Read on Twitter
1. Can we restart the economy with a two-day work week?
A short thread about the new #COVID19 proposal from a group of top systems biologists:
https://medium.com/@urialonw/containing-sars-cov-2-with-a-two-day-workweek-fbdea4030d30
It's great to have all sorts of ideas out there. I like this paper because they're thinking outside of the box.
A short thread about the new #COVID19 proposal from a group of top systems biologists:
https://medium.com/@urialonw/containing-sars-cov-2-with-a-two-day-workweek-fbdea4030d30
It's great to have all sorts of ideas out there. I like this paper because they're thinking outside of the box.
2. They're not attacking epidemiologists or playing stupid games with data. Just the simplest underlying models needed to explore a straightforward proposal:
Let's restart the economy with a 2-day work week, and stay at home the other 5 days of the week to reduce transmission.
Let's restart the economy with a 2-day work week, and stay at home the other 5 days of the week to reduce transmission.
3. Cyclic, scheduled closure makes a ton of sense, both from control purposes (control is hard!) and from scheduling purposes.
4. The problem is, I don't think it will work.
5. The model manages to just get R down below the critical threshold of 1. There is very little margin for error.
6. The estimate of R=2.4 during active days seems reasonable.
What I'm concerned about is the estimate of R=0.3 for lockdown days. (It's attributed to the Imperial College study but I'm not seeing it on a quick re-read; can someone point it out to me? https://www.imperial.ac.uk/media/imperial-college/medicine/sph/ide/gida-fellowships/Imperial-College-COVID19-NPI-modelling-16-03-2020.pdf)
What I'm concerned about is the estimate of R=0.3 for lockdown days. (It's attributed to the Imperial College study but I'm not seeing it on a quick re-read; can someone point it out to me? https://www.imperial.ac.uk/media/imperial-college/medicine/sph/ide/gida-fellowships/Imperial-College-COVID19-NPI-modelling-16-03-2020.pdf)
7. To my knowledge, the best estimate of R during lockdown comes from the Chaolong Wang study of 25,000+ Wuhan cases. https://www.medrxiv.org/content/10.1101/2020.03.03.20030593v1
This study estimates that R dropped to 0.32 after lockdown.
This study estimates that R dropped to 0.32 after lockdown.
8. But that paper also estimates a much larger value of =3.86 before lockdown. I think this is anomalously high compared to other studies, but it is concerning that the two-day work week scheme wouldn't work under the Wang et al. parameters that give us R=0.3 for lockdown.
9. My bigger concern is that even if the R=2.4 estimate is right for active days, I am dubious that we in the West can get anywhere near Wuhan's R=0.3 on lockdown days, especially if interspersed with active work days. And if we don't, the two-day work week allow epidemic growth.
10. I suppose there is always the 1-day week. That would work even with R=0.7. As a teenager I thrilled at Philip José Farmer's Dayworld trilogy where people lived one day a week for the good of the collective. This could be our chance to embrace that dystopian life.
