Re:Scientists create antibody that defeats coronavirus in lab
by mlyle on Monday May 04, 2020 @11:16PM
(#23076095)
> About 21% percent of New York City was testing positive for antibodies.
From samples returned 10 days ago, so add a bit more lag.
It's not beyond the realm of possibility that New York City has reached the herd immunity threshold. The biggest estimates of variance in contact networks and susceptibility (from fitting to case count data) predict a herd immunity threshold of ~15%. (Cited above).
> We don't seem to be able to ramp up testing
Testing, shmesting. It only provides a minimal adjustment to Rt; sensitivity isn't great, and you can't test everyone every day.
> we haven't gotten the infection rate down to the point where contact tracing would work
Nor will we. The most aggressive lockdowns produce case count decays of 40% or so per week-- possibly less. The suggested threshold for where contact tracing can work well is 1 infection per million population. The SF Bay Area probably has about 3500x that. -ln(3500)/ln(0.6) =~ 16 weeks.
IHME was predicting the virus would magically go away in California with continued shelter in place in the next 3 weeks. They just backed away from this crazy prediction that the data has never fit. However, policymakers depended upon it in making the tradeoff: aggressive policies aimed at control could be justified if they'd really get us to a sustainable containment regime in a couple months.
> And significant portions of the US seem to be giving up on control and going for a herd immunity strategy.
In the end, the only long term paths are to radically change our way of life forever, or herd immunity. There's only really one vaccine effort that could deliver tens of millions of doses to North America by mid-2021, and that's the Johnson & Johnson program. Odds of that program's success are 50-50 at best.
From samples returned 10 days ago, so add a bit more lag.
It's not beyond the realm of possibility that New York City has reached the herd immunity threshold. The biggest estimates of variance in contact networks and susceptibility (from fitting to case count data) predict a herd immunity threshold of ~15%. (Cited above).
> We don't seem to be able to ramp up testing
Testing, shmesting. It only provides a minimal adjustment to Rt; sensitivity isn't great, and you can't test everyone every day.
> we haven't gotten the infection rate down to the point where contact tracing would work
Nor will we. The most aggressive lockdowns produce case count decays of 40% or so per week-- possibly less. The suggested threshold for where contact tracing can work well is 1 infection per million population. The SF Bay Area probably has about 3500x that. -ln(3500)/ln(0.6) =~ 16 weeks.
IHME was predicting the virus would magically go away in California with continued shelter in place in the next 3 weeks. They just backed away from this crazy prediction that the data has never fit. However, policymakers depended upon it in making the tradeoff: aggressive policies aimed at control could be justified if they'd really get us to a sustainable containment regime in a couple months.
> And significant portions of the US seem to be giving up on control and going for a herd immunity strategy.
In the end, the only long term paths are to radically change our way of life forever, or herd immunity. There's only really one vaccine effort that could deliver tens of millions of doses to North America by mid-2021, and that's the Johnson & Johnson program. Odds of that program's success are 50-50 at best.