to reason about and understand exponential growth. No one cares about 12 deaths. It's the exponential growth in something that has the potential to be the flu, but roughly 10x worse.
- NicolasGorden 7y ago
  > understand exponential growth Oh I know what exponential growth is, it is a basic mathematical concept and I'm quite good at math. The thing is, extrapolating the growth rate of new phenomena is not logical. If this isn't intuitive, think of it this way: I have $1 today. I earn $100 and by tomorrow I have $101. I sell my car and the day after tomorrow I have $10,101. Would it make sense to continue projecting my increased asset accumulation at the rate of 10,000% daily for the next 10 years? Of course not, it would mean I would have more assets than exist in the world. Growth rates are higher when numbers are lower and lower when numbers are higher. It's the nature of change in any finite data set. > roughly 10x worse According to the WHO Corona virus has a 3% mortality rate, your estimate of 'at least 10x' would mean 500,000 american dead by the end of year, which would require 150,000,000 to be infected this year, or basically, 48%~ of the population getting infected this year. I'm not saying your wrong, but can you provide a link to a reputable site such as CDC/WHO that supports your claim?
  - natalyarostova 7y ago
    Sure, there are saturation parameters in the differential equations that prevent that result. There are smart epidemiologists who project those claims >https://www.theatlantic.com/health/archive/2020/02/covid-vaccine/607000/ https://www.theatlantic.com/health/archive/2020/02/covid-vac... The CDC/WHO are still trying to push for containment, so I'm unaware of any concrete predictions they are taking with regards to the final infection rate. More generally, the more I've been binge studying epidemiology (not claiming to be an authority), the more I'm learning that, as a field, it has difficulty with long-run predictions. That's not super surprising, I work in the statsy/ML forecasting world, and these types of unstable, unknown processes are very hard to forecast accurately. What we do know though is that there is good reason to believe it's parameters governing infection (e.g. secondary infection distribution) are more infectious that diseases such as H1N1, which went on to infect ~20% of the world. We do know that the mortality is 1-3%, depending on ICU saturation. There are reasonable base rate scenarios to view this as a likely outcome. We need to do our best to contain this and keep it as a slow burn, or we will be dealing with hundreds of thousands of Americans dead, and millions hospitalized. To be a little more specific, the deaths given a functioning health system are less concerning to me. It's the fact that this has a very high hospitalization rate of 5-20%. That's why we need to keep spread slow, so the hospitals can manage.

lern_too_spel 7y ago

Without containment, COVID-19 will directly kill at least 10x more people in the US this year. Worse, it will overwhelm hospitals and suffocate medical supply chains, resulting in more indirect deaths as well. You're comparing the deaths of multiple already pandemic flus to a single disease which has not yet spread throughout the United States.

