Helen MacDonald

Well, I think it's just ironic that you have this public policy of lockdown, which everyone is strictly adhering to.

And it's an incredibly prohibitive and has all kinds of damaging consequences for the economy.

But the irony would be that if in doing that policy, you haven't actually shielded and protected the very people that that policy was...

in many ways, designed to protect those most vulnerable who are a large section of the population.

Yes, death data has been something which has been fascinating us for a little while on this issue.

And it feels like you can't really read anything online without knowing how many people have died in the last 24 hours.

We had some headlines in the UK suggesting a large rise in the death rate last week, and with as many as one in five being linked to coronavirus.

But

It's hard to put that in context.

How many people do we expect to die each day?

Is coronavirus killing people who were already dying?

In other words, are people dying from the virus or just sort of with the virus of other things?

And with so much uncertainty in the models that have been used for public health policy decision making, there have been legitimate concerns about whether we might have either

overestimated or underestimated the threat posed by the virus.

So who better to talk to, we thought, than statistician David Spiegelhalter from the University of Cambridge, who's an expert in risk and evidence communication.

And what do we make of that?

Do you think that data exists?

Do you think that that spike will reassure the government advisors in some way that they have got the measures right, that we're starting to see that the virus is causing excess death?

And there was this concern that perhaps until we started to see something like that, was the lockdown worth it with all the harm that might come downstream?

So what you're saying is that, in effect, these different data sources that you could look at on death, the dates of when those deaths are occurring is important.