Sorted by New

Wiki Contributions


Many have probably seen the threads (e.g., here and here, and this Medium post by Yishan Wong) about how massively scaled up testing is the key to threading the needle between economic collapse (long-term suppression tactics) and unacceptably high mortality (mitigation only, or worse, doing nothing). Aggressive and scaled up testing infrastructure is the key enabler for contact tracing, which appears to be the cornerstone of the South Korea model, which notably does not rely on draconian suppression tactics like lockdown, and they are currently the only other country than China (that I know of) that has R0 < 1.

I'm now wondering if, in the spirit of wartime efforts, more people can be trained to help power this testing infrastructure than can be trained to help increase our healthcare capacity.

Read the Imperial College COVID-19 Response Team report tonight.

The numbers are quite starkly grim, based on an epidemiological simulation model. They conclude that mitigation strategies (only isolating symptomatic people, social distancing only at-risk people) will at best reduce the load on the healthcare system to "only" 8x current surge capacity in UK/US, leading to estimated 1.1M deaths from COVID-19 alone (i.e., not considering possible deaths from other causes due to an overloaded healthcare system). Instead, suppression strategies (everyone socially isolating) need to be followed for 12-18 months to ensure that the load from COVID-19 stays within surge capacity, minimizing total deaths to the low hundreds of thousands, and buying time for a vaccine / treatment to help beat back a second epidemic after relaxing suppression measures.

I am curious about a few assumptions in their model (wish it was open-sourced!), and how this might change the estimate.

1) They assume a fixed ICU capacity (where I think the main limit is ventilators, not just beds). Does anyone have any models/estimates of how much UK/US can expand intensive care capacity (e.g., with "wartime" style all hands on deck manufacturing and innovation)?

2) I don't see any modeling of social network effects in the mitigation scenarios. I'm thinking of the conjecture in this preprint ( that intergenerational interactions, co-residence, and commuting patterns in Italy might have unique effects on the transmission and mortality rates.

Has anyone seen discussion of these points from experts? Or done some modeling themselves that could speak to these issues?

Also curious what others think of the report.