select search filters
briefings
roundups & rapid reactions
Fiona fox's blog
The Office for National Statistics (ONS) have released their weekly figures on the number of deaths in England Wales, including deaths from COVID-19 in all settings.
Prof Sir David Spiegelhalter, Chair of the Winton Centre for Risk and Evidence Communication, University of Cambridge, said:
“The latest ONS data allow us to look at the lethal risks to which younger people in England and Wales have been exposed over 10 weeks of the epidemic.
“32 out of 7 million 15-24 year-olds have died with COVID-19, around 1 in 200,000. In normal life, nearly twice that number (56) would be expected to be killed on the roads over the same period.
“Over 80% of COVID-19 deaths in under 45s have a pre-existing medical condition, including obesity, so only around 6 COVID-19 deaths occurred in those aged between 15-24 without such a condition. The death rate in those without such a condition is more like 1 in 900,000, which is about the same size as 10 extra days risk of dying on the roads.
“For 25-34 year-olds without a pre-existing medical condition, the fatal risk was around 1 in 240,000, around 6 weeks of extra road-risk.
“For 35-44 year-olds without a pre-existing medical condition, the fatal risk was around 1 in 70,000, around 16 weeks of extra road-risk.
“These are just risks to the individuals themselves – they can still pose a risk to others and should protect the vulnerable
“These risks reflect the peak of the epidemic. Without lockdown, they would presumably have been higher. And future risks will be lower due to reduced chance of catching the virus.”
Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:
“This week’s new ONS data on deaths, based on registrations, continues to show encouraging trends. The ONS report does mention that the total number of deaths in England and Wales in the latest week available, ending 5 June, was slightly higher than the previous week. They attribute this apparent increase to the bank holiday on 25 June, which would have the effect of delaying some death registrations. I think they are certainly correct on this – the increase does show up in the data calculated according to the date when deaths were registered, but there is no sign of an increase in this latest week if deaths are instead allocated to the date on which they actually occurred. Some deaths that occurred in the most recent weeks would not have been registered by 13 June, the last registration date included in this week’s new data, but even allowing for that, the number of deaths that occurred in the week ending 5 June appears to be lower than the previous week, continuing the favourable trend of the past few weeks.
“The number of deaths where COVID-19 was mentioned on the death certificate is likewise down from the previous week, whether measured by the date of registration or the date of occurrence of the death. In the most recent week, about 14 deaths in every 100 had a COVID-19 mention; the previous week, it was about 18 deaths in every 100. That is still a considerable proportion of the deaths that occur – we are certainly not out of the woods yet – but the figures continue to move in the right direction. In the most recent week available, the numbers of COVID-19 deaths in both hospitals and care homes have fallen.
“The number of deaths in the latest week continues to be somewhat above the average for the previous five years, but again the number of these excess deaths is well down on the previous week. Measured by the date of registration, the number of excess deaths (from all causes) is less than half the figure for the previous week. This comparison may be distorted a little by the late May bank holiday – but there was also a marked fall in the number of excess deaths measured by the date of occurrence of death. Again, the trend continues in an encouraging direction. The number of excess deaths from all causes has, for a few weeks now, been less than the number of deaths where COVID-19 is mentioned. This appears to indicate that deaths from non-COVID causes are occurring less frequently than might be expected from previous years; to understand properly why this is happening would need deeper analysis of the data than is possible from these weekly data releases.”
Nigel Marriott FRSS, Independent Statistician, said:
“Excess deaths were 732 in w/e 5th June which is down from 1653 from the previous week ending 29th May. Total deaths are up but that is because these figures are registrations of deaths and the previous week included a bank holiday. As a result the number of registrations always falls at that time of year since the ONS are effectively working a 4 day week. Therefore whilst deaths were lower last week, so was the baseline (5 year average) hence why the excess or difference was higher last week.
“The latest figure of 732 is the first time excess deaths have been within a historical range since the week ending 27th March. Over the last 10 years, excess deaths for the 23rd week of the year have fluctuated between -2459 and +1290. Thus if the pandemic had never happened, an excess death figure of 732 would have normally passed unnoticed. Some people might say that because of this, the first wave of the pandemic is now over from the point of view of excess deaths. Personally, I would say it is over once excess deaths hit are near zero or negative but I can see the logic of using the other definition.
“However, the breakdown by location of death reveals some patterns that need explanation. It is now clear that the first wave breaks down into 3 parts –
I. Weeks 11-13 to 27th Mar – first COVID-19 deaths registered, excess deaths within normal range.
II. Weeks 14-17 to 24th Apr – height of first wave, record excess deaths, excess deaths exceed total number of COVID-19 registered deaths by a considerable margin meaning there were significant excess deaths not due to COVID-19. This shocked a lot of commentators at the time.
III. Weeks 18-23 to 5th June – lee of first wave, excess deaths fall until they return to historical ranges in week 23. Excess deaths broadly match COVID-19 registered deaths meaning there were no meaningful excess deaths not due to COVID-19.
“However, part 3 of the first wave hides the fact that the pattern is very different by the 4 locations used (Home, Hospital, Care Homes, Other). Since week 15 (w/e 10th April) – Hospitals non-COVID-19 excess deaths have been consistently negative (averaging 1500 per week) and this continues through to today. There have been no hospital excess deaths over the whole of part 3 of the first wave i.e. all COVID-19 registered deaths in hospitals have been cancelled out by negative excess deaths from non-COVID-19 deaths so that total excess deaths in hospitals are essentially zero. Indeed in the most recent week, hospital excess deaths are now negative. In effect during part 3, total deaths in hospitals were normal with COVID-19 deaths substituting for other causes.
“In the other 3 locations, excess deaths continue to be positive i.e. total deaths exceed the baseline and the excess deaths continue to be a mixture of COVID-19 and non-COVID-19 deaths. For care homes, non-COVID-19 deaths are now zero in the most recent week so current excess deaths there are all due to COVID-19. But in the Home, COVID-19 has barely been mentioned on death certificates and so nearly all excess deaths in the home are non-COVID-19. More than that, the scale barely changed in part 3 about 1,000 per week and is still similar to the scale seen in part 2.
“Care homes have received attention in the media but deaths in homes have not and I do think this needs to be investigated.”
All our previous output on this subject can be seen at this weblink: www.sciencemediacentre.org/tag/covid-19
Declared interests
Prof Kevin McConway: “Prof McConway is a member of the SMC Advisory Committee, but his quote above is in his capacity as a professional statistician.”
None others received.