7 ms·

Coronavirus fatality rate could be as low as 0.26%, CDC says

43 points by mrfusion 6y ago

harry8 6y ago
Back of the envelope check as an exercise. Population of northern Italy. Assume everyone contacted covid. How many dead / pop More or less than 0.26% Or new York, or London. See what numbers you come up with. Is this a reasonable thing for the cdc to say or not? Please do the exercise yourself and draw your own conclusion.
- mikem170 6y ago
  I wondered about that. But it is a reasonable number if you take into account that there were not enough tests to detect all cases. I assume that now they are getting antibody data and see that a lot more people caught this than they thought, not exhibiting symptoms or needing medical care. And there are things we still don't know, about how it spreads (turns out not so much on surfaces), and about variants that are out there (why worse in NYC and Europe than in parts of asia and west coast?) And local conditions (NYC is very dense with subways, Northern Italy with intergenerational housing and smoking, etc). More data is always good.
  - harry8 6y ago
    > But it is a reasonable number if you take into account that there were not enough tests to detect all cases. Assume every single person contracted it and testing is irrelevant in the calculation. Then you can adjust down to 90% contracted covid, or 80% or 20% or whatever to see how that affects things if you want to. Start with deaths/population and see what number comes up. It's a really useful starting point. And a really useful skill to have to be able to check a supplied number with a back of the envelope calculation. I say do it! For more than one reason it's worthwhile.
- gnusty_gnurc 6y ago
  IFR is a linear combination of age groups and their associated risk. It turns out NYC was an unmitigated disaster around nursing homes. That'll drive up the statistics quite a bit in that dead/population figure you're mentioning.
  - TheCoelacanth 6y ago
    NYC has a median age of two years younger than the country as a whole. Age distribution is not a plausible explanation for NYC having a particularly large number of deaths from the outbreak.
    - zaroth 6y ago
      It sounds like you are assuming the virus spreads uniformly through the population. It does not. The median age of the population isn’t as relevant as the prevalence of the virus within specific demographics in the population. NYC may have a slightly smaller elderly population, but that won’t matter if the prevalence of the virus in their elderly population is, say, 2x higher than the national average. I’ve only been following it loosely, but apparently NYC brought the virus into their nursing homes by discharging SARS-CoV-2 positive patients back into them, and the IFR in a nursing home is certainly not 0.2%.
      - mindslight 6y ago
        The original comment is putting a bound on the case fatality rate by simplistically assuming everyone has been infected. The actual prevalence in nursing homes is irrelevant with that assumption. Nursing homes being prone to infection just tightens the bound.
        zaroth 6y ago
        I think you are misreading OP. They are saying the prevalence in nursing homes was higher (“an unmitigated disaster”) which drives up the average IFR.
        mindslight 6y ago
        My comment applied to the comment you're referencing as well. The original comment is the one above that.
- tinus_hn 6y ago
  A useless exercise unless you also take age distribution in account. Anyway for Lombardy, the hardest hit region of Italy the answer is 0.16%.
mikem170 6y ago
>According to new estimates released by the Centers for Disease Control and Prevention (CDC), the COVID-19 infection fatality rate may be as low as 0.26%, a number that only doubles the seasonal flu but significantly lower than earlier estimates. The news has been pushing worst case scenarios at every step of the way, not doing a good job of presenting the complexities behind the facts and estimates. And lots of people seem to be blind to this. Maybe schools could be doing a better job teaching critical thinking?
- pengaru 6y ago
  > Maybe schools could be doing a better job teaching critical thinking? Critical thinkers don't make for easy consumers.
lmilcin 6y ago
I wish there was an assessment with error bars not spanning multiple orders of magnitude... Jokes aside. If the number was true, for Italy (33k reported deaths) this would give around 12M infected. For US (99k reported deaths) around 36M infected. Those numbers seem to be unrealistic. I remember Italy did some tests and even in the worst hit places less than 7% of population developed antibodies, not average 25% for entire Italy as suggested. I know this is very, very back of envelop calculation but it would be nice to somehow reconcile those numbers.
