The IHME coronavirus model now 'predicts' that peak hospital use in NJ happened on April 8, i.e. 4 days ago. Just one small problem: the data published by the state show that hospital use did not, in fact, peak on that date.

Hospital use *may* have peaked yesterday (April 11) at 7,618 patients. Too soon to be sure. And arguably IHME model was 'close enough'. But the prediction it was making in its April 10 update was inconsistent with data that had already been published. Use models with caution!!!

Here is a very nice representation of the New Jersey COVID-19 hospitalization data from @JPVMan: https://twitter.com/JPVMan/status/1249430315245998091?s=20