Update: Today the DHSC announced 813 deaths, but 156 of the English deaths happened before April 6 for which we already have good data.

This is quite a problem for updating, so I have recalibrated the model and it now gives slightly lower estimates.

1/

A cautious estimate of the total number of UK excess deaths linked to coronavirus stands today at

42,700

The profile of the virus path now matches the date of death rather than date of announcement

2/

There are always going to be revisions as we learn more. The downside of moving to the date of death is I lose the latest date figures for England and have to impute them. But it is better that way than assuming data is current, when much is almost a month old.

3/

The daily profile looks like this now and no longer has the weekend effect.

There are 3 phases to this chart. Up to 6 Apr - it is calibrated to match ONS escess death data (assumed to be true)

4/

After Apr 6, it is calibrated using NHS England & Public Health Wales data and daily announed Scottish and NI numbers.

After Apr 20, the English and Welsh data have to be projected using a deliberately low daily increase

ENDS