5 ms·

Either there's something very unusual about your personal situation, or you're using "almost certainly" to mean "well under ten percent", or your statement is s

Either there's something very unusual about your personal situation, or you're using "almost certainly" to mean "well under ten percent", or your statement is simply false. Estimates of the household secondary attack rate vary but are typically low tens of percent[1][2], and the IFR even for patients >80 years old is about 8%[3]. The probability that your mother would die if you became infected is the product of those two. To be clear, I wear a mask and I support orders requiring that, since the cost of mask-wearing seems very small to me in exchange for the possible benefit. (If anyone finds a mask uncomfortable, I'd suggest a better mask. Surgical-style masks are much easier to breathe through than fabric, and widely available near pre-pandemic prices.) We should avoid hyperbolic, misleading, or false statements in support of a noble cause, though. 1. https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(20)30471-0/fulltext https://www.thelancet.com/journals/laninf/article/PIIS1473-3... 2. This is unintuitive given the close contact sharing a household and R0 = 2.5, but consistent with the theory that spread is driven by a small fraction of super-spreaders, while most infected transmit the disease to no one at all. 3. https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(20)30243-7/fulltext https://www.thelancet.com/journals/laninf/article/PIIS1473-3...

necovek 6y ago
You might be ignoring the fact that they know what "comorbidities" their mother has. So general population percentages might not apply to them.
- tripletao 6y ago
  I started my reply with the words "Either there's something very unusual about your personal situation...". So can you give an example of comorbidities that you wouldn't consider very unusual, and your estimate of the probability that the mother would die if the child were infected in that case? Remember that: 1. The secondary household attack rate is far less than 100%; the paper I cited found 18% for people >= 60 years old sharing the same address. So even if the child becomes infected, the mother probably wouldn't be. 2. Most eighty-year-olds have a lot of comorbidities already, so the IFR for that age group accounts for that. Assuming only that the mother is in the oldest age category from each paper and the probabilities are independent, I get a probability that the mother dies if the child is infected of 1.4%. Comorbidities can indeed drive that higher, and maybe there's some dependence (sicker people both more likely to catch the coronavirus and more likely to die from it); but I think you'll find it very hard to build a plausible set of assumptions that would get you to "almost certainly", unless the mother is so sick that she's likely to die soon regardless. (Of course, that wouldn't mean her life is without value. But it would mean that when we decide we can't do something until a vaccine is available, we're deciding she can't do it ever again in her life. If it were my life, then I'd judge that calculation differently.) Most discussion of probability of death feels rather ghoulish, and the above is no exception. All of us will die, though, and it's a question only of what we can do in the time before that. I'd rather plan that time using the best empirical estimates we can make than with half-informed hyperbole. Wouldn't you?

necovek 6y ago

You might be ignoring the fact that they know what "comorbidities" their mother has. So general population percentages might not apply to them.

tripletao 6y ago
I started my reply with the words "Either there's something very unusual about your personal situation...". So can you give an example of comorbidities that you wouldn't consider very unusual, and your estimate of the probability that the mother would die if the child were infected in that case? Remember that: 1. The secondary household attack rate is far less than 100%; the paper I cited found 18% for people >= 60 years old sharing the same address. So even if the child becomes infected, the mother probably wouldn't be. 2. Most eighty-year-olds have a lot of comorbidities already, so the IFR for that age group accounts for that. Assuming only that the mother is in the oldest age category from each paper and the probabilities are independent, I get a probability that the mother dies if the child is infected of 1.4%. Comorbidities can indeed drive that higher, and maybe there's some dependence (sicker people both more likely to catch the coronavirus and more likely to die from it); but I think you'll find it very hard to build a plausible set of assumptions that would get you to "almost certainly", unless the mother is so sick that she's likely to die soon regardless. (Of course, that wouldn't mean her life is without value. But it would mean that when we decide we can't do something until a vaccine is available, we're deciding she can't do it ever again in her life. If it were my life, then I'd judge that calculation differently.) Most discussion of probability of death feels rather ghoulish, and the above is no exception. All of us will die, though, and it's a question only of what we can do in the time before that. I'd rather plan that time using the best empirical estimates we can make than with half-informed hyperbole. Wouldn't you?