The difference between the changes in COVID-19 patients in hospital and new COVID-19 hospitalisations is really odd & I'd be interested in any explanations anyone had. In the NW patients in hospital look to have grown near-exponentially yet hospitalisations look... pretty flat? https://twitter.com/BristOliver/status/1308085968117936139

Please do not @ me with some stupid point about the use of the word "exponential". I promise I know what it means.

FWIW I think that *so far* the hospitalisations series has been much more robust with the in-hospital series subject to lots of revisions but my explanation for this is "??????????". I have no idea if this is still the case now.

in response to two replies: hospitalisations should include in-hospital diagnoses

I think there's a day of week effect on hospital cases. Mondays look to have v high jumps & as a result there's one instance of a negative implied discharges+deaths in the last month.