It seems like their simulated model makes certain members of a population less ‘active’ and so at less parametric risk of infection/transmission than in a model where everyone is active. This is useful, and/but doesn’t appear to be Covid-19 specific.
All this talk of herd immunity is not terribly helpful because we don’t know what “immunity” means in practical terms.<p>The duration of immunity for other corinaviruses varies quite a bit, the disease has the potential to cause permanent respiratory damage, with uncertain consequences for the possibility of reinfection, and we’re still seeing novel symptoms arise as in the case of the children with Kawasaki-like inflammation in NYC.<p>I get the strong desire to return to a normal state of affairs, but paving over reality with scientific notions, deployed pseudoscientifically, does very little to get us there.
My question is what R0 looks like in a city with mass transit vs without. The subway in NYC is huge and puts so many people together in close contact every day.<p>Is it outrageous to think R0 could be 3.5 in NYC but 1.5 somewhere else?
> The classical herd immunity level hC is defined as hC=1−1/R0<p>How do you use this in places where transmission is less than 1? New Zealand now has a transmission rate considerably under 1, but this seems to break the calculation.<p>Does this mean that a reproduction number under 1 doesn’t need herd immunity? We do, because we can’t stay locked down indefinitely, and presumably our rate will go up when we relax restrictions.
This is also true under much weaker assumptions. Here's a simple example calculation. Say 50% of a population has R0=1.5, 50% has R0=2.5; the average R0 is 2. Then the required herd immunity is (2-1/1.5-1/2.5)/2=0.47, below the classical 1-1/2=0.5. That this always true for any number of independent subpopulations is a consequence of the harmonic-arithmetic mean inequality.
Interesting that it seems to suggest there is an optimal amount of distancing/lockdown/etc. to lower the peak, but avoid a big "rebound". I don't see a lot of discussion about how to find that optimum amount.<p>It has occurred to me that the planet is essentially conducting a gigantic experiment in virus evolution, by creating a novel viral selection environment never seen before. For example, how does it impact a virus if the strains most likely to send the host to the hospital, actually spread more than the strains most likely to be asymptomatic, because even asymptomatic hosts are in isolation?<p>Unless there is a vaccine or cure coming soon (there isn't), then the end result must be herd immunity, no matter what the policy. It would seem that we should be considering how to select for the healthiest segment of the population to provide that herd immunity, rather than trying to isolate as much as possible in the vain hope that it will die out prior to that point. If that was every possible, it has long since spread way too far to expect that to be possible.<p>There might be cases (e.g. MERS) where you can stomp out a virus before it spreads enough to cause herd immunity, but we have long since passed that point. I don't think most people (including policy makers) are thinking about the endgame properly (as this paper does).
So, when active spreaders do most of the spreading, when they have herd immunity, there is a disproportionate affect on the overall spread.<p>Meaning, let the young and stupid go out and play, it will allow for herd immunity at a lower rate than the typical randomly distributed rate.