November 17, 2020

expert reaction to the ONS release on deaths registered in England and Wales, provisional: week ending 6 November 2020

The Office for National Statistics (ONS) have released provisional counts of the number of deaths registered in England and Wales in the week ending 6 November 2020.

Prof Sheila Bird, Formerly Programme Leader, MRC Biostatistics Unit, University of Cambridge, said:

“Key data from Office for National Statistics are provisional counts of COVID-mention deaths by date of occurrence. My table, updated from last week, contrasts COVID-mention deaths in England & Wales (by week of occurrence) in early March to early May 2020 with early September to early November 2020.

“Not all deaths which occurred in the week ending 6 November 2020 have yet been registered. Taking registration delay into account, we may expect that 2170 COVID-mention deaths may have occurred in England & Wales in the week ended 6 November 2020, see Table.

“Hitherto, I have estimated next week’s final total [eg now for the week ended 13 November] on the basis that the impact of tier-measures from early October had not kicked-in. But the measures appear now to have had impact! Last week, I’d estimate that COVID-mention deaths in the week ending 6 November 2020 would be around 2700: but, on current evidence, they are likely to be below 2200 when fully registered. This is a marled and important impact against my counter-factual.

As UK’s second wave has evolved, COVID-19 deaths in England and Wales had been more than doubling fortnightly (running-rate of 2.2 = {168+243+376+523+755+1244+1640}/{76+100+168+243+376+523+755} last week). The current running rate, from the week ending 9 October, is now 10% lower at 2.0 = {761+1254+1627+2170}/{377+523+761+1254}.

“This early moderation down of COVID-mention deaths seems too early to be attributed to the start of England’s lockdown (4 November to 2 December 2020).”

Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:

“Deaths registered in England and Wales are above the 5-year average again for the latest week (ending 6 November), as they have been every week since early September. Each week deaths (from all causes) have generally become higher and higher above the 5-year average, and this week they are almost 1,500 above it, which is over 14% above the average. The position in Wales looks particularly concerning – in the latest week, total deaths there are a third higher than the average. There’s been some slightly encouraging news recently about numbers of new infections – the broad picture is that the increases might well be levelling off. But there’s no sign at all of that in these death registration figures. That could, possibly, be because, if a person unfortunately dies from Covid-19, that would typically be two or three weeks after they first had symptoms of the infection. So maybe the current measures to reduce new infections are having an effect but it’s too early to see that in death registrations. But we absolutely can’t be sure of that yet.

“My other main concern is that deaths (from all causes) in people’s private homes are yet again well above the average of the last 5 years – by almost 1,000 for the latest week. That’s more than 140 extra deaths each day. Deaths at home have been at that kind of level, over 100 above average each day, since mid-May, just after the first peak of the pandemic. I still haven’t seen any clear explanation of why this might be happening, and that worries me.”

Prof Sir David Spiegelhalter, Chair, Winton Centre for Risk and Evidence Communication, University of Cambridge, said:

“Usually around 2,500 people die at home each week at this time of year – now it is 3,500. That is 1,000 more, an extra 40%.  These extra home deaths, very few of which are from Covid, have continued since March.  it would be very good to know the quality of end-of-life care being received, and how many of these deaths might have been delayed through, for example, more rapid treatment of heart attacks and strokes.

“When we look at the main underlying cause of death in England and Wales that week, there were 1,743 deaths for which it was Covid-19, compared to 307 as influenza or pneumonia.  And usually only a small proportion of deaths from ‘influenza and pneumonia’ are from flu itself,  and so Covid is definitely not in the least like flu.”

https://www.ons.gov.uk/releases/deathsregisteredinenglandandwalesprovisionalweekending6november2020 

All our previous output on this subject can be seen at this weblink:

www.sciencemediacentre.org/tag/covid-19

Declared interests

Prof Kevin McConway: “I am a Trustee of the SMC and a member of the Advisory Committee, but my quote above is in my capacity as a professional statistician.”

Prof Sir David Spiegelhalter: “DJS is a paid non-executive director of the UK Statistics Authority, that oversees the work of the ONS.”

None others received.