Every day at about 4pm, the government releases data showing how many people have died in the UK within 28 days of testing positive for coronavirus – and how big an increase that represents over 24 hours.
Taken at face value – an interpretation encouraged by the way the government words these updates – you’d assume that meant we knew how many people had lost their lives to Covid-19 over the course of the last day.
Consequently, you’d assume we also knew all kinds of other things: how rapidly the disease is spreading, whether the latest version of lockdown is doing anything to help, and so on.
But we need to take these numbers with a pinch of salt.
On the one hand, many of the most recent deaths won’t be included, even though they should be. On the other, deaths that actually happened weeks ago are finding their way into what should be today’s data.
Here are the main issues broken down.
Deaths from days ago are only now being reported
Today’s increase includes not only deaths that occurred over the 24 hours up to 5pm yesterday, but also deaths from earlier days that have only now filtered through to the data collection taking place centrally.
A death on, for example, December 20 might take more than a day to verify – say, if there is confusion over test results, or a delay in submitting data from a particular health trust.
That person’s death might then finally make its way into the total several days later – say, January 3 – meaning it would wrongly appear to be part of the daily increase on that day.
Just 140 of the 765 deaths reported in England on April 9 had actually taken place on April 8. One of the deaths in that dataset had occurred as far back as March 16. Yet April 8 has since emerged as the single day with the highest number of UK deaths: as the days passed, more and more cases from that day were added, bringing the total for England’s hospital deaths alone to 974 (as of January 13). That means most of those 974 deaths were reported on subsequent days.
This problem isn’t confined to the first wave. In a more recent example, a death from May 8, 2020, was included in the data released on January 13, 2021.
The most recent deaths aren’t being reported yet
By the same token, the true number of deaths over the last 24 hours is likely to be higher than we so far know. Many of those cases will not yet have reached the government’s statisticians and will instead appear in future counts, artificially elevating those.
When the curve levels off, these two delays might actually cancel out – but for the moment it’s hard to know exactly how wrong we’re getting it.
Only deaths after a positive Covid-19 test result are included
A third issue is that the UK data only includes deaths of people who have died within 28 days of testing positive for coronavirus. Until April 29, they didn’t include the deaths of people who had died outside hospitals at all, though this has now been rectified and the data prior to then has been backdated – which is why you don’t remember deaths topping 1,000 last year, even though we now know it happened for four days straight in April.
The Office for National Statistics releases more comprehensive weekly data on all deaths in which coronavirus has been suspected as a factor and recorded on a death certificate, and this is reproduced on the government’s coronavirus dashboard – but these numbers are at least 11 days behind.
The most recent figures show a gap in the region of 7.000 between the two methods, with death certificates representing the higher number.
It should also be noted that dying after testing positive for Covid-19, or with Covid-19 mentioned on a death certificate, does not necessarily guarantee that Covid-19 wholly caused that person’s death, adding a further layer of ambiguity.
It was reported on January 13 that 100,000 Brits had died of coronavirus. That figure was reached through a combination of methods of counting: it combines death certificates data from the start of January with more recent daily deaths figures published since then.
In other words, it’s a blend of two methods of counting. Neither the deaths-after-a-positive-test total nor the death certificates total has yet reached 100,000 on its own, which is why you’ll probably see the number reported as a new milestone again in the future.