5 ms·

From my analysis, it appears all age groups are going to be somewhere between 50%-100% more likely to die in 2020 than any prior year. https://austingwalters.c

From my analysis, it appears all age groups are going to be somewhere between 50%-100% more likely to die in 2020 than any prior year. https://austingwalters.com/u-s-covid19-less-tests-more-deaths-no-end-in-sight/ https://austingwalters.com/u-s-covid19-less-tests-more-death... I think that's a better metric because it puts it in perspective. If you're 30, a 50% increased risk of death is still very very low. But if you're 80+ that risk is pretty substantial. Note: this is basically extrapolating the New York and New Jersey numbers, which is assumed had the highest COVID19 penetration in the U.S.

dcolkitt 6y ago
One of the most notable things about Covid is that the IFR seems to be closely proportional to baseline all-cause mortality for most groups. It's obviously not exact, so it's hard to tell whether this is just a case of coincidence and squinting to find a pattern in the noise. But still, I think there's potentially something deep going on that may tell us more about what mediates Covid lethality.
cma 6y ago
Based on New York isn't that off by 5 or so? Around 1% die each year, in New York State around .17% of coronavirus (confirmed/probable), and more likely ~.2% from coronavirus (excess deaths). To double normal deaths, reaching 2% dead for the year, it would need to be 5X higher excess deaths. There is still a quarter of the year left, but it doesn't seem enough. I think it would be closer to doubling everyone's mortality risk only in the zero mitigation scenario.
rubidium 6y ago
“ this is basically extrapolating the New York and New Jersey numbers” That’s a poor measure for a number of reasons. Not least of which is we’re getting a lot better at treating Covid medically.
- lettergram 6y ago
  I don’t think that’s necessarily true. We don’t have enough data to support that claim, probably won’t for years. Very few peer reviewed treatments have been socialized. Most evidence is that things “show promise” then end up being less effective after actual analysis.
  - chomp 6y ago
    Peer reviewed treatments, sure. But case studies have been widely shared between medical teams. Things like tummy time, steroids for people on ventilators, anti clotting agents, ventilator strategies and remdesivir are all things we've learned since the outbreak started. These aren't ground breaking cure all treatments, but they definitely moved the needle.
  - samatman 6y ago
    That's rich, coming from someone who just predicted a 75% uptick in all-cause mortality with 25% error bars!
mxcrossb 6y ago
> I think that's a better metric because it puts it in perspective. It’s interesting that you see it this way. Personally if someone said to me I have a 50% increased chance of death, I’d really struggle to make a proper risk assessment. Maybe it’s just hard to talk probabilities about ones own mortality
kalleboo 6y ago
> I think that's a better metric because it puts it in perspective It's a classic tabloid shock headline to write "XXX will increase your risk of cancer by 10 times!" (when your base risk of cancer is low). So whenever I see a statistic phrased like you did, I assume it's bullshit.

dcolkitt 6y ago

One of the most notable things about Covid is that the IFR seems to be closely proportional to baseline all-cause mortality for most groups. It's obviously not exact, so it's hard to tell whether this is just a case of coincidence and squinting to find a pattern in the noise. But still, I think there's potentially something deep going on that may tell us more about what mediates Covid lethality.

cma 6y ago

Based on New York isn't that off by 5 or so? Around 1% die each year, in New York State around .17% of coronavirus (confirmed/probable), and more likely ~.2% from coronavirus (excess deaths). To double normal deaths, reaching 2% dead for the year, it would need to be 5X higher excess deaths. There is still a quarter of the year left, but it doesn't seem enough. I think it would be closer to doubling everyone's mortality risk only in the zero mitigation scenario.

rubidium 6y ago

“ this is basically extrapolating the New York and New Jersey numbers” That’s a poor measure for a number of reasons. Not least of which is we’re getting a lot better at treating Covid medically.

lettergram 6y ago
I don’t think that’s necessarily true. We don’t have enough data to support that claim, probably won’t for years. Very few peer reviewed treatments have been socialized. Most evidence is that things “show promise” then end up being less effective after actual analysis.
- chomp 6y ago
  Peer reviewed treatments, sure. But case studies have been widely shared between medical teams. Things like tummy time, steroids for people on ventilators, anti clotting agents, ventilator strategies and remdesivir are all things we've learned since the outbreak started. These aren't ground breaking cure all treatments, but they definitely moved the needle.
- samatman 6y ago
  That's rich, coming from someone who just predicted a 75% uptick in all-cause mortality with 25% error bars!

mxcrossb 6y ago

> I think that's a better metric because it puts it in perspective. It’s interesting that you see it this way. Personally if someone said to me I have a 50% increased chance of death, I’d really struggle to make a proper risk assessment. Maybe it’s just hard to talk probabilities about ones own mortality

kalleboo 6y ago

> I think that's a better metric because it puts it in perspective It's a classic tabloid shock headline to write "XXX will increase your risk of cancer by 10 times!" (when your base risk of cancer is low). So whenever I see a statistic phrased like you did, I assume it's bullshit.