5 ms·

I think the article groups 20-54 because some people believe that only older people are much affected by the disease. They picture hospitals filled with the el

I think the article groups 20-54 because some people believe that only older people are much affected by the disease. They picture hospitals filled with the elderly. However, the data shows here that if you visited a US hospital filled with these COVID-19 patients that close to 40% of them would actually be younger people. The article is not claiming that 20-54 year olds with the disease have a 40% chance requiring hospitalization. The stats quoted also do not show that the 20-44 or 45-54 cohorts have a 20% chance of requiring hospitalization. Instead, it's saying that the 20-54 cohort makes up close to 40% of the people who require hospitalization. I.e. I think you have the probability conditional backward. It's not P(need hospital|age 20-54) ~= 40%. Instead it's P(age 20-54|need hospital) ~= 40% The point is that it's not only the elderly who are affected here. They may be most at risk, and most likely to die even with hospitalization, but even younger people get sick enough to require hospitalization, without which the fatality rate of that cohort may rise as well. If we don't "flatten the curve" then adults of all ages may not get the hospitalization they need and start dying. (edit: corrected age bracket from 20-64 to 20-54)

YeGoblynQueenne 7y ago
>> The article is not claiming that 20-64 year olds with the disease have a 40% chance requiring hospitalization. The article subtitle says: "nearly 40 percent of patients sick enough to be hospitalized were aged 20 to 54." >> I.e. I think you have the probability conditional backward. It's not P(need hospital|age 20-64) ~= 40%. Instead it's P(age 20-64|need hospital) ~= 40% The article is not making any such formal distinction.
- ddlatham 7y ago
  Your comment asked: >> Would it weaken the arcile's point to report that 20-44 year olds and 45-54 year olds both had around 20% chance of being hospitalised? But I think that's unfair to the article. The data are not showing that 20% of age 20-44 are hospitalized (which is how I read your question). Rather the data say that 20% of the hospitalized are age 20-44. In fact, the article does report that explicitly in the 6th paragraph: In the C.D.C. report, 20 percent of the hospitalized patients and 12 percent of the intensive care patients were between the ages of 20 and 44, basically spanning the millennial generation. So the article is being fair to the report, it is just summarizing the report data in the lede.
  - YeGoblynQueenne 7y ago
    >> The data are not showing that 20% of age 20-44 are hospitalized (which is how I read your question). That is clearly not what I say. I quoted the CDC report that clarifies that a) 20% of those hospitalised were in the age range 20-44 and 18% in the range 45-54 and that b) that is only those for whom there is age data, which were only a small proportion of all confirmed cases. Please give me the benefit of the doubt that I know how to read English. At the very least I must have read and understood the passage I quoted! The article is pulling the CDC data this way and that and tries to argue that "younger adults make up [a] big proportion of coronavirus hospitalisations from the US", but there is no reason to base such a claim on the CDC data and there is no such claim in the CDC report. The Times article simply mashed up some of the data in the CDC report to make a sensational title. My comment suggests that, if they want to attach a probability to the hospitalisation of different age ranges, they should not try to aggregate the CDC age ranges just to get to that "almost 40%" and instead use the best approximation that can be derived from the CDC data, which is about 20% for those in each age range, 20-44 and 45-54. And you are pulling my comment this way and that to show ... I'm not sure what. Do you agree with the article, that "younger adults make up [a] big proportion of coronavirus hospitalisation from the US"? If not, what exactly are you disagreeing with in my comment?

YeGoblynQueenne 7y ago

>> The article is not claiming that 20-64 year olds with the disease have a 40% chance requiring hospitalization. The article subtitle says: "nearly 40 percent of patients sick enough to be hospitalized were aged 20 to 54." >> I.e. I think you have the probability conditional backward. It's not P(need hospital|age 20-64) ~= 40%. Instead it's P(age 20-64|need hospital) ~= 40% The article is not making any such formal distinction.

ddlatham 7y ago
Your comment asked: >> Would it weaken the arcile's point to report that 20-44 year olds and 45-54 year olds both had around 20% chance of being hospitalised? But I think that's unfair to the article. The data are not showing that 20% of age 20-44 are hospitalized (which is how I read your question). Rather the data say that 20% of the hospitalized are age 20-44. In fact, the article does report that explicitly in the 6th paragraph: In the C.D.C. report, 20 percent of the hospitalized patients and 12 percent of the intensive care patients were between the ages of 20 and 44, basically spanning the millennial generation. So the article is being fair to the report, it is just summarizing the report data in the lede.
- YeGoblynQueenne 7y ago
  >> The data are not showing that 20% of age 20-44 are hospitalized (which is how I read your question). That is clearly not what I say. I quoted the CDC report that clarifies that a) 20% of those hospitalised were in the age range 20-44 and 18% in the range 45-54 and that b) that is only those for whom there is age data, which were only a small proportion of all confirmed cases. Please give me the benefit of the doubt that I know how to read English. At the very least I must have read and understood the passage I quoted! The article is pulling the CDC data this way and that and tries to argue that "younger adults make up [a] big proportion of coronavirus hospitalisations from the US", but there is no reason to base such a claim on the CDC data and there is no such claim in the CDC report. The Times article simply mashed up some of the data in the CDC report to make a sensational title. My comment suggests that, if they want to attach a probability to the hospitalisation of different age ranges, they should not try to aggregate the CDC age ranges just to get to that "almost 40%" and instead use the best approximation that can be derived from the CDC data, which is about 20% for those in each age range, 20-44 and 45-54. And you are pulling my comment this way and that to show ... I'm not sure what. Do you agree with the article, that "younger adults make up [a] big proportion of coronavirus hospitalisation from the US"? If not, what exactly are you disagreeing with in my comment?