NicolasGorden 7y ago
> kill at least 10x more people in the US this year Since the Corona virus has a 3% mortality rate, your estimate of 'at least 10x' would mean 500,000 american dead by the end of year, which would require 150,000,000 to be infected this year, or basically, 48%~ of the population getting infected this year. I'm not saying your wrong, but can you provide a link to a reputable site such as CDC/WHO that supports your claim?
- lern_too_spel 7y ago
  Unlike the flu, nobody has prior immunity to the disease, so everyone is susceptible to getting the disease. Without containment, nearly everyone will get the disease. You have also undercounted the US population by 10%.
  - NicolasGorden 7y ago
    > Without containment, nearly everyone will get the disease You might be right in your claims. But can you provide me with any reputable sources to verify your claim?
    - lern_too_spel 7y ago
      This is pure logic. Nobody has to say it because the conclusion follows trivially. Which part of the deduction were you not able to follow?
      - NicolasGorden 7y ago
        You are absolutely right to ask if I'm not following. Sometimes I fail to follow. In terms of Pure logic, I'd love to learn about that one day. Anyway, deducting a conclusion from reality can be possible, though I'm not sure I agree trivial. I will say I think using statistical past experiences to model future outcomes without understanding underlying mechanisms of action would normally be consider inductive reasoning, not deductive. But even if you arrived at your conclusion using deductive reasoning you could still create a logically valid statement that is unsound. Which is why I ask, what were the original observations that allowed you to deduct that the infection rate will be near 100%.
        lern_too_spel 7y ago
        I didn't capitalize pure logic. You did. Don't put words in my mouth. Clearly, I am using pure to mean "only" or "100%." We can deduce that the infection rate will be nearly 100% because there is no significant subpopulation that is isolated from the rest of the population. Without isolation, the infection spreads. See also: rhinovirus-based colds, which have infected 100% of the US population despite different people having prior immunity to different strains, slowing its spread.
        NicolasGorden 7y ago
        Thanks for your reference. I looked up Rhinovirus. According to wikipedia, it seems that: > 99 serotypes of human rhinoviruses affecting humans have been sequenced. Of which, one seems to be the most contagious: > The human Rhinovirus Strain 1A (which was at the time held responsible for at least 30 percent of the common cold infections) Since the 'cold' is something that about 10% of the population get, and since one strain of Rhinovirus once managed to account for 30% of common cold infections, it would stand to reason that the most infectious Rhinovirus managed to infect 3% of the population in a year. Please help me understand how that allows you to 'deduct' that COVID-19 will have a near 100% infection rate?
- Simon_says 7y ago
  Half the population becoming infected this year is not that unlikely if no preventive measures are taken. The CDC and WHO stopped being reputable sites about a month ago.
  - NicolasGorden 7y ago
    Ok, so then where did the stats come from? What is your reputable source of information if you think the CDC and WHO aren't reputable. I'm open to learning, maybe you can teach me.
    - claudeganon 7y ago
      The current rate of estimated infection doubling from genomic analysis is 7 days, absent containment measures and other interventions. Assume there are 1000 cases today and do the math. You get there pretty quick: https://bedford.io/blog/ncov-cryptic-transmission/ https://bedford.io/blog/ncov-cryptic-transmission/
      - NicolasGorden 7y ago
        > Half the population becoming infected this year is not that unlikely > where did the stats come from? Maybe I misread it, but your article does not corroborate the 50% infection of the US population.
        lern_too_spel 7y ago
        Assuming a mortality rate of 1%, only 55 million people, or about 1/6 of the US population will need to be infected to reach 550k deaths directly from COVID-19. The 1/6 and 1% numbers are both conservative. Hence, at least 10x the flu. The indirect deaths will be larger still because 10x more will be hospitalized, which the medical infrastructure is not prepared to handle.
        claudeganon 7y ago
        Apologies, I linked the wrong thing. This can be extrapolated from what the American Hospital Association is projecting: https://www.businessinsider.com/presentation-us-hospitals-preparing-for-millions-of-hospitalizations-2020-3 https://www.businessinsider.com/presentation-us-hospitals-pr... As the article also references, Marc Lipsitch at Harvard projects the same.
- pcmonk 7y ago
  You're off by an order of magnitude. 3% of 150m is 4.5m.
  - NicolasGorden 7y ago
    Thanks. You are 100% right.
- Bombthecat 7y ago
  Wasnt someone from who or was it a German scientist? I forgot, but he basically excepts that 60 to 70 percent will be infected before it slows down.
  - NicolasGorden 7y ago
    IF the WHO estimated 80% infection rates I would be very concerned and would not have written any of this. Please, if you have a WHO link that predicts what you mention, please share it with me.
    - Bombthecat 7y ago
      A yeah, was a specialist from Germany, Berlin hospital charité: https://www.google.com/amp/s/www.sueddeutsche.de/gesundheit/krankheiten-experte-erwartet-60-bis-70-prozent-infizierte-in-deutschland-dpa.urn-newsml-dpa-com-20090101-200228-99-108884!amp https://www.google.com/amp/s/www.sueddeutsche.de/gesundheit/... Google translate should give you the jist.

natalyarostova 7y ago

> The extra dozen deaths from this new virus really don't change my existing habits This is either deliberately misleading, or you are not properly able to reason about and understand exponential growth. No one cares about 12 deaths. It's the exponential growth in something that has the potential to be the flu, but roughly 10x worse.

NicolasGorden 7y ago
> understand exponential growth Oh I know what exponential growth is, it is a basic mathematical concept and I'm quite good at math. The thing is, extrapolating the growth rate of new phenomena is not logical. If this isn't intuitive, think of it this way: I have $1 today. I earn $100 and by tomorrow I have $101. I sell my car and the day after tomorrow I have $10,101. Would it make sense to continue projecting my increased asset accumulation at the rate of 10,000% daily for the next 10 years? Of course not, it would mean I would have more assets than exist in the world. Growth rates are higher when numbers are lower and lower when numbers are higher. It's the nature of change in any finite data set. > roughly 10x worse According to the WHO Corona virus has a 3% mortality rate, your estimate of 'at least 10x' would mean 500,000 american dead by the end of year, which would require 150,000,000 to be infected this year, or basically, 48%~ of the population getting infected this year. I'm not saying your wrong, but can you provide a link to a reputable site such as CDC/WHO that supports your claim?
- natalyarostova 7y ago
  Sure, there are saturation parameters in the differential equations that prevent that result. There are smart epidemiologists who project those claims >https://www.theatlantic.com/health/archive/2020/02/covid-vaccine/607000/ https://www.theatlantic.com/health/archive/2020/02/covid-vac... The CDC/WHO are still trying to push for containment, so I'm unaware of any concrete predictions they are taking with regards to the final infection rate. More generally, the more I've been binge studying epidemiology (not claiming to be an authority), the more I'm learning that, as a field, it has difficulty with long-run predictions. That's not super surprising, I work in the statsy/ML forecasting world, and these types of unstable, unknown processes are very hard to forecast accurately. What we do know though is that there is good reason to believe it's parameters governing infection (e.g. secondary infection distribution) are more infectious that diseases such as H1N1, which went on to infect ~20% of the world. We do know that the mortality is 1-3%, depending on ICU saturation. There are reasonable base rate scenarios to view this as a likely outcome. We need to do our best to contain this and keep it as a slow burn, or we will be dealing with hundreds of thousands of Americans dead, and millions hospitalized. To be a little more specific, the deaths given a functioning health system are less concerning to me. It's the fact that this has a very high hospitalization rate of 5-20%. That's why we need to keep spread slow, so the hospitals can manage.