select search filters
briefings
roundups & rapid reactions
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 14 May 2021.
Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:
“This week’s ONS release of provisional numbers of deaths registered in England and Wales is again rather difficult to interpret, because of the early May bank holiday. The latest figures are for the week of 8-14 May. There wasn’t a bank holiday that week, but there was the week before. There are fewer deaths registered than usual in bank holiday weeks, because registry offices don’t operate as normal. But the deaths that aren’t registered in bank holiday weeks still have to be registered some time, so the numbers registered in weeks after bank holidays tend to be higher. This disturbance to the patterns also affects the ONS comparisons with average numbers of deaths in the corresponding weeks in 2015-2019, because the bank holiday doesn’t always come in the same corresponding week each year. It would be less confusing to use number of deaths classified by when the person’s death actually occurred rather than when it was registered, and ONS do provide the numbers by date of occurrence too. But the numbers for recent weeks, particularly the most recent week, are always too low when they are first published, because not all deaths will have been registered yet. That’s why these are provisional figures – they will be revised upwards in future bulletins as the later registrations come in. That happens even when there aren’t bank holidays interfering. So I can’t use those figures by date of occurrence of deaths to sort out what the real pattern of short-term change was.
“That said, the data for the latest week on numbers of deaths where Covid-19 is mentioned on the death certificate are very encouraging. It’s true that the total number registered in England and Wales for the most recent week, 151, is a bit higher than the previous (bank holiday) week, 129. But I think this increase is entirely explained by the bank holiday. The number of deaths involving Covid-19 in the latest week was more than a quarter down on the figure from two weeks before that, 205, before the bank holiday. That’s a pretty big decrease, and taking account of the fact that registrations in the week after a bank holiday tend to be relatively high, that makes it even more impressive. In the latest week, just 1.5% of all registered deaths had Covid-19 on the death certificate – that’s lower than the previous week, and in fact those percentages have fallen consistently every week since the end of January, when almost half (46%) of all registered deaths had Covid-19 on the death certificate. Registered deaths involving Covid-19 are down in the latest week, compared to two weeks before (before the bank holiday), in England and Wales taken separately, in most of the English regions, and in nearly every age group. (When numbers get to be this low, you’d expect to see increases in a few regions or age groups just because of statistical variability.) These falls in numbers of death involving Covid-19 match the pattern on the dashboard at coronavirus.data.gov.uk, where the numbers of deaths (defined in a different way) are also continuing to fall.
“So, really, this is all good news. There aren’t really any signs yet of increases in deaths involving Covid-19 because of the relaxation of restrictions and lockdowns, or from the effect of new variants of the virus. It’s not really possible to say much about when the people, who, sadly, died in the latest week would have been infected, because the length of time between infection and death is very variable – but it’s likely to have been at least two weeks before at least, and usually longer, so the end of April at latest. That’s after the removals of restrictions that occurred around Easter, but well before the reopening of indoor hospitality and so on in England on May 17, and before the numbers of cases involving the Indian variant B.1.617.2 had grown much. The vaccines, and other interventions, really are working against deaths. That’s not to say that there might be increases because of loosening of restrictions or variants in later weeks – I expect that there probably will be at some point – but there’s no sign yet.
“The picture on deaths from all causes is less clear. The number of registered deaths from all causes in the latest week was 10,164, which is not only higher than the previous week by over 2,000, but is higher than the two weeks before that, so before the bank holiday, as well. The total number of registered deaths from all causes is above the 5-year average in the latest week, having been below the 5-year average every week since early March. However, this does not concern me greatly. Numbers of death registrations in the week after a bank holiday, like the latest week in these figures (8-14 May), are relatively high because of late registrations that would otherwise have occurred in the bank holiday week. ONS estimate that the number of deaths that actually occurred in the latest week was between 7,728 and 9,682. (They give a range of plausible values because it’s hard to tie down the exact number at this stage.) Both this figures are quite a lot less than the number of registrations, 10,164, and even if the actual number of deaths is at the top of the ONS range, 9,682, that would fit with a slowly reducing number of deaths from all causes. It’s difficult to say anything definite, though, because of the bank holiday effects and other uncertainties, and because we don’t yet have much clear information on the causes of death other than Covid-19. So I won’t say more – things will be clearer in next week’s ONS bulletin, but then we only have two weeks of normal service before the next bank holiday disrupts the figures again.”
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 its Advisory Committee. I am also a member of the Public Data Advisory Group, which provides expert advice to the Cabinet Office on aspects of public understanding of data during the pandemic. My quote above is in my capacity as an independent professional statistician.”