The Office for National Statistics (ONS) have released provisional counts of the number of deaths registered in England and Wales in the week ending 16 April 2021.
Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:
“This is another week where the numbers and short-term trends in the provisional death registration data for England and Wales are difficult to interpret because of the recent Easter holidays. The latest figures are for the week 10-16 April – that’s after all the bank holidays, so that the changes in late registrations because of the holidays should have ended and the actual counts for the latest week are probably comparable to other non-holiday weeks – but the numbers from the previous week, and the week before that, definitely weren’t, because they included the Easter weekend holidays. So comparisons between the latest week and those weeks on registered deaths won’t necessarily make much sense. In a couple more weeks, we’ll be able to get a more accurate idea by using data classified by when the death occurred rather than when it was registered – but the usual pattern of registration delays means that we can’t yet get good enough figures for the week ending 16 April by date of occurrence of death.
“However, there’s still clear good news on deaths involving Covid-19. The number of deaths with Covid-19 mentioned on the death certificate in England and Wales, for the latest week, is down compared to the previous week, even though the number of deaths from all causes was up quite a lot compared to the previous week. To get a clearer idea of trends in deaths involving Covid-19, I’ve compared the numbers in the latest week (ending 16 April) with the numbers for three weeks earlier, the week ending 26 March, before the holidays. Across all age groups, the number of deaths involving Covid-19 fell over that period from 719 to 362 – so that it almost halved. In people aged 50 and over, it decreased by 51%, so slightly more than halved. In the under 50s the reduction was proportionally smaller, 12%, but the numbers of deaths in the under 50s is much smaller anyway. The actual count of registrations in under 50s went down from 26 to 23, and in both those counts about half the deaths were in people in the age group just below 50 (45-49). Altogether this emphasises the impact of vaccines on Covid-19 deaths in older people, and indeed it shows again what a small proportion of the Covid-19 deaths are in younger people – but even in younger people, and with the small numbers of deaths involved, the figures moved in the right direction.
“It’s not so clear what’s going on in total deaths from all causes, though. They are 1,340 up on the previous week, but that comparison might be affected by the bank holidays. However, they are also about 400 up on the figure from three weeks earlier, before all the holidays, and at a time of year when numbers of deaths fall, on average. ONS correctly report that they are still running below the average for 2010-2015 for the corresponding week, and that that’s been the case for six consecutive weeks now. But in the latest week, all-cause deaths are only 82 below the five-year average. That’s fewer than one death in a hundred less than the five-year average. (10,438 deaths registered in the latest week, compared to a five-year average of 10,520.) It’s probably clearer to say that total deaths are running at about the average level. In previous weeks, deaths were running at several hundred below the long-term average – indeed well over a thousand in the previous two weeks, though they were affected by the bank holidays. I, and others, have been pointing out that this could be due to continuing effects of lockdown, particularly on deaths from respiratory diseases, and to the sad fact that some people, who would otherwise have died in those weeks, would have died earlier in the year or last year from Covid-19. So why has this pattern stopped? Difficult so be sure so far, I think, without more data on what actually caused the deaths. Apart from Covid-19, these weekly ONS bulletins give numbers of deaths from only one other cause, influenza and pneumonia. Deaths from those causes have remained fairly constant over the past few weeks, and are still running a long way below the five-year average level of deaths from influenza and pneumonia, so that’s not part of the explanation. I’ve looked at patterns of all-cause deaths at different ages and haven’t seen any obvious explanation there either. I don’t think there is anything to provide a clear explanation in the data on where people died. Again comparisons are difficult because of the bank holidays. Deaths in people’s own homes are up compared to the weeks affected by bank holidays, but only to a level slightly higher than it has been for many months.
“I will comment again on a different aspect, though, and I make no apology for saying again what I said last week about the fact that, of the deaths where Covid-19 is mentioned on the death certificate, not all are coded as having Covid-19 as the underlying cause. In the latest week in England and Wales, of the 362 deaths with Covid-19 on the death certificate, 275 had it recorded as the underlying cause of death (76% of the total). I said last week that this percentage has gone down compared to the height of the pandemic wave, but that this is what happens when death numbers are low and falling. But I’ve seen more comment in the last week that this means that the other quarter of people, who had Covid-19 on their death certificate but not coded as the underlying cause of death, weren’t really affected by Covid-19 at all, and that they would just have Covid-19 on the certificate because the certifying doctor has to put it on the certificate if they had had a positive test. That’s simply not true at all. The certifying doctor fills in diseases or conditions related to the death in two possible places on the certificate. Part I has some boxes where the doctor lists the sequence of diseases or conditions that led to death, and this generally identifies clearly what the doctor considers to be the underlying cause. Very often, if Covid-19 is mentioned in those boxes, it would be identified as the underlying cause, but if not, that would mean that the doctor considered it to be part of the chain of events that led to the death, so that if the person had not had Covid-19, the death would not have occurred. The other place Covid-19 could be mentioned is in Part II, where the doctor lists “Other significant conditions contributing to death but not related to the disease or condition causing it.” For Covid-19 to be mentioned here, the doctor has to believe it was contributing to the person’s death, perhaps by making it occur earlier than it would have if they weren’t infected, perhaps substantially earlier. So someone who, for instance, had had a positive Covid-19 test but was killed in a car accident, should not have Covid-19 mentioned on their death certificate at all. So, for these deaths involving Covid-19 but not having Covid-19 as the underlying cause, the virus could have had a substantial effect on the patient, made their last days much more uncomfortable, or even shortened their life by a substantial amount. They very much aren’t deaths that have nothing to do with Covid-19, and they very much aren’t like deaths under the 28-day definition of counting deaths that occur within 28 days of a positive test, whatever the actual cause of death.”
Prof Paul Hunter, Professor in Medicine, The Norwich School of Medicine, University of East Anglia, said:
“That there was a jump of 1,340 in the number of deaths recorded in the week ending 16th April over the previous week is not a cause for concern. The previous weeks figures were artificially low because of deaths not being recorded on the run up to Easter so this last week’s figures include deaths what would normally have been included in the previous week. Overall the number of deaths with COVID is following the same downward trend as is seen in the daily reports of deaths on the DHSC Dashboard.”
All our previous output on this subject can be seen at this weblink: