I just finished reading this paper and it's absolutely terrifying. They predict 500,000 deaths in the UK and more than 2 millions in the US in the absence of any intervention to mitigate the epidemic. They estimate the death toll could be halved by adopting very restrictive 1/n
interventions, but they would have to be maintained for most of the period until a vaccine is available, which is probably going to take at least 18 months. But note that their estimate of the death toll is based on the assumption, which they know to be completely false, that 2/n
the fatality rate will not increase due to the overwhelming of ICU beds capacity, even though their simulations predict that, even if the optimal mitigation strategy is adopted, it will result in an *8-fold* higher peak demand on critical care beds over and above the 3/n
available surge capacity in both GB and the US. The assumptions they make, especially about compliance vis-à-vis the restrictive measures, strike me as a bit pessimistic, as the authors themselves acknowledge, but not crazy. They do a sensitivity analysis which shows that the 4/n
results are robust to various assumptions about the reproduction rate of the virus and other parameters of the model. Of course, the model could be off, though apparently this kind of model has been successfully used in the case of influenza epidemics, but even if it 5/n
overstates the ICU beds capacity required by a factor of 2 or even more than that, it means that a ridiculous number of people are going to die even if we adopt extremely restrictive measures that will destroy the economy. I though I knew how bad this was, but I never 6/n
imagined it could be *that* bad. I will probably write another blog post on this once I have read he papers that describe the kind of models used more fully, though it will be in French. I strongly encourage everyone to read this paper right now and talk about it. 7/7
