David Spiegelhalter

How many cases, what proportion of these extra is that sort of case?

And this is very important because this could influence what people feel about releasing the lockdown and so on.

People are going to be arguing about this for ages.

trying to work out what sort of analyses, what sort of data would help tell you to try to separate between those things.

Very difficult.

And so, you know, people have got ideas, which we're already starting to look at.

For example, if it's due to lockdown, that should be uniform across the whole country, you know, because everyone's experiencing the same impact.

So that should be a uniform effect.

Whereas if it's unregistered COVID deaths, that should tend to occur in hotspots of COVID, you know, in West Midlands and London and so on.

At the moment, we're only getting regional data, very broad areas.

If we could have regional data

data in much finer areas.

So you could look explicitly at areas in which there are some areas of the country with very little COVID indeed,

And if they are having a rise in non-COVID deaths, then that would suggest it's due to the intervention rather than perhaps the disease.

But it's extremely difficult to tease this out, partly because the COVID deaths, we know who those are.

Those have got names.

We can look at their comorbidities and so on.

But the non-COVID deaths, this excess of 2,500, of course,

We don't know who those are.

We've got to look at the whole 12 and a half thousand per week of non-COVID.