select search filters
briefings
roundups & rapid reactions
before the headlines
Fiona fox's blog
The Office for National Statistics (ONS) have released provisional counts of the number of deaths registered in England and Wales in the week ending 11 December 2020.
Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:
“On average over the last five years, the number of deaths increases a bit each week at this time of year as we get into the winter. ONS have released the numbers of deaths registered in England and Wales each week up to the week ending 11 December. For this latest week, the number of deaths hasn’t increased – in fact it’s fallen very slightly by 11 deaths. The number of excess deaths – the difference between weekly deaths this year and the average for the previous five years – is down too. So is the number of deaths registered where Covid-19 is mentioned on the death certificate, and that was increasing since early September until the latest two weeks.
“That is positive news, but it’s certainly no reason to be complacent or to drop our guard. The number of deaths in the latest week is still 14% above the five-year average – that’s over 1,500 deaths registered in the latest week more than might have been expected. There were still 2,756 registered deaths involving Covid-19, which is 22% of all the deaths registered in the latest week. Deaths involving Covid-19 were down on the previous week in many regions of England, but not in the East Midlands, the East of England, and the South East, and they were up in Wales too – and in every region of England and in Wales there were more deaths registered in the most recent week than the average for the previous five years.
“Also, these figures go up only to the week ending 11 December. Everyone will have seen reports about a rapid increase in Covid-19 cases very recently, but that didn’t begin in England (taking the totals for the whole country) until the very end of November, and in Wales till mid-November. Before those upturns numbers of cases had been decreasing for a time. In the sad event that someone dies from Covid-19, that will generally occur some time after they were first infected – maybe three weeks or more on average. So, in the totals for England, I wouldn’t expect the numbers of deaths to have increased by 11 December as a consequence of the recent rises in infections. I might have expected some upturn in deaths in the week ending 11 December in Wales, because the increase in cases started earlier there than in England, and in fact the numbers of deaths registered in Wales involving Covid-19 were about 8% higher in Wales in that week compared to the week before. These trends are in line with the data on deaths available on the dashboard at coronavirus.data.gov.uk. The current data on deaths on the dashboard are complete only up to 13 December (for 7-day averages – the daily counts are really too affected by random variation to be very helpful). For England, there is not yet a sign of an increase up to that date, but in Wales, numbers of deaths have been increasing since about the start of December. Give then trends in confirmed cases on the dashboard, I’d unfortunately expect total Covid-19 deaths in England to start increasing soon, as indeed they already are in London, the South East, and the East, and I’d also expect the deaths in Wales to continue increasing. I hope I’m wrong on this, and we’ll see when more data come in over time, but these things are unfortunately fairly inevitable.
“As you are probably very aware by now, there are complications in looking at trends in numbers of deaths, because they can be counted in different ways (on the basis of registrations or past positive tests, and on the basis of the rate when the death was recorded or registered or when the person actually died). These differences don’t, in the short term anyway, tend to affect the general trends, though they do affect the actual numbers that are recorded. To check, I have compared the trends for the latest week from these different definitions, and I saw the same general trends from all the sources, though the actual numbers that I’ve quoted are based on the date when deaths are registered.
“Last week’s potentially positive news about deaths at people’s homes, that the excess over the five-year average had fallen, hasn’t continued into the most recent week. There were 857 more deaths at home in the week ending 11 December than there were on average in that week in the previous five years. That means that deaths at home are running at about one third above the usual level. The great majority of these deaths are not recorded as involving Covid-19.”
Prof Sheila Bird, Formerly Programme Leader, MRC Biostatistics Unit, University of Cambridge, said:
“Provisional counts for COVID-mention deaths by date of occurrence are key data. They show November slow-down but the uncertain glimmer of early December down-turn in England and Wales. We must up our guard as there are lives to save over the next 11 winter weeks.
“My table, updated from last week, shows COVID-mention deaths in England & Wales (by week of occurrence) from early September to early December 2020. The cumulative 2nd wave total of COVID-mention deaths that have occurred by 11 December is at least 21,500 (to the nearest 100).
“Not all deaths which occurred in the week ending 11 December 2020 have yet been registered. Taking registration delay into account, we can expect that 2,780 COVID-mention deaths may have occurred in England & Wales in the week ended 11 December 2020, see Table. But recent multipliers add uncertainty (e.g. up to 3050).
“This week introduces a frisson of uncertainty about down-turn in COVID-mention deaths as the most recent fortnightly multipliers in yellow (2880/2841 = 1.01) and turquoise (3009/2603= 1.16), although decreasing, are both above 1. If the yellow multiplier continued to apply, we’d expect around 3050 COVID-mention to have occurred in the week ending 11 December. But, using registration-day as the basis for my estimation, we’d be running noticeably lower at 2780. Both estimates are shown as the new variant could have thrown a spanner in the works that could affect COVID-mentions deaths after mid-November.
“The summary remains that COVID-mention deaths in Wave 2 show the glimmer of a turned-down but there is additional uncertainty and so we must be ultra on-guard over the next 4 weeks so that the lustre lingers rather than the light be dimmed.
“For more detailed forecasting for English regions, see MRC Biostatistics Unit’s nowcasting and forecasting.”
Prof Lawrence Young, Professor of Molecular Oncology, Warwick Medical School, said:
“The ONS data provides provisional counts of the number of deaths registered in England and wales including deaths involving COVID-19. The data for the week ending 11th December 2020 shows that coronavirus associated deaths in England and Wales have fallen for the second week in a row despite the surge in cases associated with the new virus variant. The number of deaths involving COVID-19 decreased in all English regions except the East Midlands, the East of England and the South East of England. Deaths associated with COVID-19 appear to be levelling off in the West Midlands. But this is against a background of rising cases (up by a third across the country last week) and increasing hospitalisations.”
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.”
None others received.