- emeerson 6y ago
  Initially I thought of those numbers (36m in the US) as totally unrealistic. But then comparing to seasonal influenza case counts, they seem to be within that order of magnitude (39m-56m cases): https://www.cdc.gov/flu/about/burden/preliminary-in-season-estimates.htm https://www.cdc.gov/flu/about/burden/preliminary-in-season-e... would that be implausible? am I misunderstanding or comparing the wrong metrics? I'm actually curious to see the time interval of when the first true COVID-19 case appeared plotted over first true 2019-2020 influenza cases. If along the same timeline, you'd figure that social distancing would have affected the R0 of both. Assuming that SARS-CoV-2 is truly more (or comparably) contagious, 36M could be plausible. To be clear: I am not equivalating COVID-19 fatality rate w/ the flu, or suggesting it is "innocuous" as the flu is.
  - rusticpenn 6y ago
    But these numbers cannot be compared as a lockdown is in effect..
- rurban 6y ago
  Italy is one of the 7 outlier countries. It has an IFR even above 1%. But Italy is always the highest flu outlier worldwide (regularily above 1%), so there no reason to bring Italy into the discussion. The other outliers are much more interesting, like Belgium, which has also a reporting problem. UK, NL, Spain,... 0.3% has been the established number for a few months already. Reporting always concentrated on the meaningless CFR at 3.4%, for political reasons. 10x more asymptotic cases also reflect all the antibody tests. Asymptotic means an ordinary cold. That's why it's so dangerous. Low fatality rate, no symptoms, high reproduction.
op00to 6y ago
What about permanent cardiovascular damage? What’s that rate for flu and then for coronavirus?
- xt00 6y ago
  Yes, it would be very helpful to have the following numbers: 1) mortality rate 2) no symptom rate 3) negative outcome rate ->(some clear impact to patient who has survived, but has life impacting effects) So if you find for example that within a population of 100000 people that you have 35% with no symptoms, and 0.26% died, then 64%+ had symptoms, so then how many people had symptoms that likely are long-lasting. If for example 10% of people had long lasting impacts, then that is a far greater number than 0.26%.. so before people say "look it ain't so bad".. then we need to understand that number before people invite 100 people over to their house party.
sharemywin 6y ago
330M*.0026 = 857K. in US. Herd immunity is like 60% of that number. That's a lot of people. That could be your parent or grandparent or you. I get so tired of people comparing this to the flu. It's over taken the number of deaths from the flu in 2 months not 12. And that's while we were in locked down. Also, a whole lot of people get symptoms. I don't think having flu like symptoms for 2 weeks(not 2-3 days) sounds like fun either.
- vb6sp6 6y ago
  It's plain greed. The only people clamoring to "open up" always have a vested interest. Pure scum.
  - samdamsamm 6y ago
    In the United States a lot of people I know are eager to get back to work because they don’t want to get evicted or fall behind on car payments and fall into the Spiraling USA Debt Trap that ruins so many lives. I don’t blame them in the least. I was laid off in March and have still not received any unemployment insurance myself. If that is still the case this time next month I will be freaking out all the same.
    - sharemywin 6y ago
      I by no means complaining about a people that have to go back to work. It just sucks that people have to make a choice between potentially risking their's or their family's lives for a paycheck. I just don't think people should be understating or under estimating this thing.
      - samdamsamm 6y ago
        So the problem is the catastrophic dismantling of the welfare state over the past several decades. Ironically, most Americans complaining about getting the economy up and running again also vote against having a basic safety net. I’d say you’re getting what you asked for. Should have at least voted for Medicare For All.
WheelsAtLarge 6y ago
The death rate is one thing but the hospitalization rate should not be ignored. We've seen the overflowing capacity in New York and other places so we need to be careful with just quoting the death rate and feeling that it's no big deal. Whatever the death rate is we know that over 100k people have died. And who knows how many thousands of survivors have dealt and will deal with the damage the virus inflicts on their bodies. We should not take that lightly. I'm one that thinks that face masks and keeping social distance is the best way to go. If we followed that without fail the transmission rate would collapse and we would be close to normal.
