Thread by @drsanjaygupta

There has been a lot of talk recently about how many people will be infected by coronavirus – and in particular, how many people could die. One thing people have been focusing on in particular: models. Here's why they matter. (1/14)

Models aren't perfect. They use the past to predict the future, which means they're only as good as the assumptions – and data – that are fed into them. But the White House said Sunday that models helped officials decide to extend social distancing through the end of April.(2/14)

One model that people have been talking about lately is from @IHME_UW at the University of Washington. It assumes that social distancing measures will remain in place through May, but its predictions are still grim: about 82,000 people could die by August. (3/14)

A lot of these deaths will happen in places hit early by the virus – in other words, places that haven't as much time time to prepare, and where social distancing measures may have come too late. (4/14)

New York, for example, is estimated to see 15,500 deaths by August – and will see its peak count of daily deaths in ten days. California, with millions more residents, is expected to peak more than three weeks from now, and is projected to have 4,300 deaths by August. (5/14 )

If you want to see what the model predicts for your state - you can use the dropdown at the top of the page. (6/14)
covid19.healthdata.org/projections

Any loss of life is tragic – and tens of thousands of deaths are especially painful to see. But there are things we can do now to save lives, and to ensure that things don't get worse. (7/14)

For starters, the model was based in part on data from Wuhan, China, where officials were far more aggressive in their lockdown than the US has been. (8/14)

The predictions assume that social distancing measures in the US could be as effective as China's, but the researchers note that "it will take time to evaluate whether social distancing adherence is fundamentally different in the US compared to Wuhan." (9/14)

It's incumbent on us to follow every last recommendation – even if they're not legally binding – so that we have the best chance of success. If tens of thousands of people could die *with* social distancing, imagine how many could die *without* (10/14)

In states that are expected to see their "peak" later on, officials can take other steps too: expanding hospital capacity before it's too late, ramping up production of supplies and creating the infrastructure to handle a possible surge in cases. (11/14)

The IHME model, for example, estimates that New York will face a shortage of more than 58,000 hospital beds in less than two weeks. Other states need to act now to ensure that they don't face a similar crisis. (12/14)

While some states – like California – are predicted to have enough capacity at the moment, the model is updated regularly. A budding outbreak could change the state's trajectory, especially if people don't stay home. (13/14)

Models can be scary. 82,000 deaths is a lot. But it's important to be honest and realistic – and know that, if we don't keep our distance and wash our hands, things could get even worse. (14/14)