The data released by the Office for National Statistics (ONS) gives us the best indication so far about how many lives are being lost to Covid-19 across England and Wales.
These are sobering statistics, setting new and unwanted records.
In the week ending 3 April, there were about 6,000 more deaths registered than in an average week and nearly 3,500 of them mentioned Covid-19 on the death certificate.
And when you dig down into the data at regional levels, some of the numbers look even worse.
They show how many lives have been lost in London, for example, where the coronavirus pandemic first spread rapidly in the UK.
Until the latest figures came out, the highest number of registered deaths per week in London – since the ONS started keeping weekly records in 2005 – was 1,549. That was back in January 2015.
For the week ending 3 April this year the number was 2,511. That’s 62% more than the previous record high, and more than double the average of 1068 registered deaths for this week of the year.
And it happened well before the outbreak reached its current levels.
The chart of deaths in London shows just how high the new figure is compared with an average weekly number of deaths over a five year period.
The number of deaths varies from week to week, but for most of the year the variations are relatively small – it is only slightly higher or lower than the average.
Generally, there will be only one or two weeks in the year when there are significant variations, usually in the middle of winter, when seasonal flu is at its height.
That the figure was already so far outside its normal range three weeks ago shows the scale of the threat that Covid-19 poses.
It’s also worth remembering that the ONS relies on death registrations (in hospital and in the wider community in places such as care homes) for its figures.
It suggests there is a time lag of about five days from when deaths occur, to when they are actually registered. So, most of these deaths actually occurred in the last week of March.
The figures for the Midlands (combining ONS data for West Midlands and East Midlands) also show a significant spike in the most recent figures.
Once again, the figures for the Midlands are the highest per week since records began, but they are only a little higher (3,058 registered deaths compared with 3,041) than the next highest week in January 2015. That was at the height of the worst winter flu outbreak in recent years.
Other regions in England and Wales show that there are still weeks in January 2015 and January 2018 when the number of registered deaths was slightly higher than the most recent figures.
The North-West for example had its second worst week for registered deaths – 2,137 in the week ending 3 April compared with 2,282 in the week ending 9 January 2015.
There were 920 deaths registered in Wales, compared with 1,031 in the week ending 9 January 2015.
It is worth remembering that other regions are behind London in the curve of the pandemic, and figures released next week are likely to be worse.
“We can see that London and the Midlands are seeing the effects of the epidemic ahead of the rest of the country,” said Dr Sarah Deeny, assistant director of data analytics at the Health Foundation.
“Unfortunately, we can expect the total number of deaths related to Covid-19 recorded by the ONS to increase further in coming weeks across England and Wales.”
The ONS says 47% of all the deaths registered in London mentioned Covid-19 on the death certificate, including a considerable number in the community. What is not clear yet is how many of the other 53% might have been linked to the disease.
One of the biggest issues for policymakers over the coming weeks will be to find out what is causing all these excess deaths.
We know many of them are deaths from Covid-19, and that they will continue to happen for a few weeks despite the lockdown.
But it will also be vital to establish how many deaths may be happening because of the lockdown, if people are not getting the treatment or support they need for other health conditions.
“The steep rise in figures for the capital is reflected in the levels of pressure we have seen on hospital intensive and critical care units and ambulance services in the region,” said Sarah Scobie from healthcare thinktank the Nuffield Trust.
“But it is too early to tell at this point whether deaths where Covid-19 was not mentioned are the result of some cases not being recognised as coronavirus, or a substantial increase in people dying from other conditions.
National Records Scotland releases figures on a slightly different timescale. In the week to 5 April, there were 1,741 deaths registered in Scotland, up from the five-year average of 1,098. Of those, 282 mentioned Covid-19 on the death certificate.
The Glasgow area has been by far the worst hit by the virus.
In Northern Ireland for the week ending 3 April there were 434 deaths registered, up from the five-year average of 298. Covid-19 was mentioned on 55 death certificates.