- nikolay 6y ago
  This is the biggest thing people don't get - this is not the flu and surviving it or having mild symptoms does not mean that you didn't signed up a chornic CVD or metabolic disease based on the early data that the virus causes a permanent myocardic injury, lung scaring, and God knows what else. We'll be studying the damage for the decades to come! I think many people forget that even the "benign" HPV causes different types of cancer, and we recently found it. So, a new virus means being extra careful, but how can you explain this to people who strongly believe that Bill Gates wants to chip them thru vaccines and that 5G kills birds and people?
  - tinus_hn 6y ago
    You are putting up a straw man and then taking it down. Not everyone who is not panicking about COVID-19 is a conspiracy theorist.
    - WheelsAtLarge 6y ago
      It does not change the primary narrative. COVID is dangerous and highly contagious so we need to be careful and not treat it lightly.
      - tinus_hn 6y ago
        Many things are dangerous. Being careful means evaluating risks, not hysterically avoiding all risks at any cost. Smoking costs more lives than COVID-19 and it is much easier to save these lives. Yet it’s acceptable to do very little. But somehow COVID-19 warrants endless lockdowns.
cozzyd 6y ago
Considering that that's roughly the population fatality rate in NYC, and significantly less than than the population fatality rate in Bergamo province, that seems unlikely. Is the methodology available somewhere? Scenario 5 looks suspiciously like the geometric average of 0.2 and 1...
- rogerkirkness 6y ago
  PFR in Italy based on one of the oldest populations in the world, with the worst NOx pollution in the world because of the mountains and industry, without controlling for confounding reasons for dying (e.g. all severe cases that hit the hospital counted and all mild cases that don't aren't).
  - cozzyd 6y ago
    The population fatality rate in Bergamo is based on the total number of COVID deaths (estimated from excess mortality) and the total population. See https://www.medrxiv.org/content/10.1101/2020.04.15.20067074v3 https://www.medrxiv.org/content/10.1101/2020.04.15.20067074v...
zaroth 6y ago
The population-wide total IFR could be as low as 0.26% but it would depend heavily on the demographics of the people infected. If you infect 100 people under the age of 40 your IFR will likely be much lower than even 0.26%. If just 10 of those 100 are over 80 then it’s a totally different story. Not all diseases have such a high spread in the mortality rate versus age as COVID, which spans effectively 0%-20%. It makes discussions on the average IFR less useful than they might otherwise be.
tinus_hn 6y ago
It would probably better to link to the linked page https://www.cdc.gov/coronavirus/2019-ncov/hcp/planning-scenarios.html https://www.cdc.gov/coronavirus/2019-ncov/hcp/planning-scena... Note that the risks are dramatically different for different age groups, so if one calculates a grand total this requires assumptions about age distribution. The page notes the data is about the US.
randyrand 6y ago
0.26% is well below the % of people that die every year (~1%). I wonder how many of those people would have died this year regardless. We will be able to tell in retrospect if the annual death rate is not significantly moved versus last year.
- cecja 6y ago
  There are studies that covid did reduce the lifespan of most of the dead by 10 years. Your argument is as retarded as your government. https://www.businessinsider.com/study-finds-that-people-who-have-coronavirus-could-die-10-years-early-2020-5?r=DE&IR=T https://www.businessinsider.com/study-finds-that-people-who-...
drinkcocacola 6y ago
Well, I've been following numbers for the last two months in different countries from different sources. My initial best estimations were that the fatality rate was about 0.77% given the numbers in South Korea (high number of tests). Other countries such Spain were reporting rates of about 10%, however testing was quite low in comparison. Recently Spain did a seroprevalence test (number of people with antibodies for Coronavirus) with a large sample of people. They found that 5% of the population had antibodies (2.4M people). Which means that the fatality rate is actually about 1.1% and not 10%. This numbers (from 0.77% to 1.16%) have been constant in different countries, so let me express my disbelief with the numbers published by the CDC.

