3 ms·

Or more simply, 5x as contagious but 1/2 as deadly? That's still a lot more deaths.

by aqme28 5y ago

kbelder 5y ago
Good chance it means more deaths sooner, but fewer total deaths.
oneoff786 5y ago
No it’s not. Contagion spread is logarithmic, and has obvious upper bounds. E.g. about a third of the us has gotten Covid. 5x contagious does not give you 166% of the population infected. Nor does it give you 100%. Probably not even 90%. I would bet no more than 50% over the next year. Half as deadly on the other hand is a big deal. It’s not just deaths, it’s also hospitalizations. If you imagine the bell curve of outcomes, the deaths are the small tail on the right and the hospitalizations are the fatter bit on the right. If you get to move everything to the left a bit, you remove the biggest chunk of the population from getting hospitalized in the first place.
- renaudg 5y ago
  It sounds like you don't know the first thing about epidemiology modelling : https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6962332/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6962332/
  - oneoff786 5y ago
    Read your own damn paper. The exponential phase of a disease spread is at the very beginning when the total number of cases is small. The rate rapidly decays to a point of being mostly linear and continuing to drop off. People think exponential is an accelerating rocket ship to 100% infected. It is not. It’s a decelerating trend likely to stall out well below 100%. If the R factor increases from 1 to 2, over a LONG period, you would be wrong to expect exponentially more cases. You would probably see fewer than 2x the cases. You would see it expand much, much faster initially. But both would see slow downs at similar inflection points. The very first sentence of your own linked abstract: > The INITIAL exponential growth rate of an epidemic is an important measure of the severeness of the epidemic, and is also closely related to the basic reproduction number. Estimating the growth rate from the epidemic curve can be a challenge, because of its DECAYS with time. the long term model is not exponential. This really should be obvious, but you cannot create an accurate disease model that does not show the impact of population size. Pure exponential models are only relevant on a theoretical infinitely large population that is well and randomly mixed. When you establish that there’s only so many people, the growth rate has to decline, and not at 100% of the population. It stalls out well before that. Look into random graph models.

Or more simply, 5x as contagious but 1/2 as deadly? That's still a lot more deaths.

by aqme28 5y ago

Or more simply, 5x as contagious but 1/2 as deadly? That's still a lot more deaths.

kbelder 5y ago

Good chance it means more deaths sooner, but fewer total deaths.

oneoff786 5y ago

No it’s not. Contagion spread is logarithmic, and has obvious upper bounds. E.g. about a third of the us has gotten Covid. 5x contagious does not give you 166% of the population infected. Nor does it give you 100%. Probably not even 90%. I would bet no more than 50% over the next year. Half as deadly on the other hand is a big deal. It’s not just deaths, it’s also hospitalizations. If you imagine the bell curve of outcomes, the deaths are the small tail on the right and the hospitalizations are the fatter bit on the right. If you get to move everything to the left a bit, you remove the biggest chunk of the population from getting hospitalized in the first place.

renaudg 5y ago
It sounds like you don't know the first thing about epidemiology modelling : https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6962332/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6962332/
- oneoff786 5y ago
  Read your own damn paper. The exponential phase of a disease spread is at the very beginning when the total number of cases is small. The rate rapidly decays to a point of being mostly linear and continuing to drop off. People think exponential is an accelerating rocket ship to 100% infected. It is not. It’s a decelerating trend likely to stall out well below 100%. If the R factor increases from 1 to 2, over a LONG period, you would be wrong to expect exponentially more cases. You would probably see fewer than 2x the cases. You would see it expand much, much faster initially. But both would see slow downs at similar inflection points. The very first sentence of your own linked abstract: > The INITIAL exponential growth rate of an epidemic is an important measure of the severeness of the epidemic, and is also closely related to the basic reproduction number. Estimating the growth rate from the epidemic curve can be a challenge, because of its DECAYS with time. the long term model is not exponential. This really should be obvious, but you cannot create an accurate disease model that does not show the impact of population size. Pure exponential models are only relevant on a theoretical infinitely large population that is well and randomly mixed. When you establish that there’s only so many people, the growth rate has to decline, and not at 100% of the population. It stalls out well before that. Look into random graph models.