Second thread of the evening – I was invited onto @MSNBC earlier this evening to talk modeling of infectious disease. I got bumped by the @WhiteHouse press briefing running over. Here are a few of the things I was planning to say after watching it 1/n
First, at some point Tony Fauci made the comment that ‘models are only as good as their assumptions’ or similar and apologized for dissing his modeling colleagues. Tony, your modeling colleagues agree. You're good 2/n
Looking at the models referred to in the press briefing: the @IHME_UW modelmakes very strong assumptions based on the previous outbreaks elsewhere. This sounds reasonable, but if we had only seen Wuhan, we wouldn’t have gotten Italy right 3/n
Worse, the curves presented by @IHME_UW assume that Wuhan style reductions will be achieved, with less effort than there. This might be achievable, but there’s not actually any evidence for it 4/n
These curves are also just saying "we think things will go up and then down like they have elsewhere" with no mechanistic reason. The results are overly optimistic, and should only be used to alert people to the threat, not to determine the effort necessary to respond to it 5/n
Next up, the @imperialcollege model which the president spent some time talking about. It estimates over 2 million deaths in the US if we do nothing. This is a model which factors in estimates of transmission rates and deaths, and permits us to game out interventions 6/n
The @GOVUK who I have dissed elsewhere in this account, nevertheless changed their overall policy to be stronger after this group updated their model TWO WEEKS ago. That report included details of the expectations for the US (find it and others here) imperial.ac.uk/mrc-global-inf… 7/n
The current policy of the FedGovt has improved, though it is still weak and should pressure some states more. sounds like people are starting to realize that this is a #pandemic; just about the most serious threat we face as human beings. This is the start of a long fight 8/fin