43 points by mrfusion 6y ago

Back of the envelope check as an exercise. Population of northern Italy. Assume everyone contacted covid. How many dead / pop More or less than 0.26% Or new York, or London. See what numbers you come up with. Is this a reasonable thing for the cdc to say or not? Please do the exercise yourself and draw your own conclusion.

mikem170 6y ago
I wondered about that. But it is a reasonable number if you take into account that there were not enough tests to detect all cases. I assume that now they are getting antibody data and see that a lot more people caught this than they thought, not exhibiting symptoms or needing medical care. And there are things we still don't know, about how it spreads (turns out not so much on surfaces), and about variants that are out there (why worse in NYC and Europe than in parts of asia and west coast?) And local conditions (NYC is very dense with subways, Northern Italy with intergenerational housing and smoking, etc). More data is always good.
- harry8 6y ago
  > But it is a reasonable number if you take into account that there were not enough tests to detect all cases. Assume every single person contracted it and testing is irrelevant in the calculation. Then you can adjust down to 90% contracted covid, or 80% or 20% or whatever to see how that affects things if you want to. Start with deaths/population and see what number comes up. It's a really useful starting point. And a really useful skill to have to be able to check a supplied number with a back of the envelope calculation. I say do it! For more than one reason it's worthwhile.
gnusty_gnurc 6y ago
IFR is a linear combination of age groups and their associated risk. It turns out NYC was an unmitigated disaster around nursing homes. That'll drive up the statistics quite a bit in that dead/population figure you're mentioning.
- TheCoelacanth 6y ago
  NYC has a median age of two years younger than the country as a whole. Age distribution is not a plausible explanation for NYC having a particularly large number of deaths from the outbreak.
  - zaroth 6y ago
    It sounds like you are assuming the virus spreads uniformly through the population. It does not. The median age of the population isn’t as relevant as the prevalence of the virus within specific demographics in the population. NYC may have a slightly smaller elderly population, but that won’t matter if the prevalence of the virus in their elderly population is, say, 2x higher than the national average. I’ve only been following it loosely, but apparently NYC brought the virus into their nursing homes by discharging SARS-CoV-2 positive patients back into them, and the IFR in a nursing home is certainly not 0.2%.
    - mindslight 6y ago
      The original comment is putting a bound on the case fatality rate by simplistically assuming everyone has been infected. The actual prevalence in nursing homes is irrelevant with that assumption. Nursing homes being prone to infection just tightens the bound.
      - zaroth 6y ago
        I think you are misreading OP. They are saying the prevalence in nursing homes was higher (“an unmitigated disaster”) which drives up the average IFR.
        mindslight 6y ago
        My comment applied to the comment you're referencing as well. The original comment is the one above that.
tinus_hn 6y ago
A useless exercise unless you also take age distribution in account. Anyway for Lombardy, the hardest hit region of Italy the answer is 0.16%.

mikem170 6y ago

>According to new estimates released by the Centers for Disease Control and Prevention (CDC), the COVID-19 infection fatality rate may be as low as 0.26%, a number that only doubles the seasonal flu but significantly lower than earlier estimates. The news has been pushing worst case scenarios at every step of the way, not doing a good job of presenting the complexities behind the facts and estimates. And lots of people seem to be blind to this. Maybe schools could be doing a better job teaching critical thinking?

pengaru 6y ago
> Maybe schools could be doing a better job teaching critical thinking? Critical thinkers don't make for easy consumers.

lmilcin 6y ago

I wish there was an assessment with error bars not spanning multiple orders of magnitude... Jokes aside. If the number was true, for Italy (33k reported deaths) this would give around 12M infected. For US (99k reported deaths) around 36M infected. Those numbers seem to be unrealistic. I remember Italy did some tests and even in the worst hit places less than 7% of population developed antibodies, not average 25% for entire Italy as suggested. I know this is very, very back of envelop calculation but it would be nice to somehow reconcile those numbers.

