Originally Posted by
spacediver
Regarding our current situation, here are the options I can make out.
A
We heavily prioritize minimization of total deaths. This involves keeping transmission at essentially a screeching halt until anti-virals and/or vaccine are ready.
This is likely to be at least a year, and quite possibly years (I may update my thinking on this once I learn more about vaccines).
Our society will be damaged beyond belief if this strategy is taken, and it will only be effective if everyone plays the same game.
B
We go back to "normal", and achieve rapid herd immunity. Assuming an infection CFR of 0.5 percent, this will mean about 35 million deaths worldwide, and about 1.5 million deaths in the United States.
This would be catastrophic, to be sure, and may damage our society in ways more horrific than we can imagine.
C
We contract and expand periodically, essentially allowing for periodic "blood letting". We eventually reach herd immunity. Even if anti-virals and vaccines aren't developed, less people die in total compared to B, because herd immunity protects those who haven't achieved immunity. I think you need ~70 percent of population to be immune for herd immunity, but I haven't yet gone down that rabbit hole. And the slower timescale of this strategy compared to B allows for more time for antivirals/vaccine development.
D
We find a stable middle ground where transmission is allowed, but at a smaller rate. This means that peak demand on healthcare system is not "overwhelming", but deaths will still occur, both in the obviously vulnerable, and in the smaller cohort of young healthy adults that this virus seems to capriciously target. As in C, herd immunity is eventually achieved, and more time is granted for antivirals/vaccine development.
The best strategy may differ from community to community, and here's the bitter twist:
The more information we have (about the virus, or about how many people are currently infected/already recovered), the more optimal our decision making will be. But we need time for more information, which increases the damage of what we are currently doing.
This is an optimization problem, to be sure, and it may not be an easy one.
Are there any options I've missed? Does anyone disagree with any of the premises?