Thread by @uk_domain_names

THREAD
1/ We know that hospital admissions correlate closely with COVID-19 deaths, but it's worth reminding ourselves just how much that's true...

Here's what happens when you plot admissions vs deaths for England, with deaths shifted by 14 days (people don't die immediately).

2/ Even though the two sets of data are plotted on different scales, you can see they track each other almost perfectly, with the exception of a period during the first wave in which deaths outstripped the long-term mortality ratio. (That's almost certainly due to under-testing.)

3/ Using the same data, we can get a crude estimate of the mortality rate, i.e. what percentage of people entering hospital and testing positive for COVID-19 will subsequently die in hospital.

(For comparison, influenza has a mortality rate around 5% for hospitalised cases.)

4/ As you can see, overwhelming hospitals doesn't just result in more deaths, but also disproportionally worse outcomes.

When they had few patients last summer, a much higher percentage of them survived. But now that resources are stretched, a higher percentage are dying again.

5/ In other words, if you can keep hospitals relatively empty, even those people affected seriously enough to need hospitalisation have a good chance of survival.

But if the infection gets out of control, then even in hospital each patient's chances dwindle signficantly.

6/ DATA SOURCES

NHS England COVID-19 all announced deaths
england.nhs.uk/statistics/sta…

Coronavirus Dashboard patients admitted to hospital (by nation)
coronavirus.data.gov.uk/details/health…

/END