emeerson 6y ago
Initially I thought of those numbers (36m in the US) as totally unrealistic. But then comparing to seasonal influenza case counts, they seem to be within that order of magnitude (39m-56m cases): https://www.cdc.gov/flu/about/burden/preliminary-in-season-estimates.htm https://www.cdc.gov/flu/about/burden/preliminary-in-season-e... would that be implausible? am I misunderstanding or comparing the wrong metrics? I'm actually curious to see the time interval of when the first true COVID-19 case appeared plotted over first true 2019-2020 influenza cases. If along the same timeline, you'd figure that social distancing would have affected the R0 of both. Assuming that SARS-CoV-2 is truly more (or comparably) contagious, 36M could be plausible. To be clear: I am not equivalating COVID-19 fatality rate w/ the flu, or suggesting it is "innocuous" as the flu is.
- rusticpenn 6y ago
  But these numbers cannot be compared as a lockdown is in effect..
rurban 6y ago
Italy is one of the 7 outlier countries. It has an IFR even above 1%. But Italy is always the highest flu outlier worldwide (regularily above 1%), so there no reason to bring Italy into the discussion. The other outliers are much more interesting, like Belgium, which has also a reporting problem. UK, NL, Spain,... 0.3% has been the established number for a few months already. Reporting always concentrated on the meaningless CFR at 3.4%, for political reasons. 10x more asymptotic cases also reflect all the antibody tests. Asymptotic means an ordinary cold. That's why it's so dangerous. Low fatality rate, no symptoms, high reproduction.

op00to 6y ago

What about permanent cardiovascular damage? What’s that rate for flu and then for coronavirus?

xt00 6y ago
Yes, it would be very helpful to have the following numbers: 1) mortality rate 2) no symptom rate 3) negative outcome rate ->(some clear impact to patient who has survived, but has life impacting effects) So if you find for example that within a population of 100000 people that you have 35% with no symptoms, and 0.26% died, then 64%+ had symptoms, so then how many people had symptoms that likely are long-lasting. If for example 10% of people had long lasting impacts, then that is a far greater number than 0.26%.. so before people say "look it ain't so bad".. then we need to understand that number before people invite 100 people over to their house party.

sharemywin 6y ago

330M*.0026 = 857K. in US. Herd immunity is like 60% of that number. That's a lot of people. That could be your parent or grandparent or you. I get so tired of people comparing this to the flu. It's over taken the number of deaths from the flu in 2 months not 12. And that's while we were in locked down. Also, a whole lot of people get symptoms. I don't think having flu like symptoms for 2 weeks(not 2-3 days) sounds like fun either.

vb6sp6 6y ago
It's plain greed. The only people clamoring to "open up" always have a vested interest. Pure scum.
- samdamsamm 6y ago
  In the United States a lot of people I know are eager to get back to work because they don’t want to get evicted or fall behind on car payments and fall into the Spiraling USA Debt Trap that ruins so many lives. I don’t blame them in the least. I was laid off in March and have still not received any unemployment insurance myself. If that is still the case this time next month I will be freaking out all the same.
  - sharemywin 6y ago
    I by no means complaining about a people that have to go back to work. It just sucks that people have to make a choice between potentially risking their's or their family's lives for a paycheck. I just don't think people should be understating or under estimating this thing.
    - samdamsamm 6y ago
      So the problem is the catastrophic dismantling of the welfare state over the past several decades. Ironically, most Americans complaining about getting the economy up and running again also vote against having a basic safety net. I’d say you’re getting what you asked for. Should have at least voted for Medicare For All.

