The UK has repeatedly set and broken new records for the number of Covid-19 deaths reported in a single day throughout January.
Amid the second wave of the pandemic, these escalating totals have been echoed by the media and widely shared as grim markers of the extent to which the virus has ravaged the nation.
But while it’s important to pay attention to these daily totals, they don’t tell us the whole story.
How the data works
The government presents daily deaths in two ways on its dashboard – the first, deaths within 28 days of positive test recorded by date of death and the second, deaths within 28 days of positive test recorded by date reported.
The second of these is the number that gets reported by the media each day, and is not a reflection of how many people actually have died in the previous 24 hours, which is information that takes longer to establish.
For example, 1,000 people could be reported dead in a single day, but those deaths could have actually taken place on any day of the pandemic, and only just been reported as Covid-19 deaths.
The most recent record for a daily report was set on Tuesday, which saw 1,610 deaths recorded in a single 24-hour period.
But if we’re looking at what the actual deadliest day of the pandemic is, we need to know when those people died, which this figure alone doesn’t tell us.
So when did the most people actually die?
Despite much greater daily totals of deaths reported being announced right now, during the second wave of the pandemic, the current record for the number of Covid-19 death by actual date of death is still April 8, which saw 1,073 – one of which was reported as recently as this week.
Even though more than nine months have now passed since that date new deaths are continually added as the records are updated and reflected in the overall toll – though this happens much more rarely now the vast majority of Covid deaths on April 8 have now been recorded.
But the April 8 record could soon be surpassed by a much more recent date, January 11, which currently stands at 1,064 and is likely to more new individual deaths added to it than April 8 as the days and weeks go by.
The three following days, January 12 to 14, all currently have death tolls (by date of death) of 1,000 or more, and all could easily exceed the 1,073 record set in April if the pattern of late additions to the overall toll continue.
These are the 20 deadliest days of the pandemic so far according to the most recent data, though these totals – especially the later ones – are likely to change as records update.
- April 8, 2020 – 1,073
- January 11, 2021 – 1,064
- January 12, 2021 – 1,052
- January 13, 2021 – 1,018
- January 14, 2021 – 1,000
- April 7, 2020 – 999
- April 9, 2020 – 999
- January 7, 2021 – 967
- January 10, 2021 – 959
- April 12, 2020 – 957
- April 11, 2020 – 956
- January 9, 2021 – 956
- April 10, 2020 – 941
- April 5, 2020 – 915
- January 8, 2021 – 906
- April 4, 2020 – 905
- January 6, 2021 – 896
- April 13, 2020 – 893
- April 6, 2020 – 892
- April 15, 2020 – 879
As you can see, four of the top five, and 10 of the top 20, have taken place in January, which seems likely to overtake April as the deadliest month so far.
Why don’t we report on these figures instead?
The main reason the daily increase – that is, deaths by date of report rather than by date of death – makes headlines is that it comes reliably and regularly. Every day from 4pm, the government’s dashboard is updated.
By contrast, the information about fatalities by date of death doesn’t come to light immediately.
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, or simply because it has to work its way through a number of databases before reaching the government.
So that person’s death wouldn’t make it into the December 21 data, which would therefore be inaccurately low if we reported it on the day.
How bad is the problem? Well, in reality, the majority of deaths aren’t reported the next day, meaning this method of reporting would be extremely inaccurate if we did it in real-time.
Just 140 of the 765 hospital 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 months passed, more and more fatalities on that day were reported, bringing the total for England’s hospital deaths alone to 974. That means most of those 974 deaths were reported on subsequent days.
So reporting daily increases by the other method isn’t perfect, but it’s the best data available on the day. We just need to remember what it shows us, and what it doesn’t.