A very important point has been made on Twitter.
Since yesterday, they have been reporting the number of deaths of those hospitalised. Prior to that they didn't mention hospitalisation, only deaths of those tested positive...
I hope this isn't a deliberate skewing of the data.
The difference is probably quite small, you'd imagine that the vast majority of people who've died have been admitted, though, as time goes on and the system gets more stretched, the % of people who die at home will likely increase.
Eventually, the numbers of people who died without being admitted will filter through the system ( PM, Cause of death registered etc ).
We won't know the full story for months though. Though horrible, the numbers we're seeing
at the moment aren't going to make much difference to the overall number of "expected" deaths for the period. Harsh though it sounds, some people with COVID-19 would have died in the near future anyway ( plainly that's no consolation to their families, but it's true ).
There's a perception among many that the numbers we see should be 100% accurate, but that's naive. There's always going to be grey areas.
In any given year, you'd expect 550 to 600k deaths per year. The overall aim has to be to have a number as close as possible to what we'd normally expect.