WheelsAtLarge 6y ago

The death rate is one thing but the hospitalization rate should not be ignored. We've seen the overflowing capacity in New York and other places so we need to be careful with just quoting the death rate and feeling that it's no big deal. Whatever the death rate is we know that over 100k people have died. And who knows how many thousands of survivors have dealt and will deal with the damage the virus inflicts on their bodies. We should not take that lightly. I'm one that thinks that face masks and keeping social distance is the best way to go. If we followed that without fail the transmission rate would collapse and we would be close to normal.

nikolay 6y ago
This is the biggest thing people don't get - this is not the flu and surviving it or having mild symptoms does not mean that you didn't signed up a chornic CVD or metabolic disease based on the early data that the virus causes a permanent myocardic injury, lung scaring, and God knows what else. We'll be studying the damage for the decades to come! I think many people forget that even the "benign" HPV causes different types of cancer, and we recently found it. So, a new virus means being extra careful, but how can you explain this to people who strongly believe that Bill Gates wants to chip them thru vaccines and that 5G kills birds and people?
- tinus_hn 6y ago
  You are putting up a straw man and then taking it down. Not everyone who is not panicking about COVID-19 is a conspiracy theorist.
  - WheelsAtLarge 6y ago
    It does not change the primary narrative. COVID is dangerous and highly contagious so we need to be careful and not treat it lightly.
    - tinus_hn 6y ago
      Many things are dangerous. Being careful means evaluating risks, not hysterically avoiding all risks at any cost. Smoking costs more lives than COVID-19 and it is much easier to save these lives. Yet it’s acceptable to do very little. But somehow COVID-19 warrants endless lockdowns.

cozzyd 6y ago

Considering that that's roughly the population fatality rate in NYC, and significantly less than than the population fatality rate in Bergamo province, that seems unlikely. Is the methodology available somewhere? Scenario 5 looks suspiciously like the geometric average of 0.2 and 1...

rogerkirkness 6y ago
PFR in Italy based on one of the oldest populations in the world, with the worst NOx pollution in the world because of the mountains and industry, without controlling for confounding reasons for dying (e.g. all severe cases that hit the hospital counted and all mild cases that don't aren't).
- cozzyd 6y ago
  The population fatality rate in Bergamo is based on the total number of COVID deaths (estimated from excess mortality) and the total population. See https://www.medrxiv.org/content/10.1101/2020.04.15.20067074v3 https://www.medrxiv.org/content/10.1101/2020.04.15.20067074v...

zaroth 6y ago

The population-wide total IFR could be as low as 0.26% but it would depend heavily on the demographics of the people infected. If you infect 100 people under the age of 40 your IFR will likely be much lower than even 0.26%. If just 10 of those 100 are over 80 then it’s a totally different story. Not all diseases have such a high spread in the mortality rate versus age as COVID, which spans effectively 0%-20%. It makes discussions on the average IFR less useful than they might otherwise be.

tinus_hn 6y ago

It would probably better to link to the linked page https://www.cdc.gov/coronavirus/2019-ncov/hcp/planning-scenarios.html https://www.cdc.gov/coronavirus/2019-ncov/hcp/planning-scena... Note that the risks are dramatically different for different age groups, so if one calculates a grand total this requires assumptions about age distribution. The page notes the data is about the US.

randyrand 6y ago

0.26% is well below the % of people that die every year (~1%). I wonder how many of those people would have died this year regardless. We will be able to tell in retrospect if the annual death rate is not significantly moved versus last year.

cecja 6y ago
There are studies that covid did reduce the lifespan of most of the dead by 10 years. Your argument is as retarded as your government. https://www.businessinsider.com/study-finds-that-people-who-have-coronavirus-could-die-10-years-early-2020-5?r=DE&IR=T https://www.businessinsider.com/study-finds-that-people-who-...

drinkcocacola 6y ago

Well, I've been following numbers for the last two months in different countries from different sources. My initial best estimations were that the fatality rate was about 0.77% given the numbers in South Korea (high number of tests). Other countries such Spain were reporting rates of about 10%, however testing was quite low in comparison. Recently Spain did a seroprevalence test (number of people with antibodies for Coronavirus) with a large sample of people. They found that 5% of the population had antibodies (2.4M people). Which means that the fatality rate is actually about 1.1% and not 10%. This numbers (from 0.77% to 1.16%) have been constant in different countries, so let me express my disbelief with the numbers published by the CDC.