As the UK adjusts to life under lockdown, one daily event is becoming a morbid mainstay of people’s new routine – the government’s release of coronavirus infections and deaths.
The ever upward trend of figures is a necessary yet chilling reminder of why our lives have changed beyond recognition in just a few weeks.
But as more and more data become available, it can be hard to wade through it and know what we should and shouldn’t pay attention to. Last week, for instance, marked the first time that an alternative set of weekly deaths figures was released – for a fortnight previously.
Let’s start with the basics...
What are the current stats?
The UK government currently reports two figures on a daily basis – total deaths and total infections.
Total deaths
The Department of Health and Social Care (DHSC) records and reports deaths during a 24-hour period from 5pm to 5pm, and releases those figures at 2pm the next day (though these have often been delayed).
As of 5pm on April 8, the DHSC said, 7,097 patients in the UK who had tested positive for coronavirus had died. Those figures were released on Wednesday.
Confusing matters slightly, each of the four NHS systems (England, Wales, Scotland and Northern Ireland) have continued to release their own deaths figures throughout the day, which often means the four devolved authorities end up releasing their own individual death counts first.
So can you add them all up to get the UK figure? Sadly, it’s not that simple.
HuffPost UK spent some time in discussion with the department to understand why the DHSC figure never matched the four totals. Full detail is here.
In summary, though, the four nations continue working on their own numbers even after releasing them to the government, so what they release separately could be more up-to-date than the DHSC’s own release.
That doesn’t, of course, explain how the DHSC ends up releasing data after each of the nations has reported.
Total infections
These data show the numbers of people in the UK who have been tested for coronavirus and, of those, how many have tested positive.
Each day, the figure represents the total tests recorded as of 9am on that day.
As of April 8, a total of 232,708 people have been tested, of whom 99,198 were confirmed negative and 60,733 were confirmed positive.
The weekly figures
On March 30 it was announced the Office for National Statistics (ONS) would begin publishing a weekly death toll based on registrations rather than just hospital reports.
This means anyone who dies outside of a hospital will be included if a doctor or coroner, for example, notes Covid-19 as a factor in their death.
It means the ONS figures are more accurate and slightly higher than the daily figures over a given week.
But the trade-off is that there is a lag in reporting. The first set of figures came over a week after the time period which they covered.
What do they tell us?
You’d be forgiven for thinking the official UK government figures on total infections and deaths from coronavirus would tell you exactly what they say on the tin, but unfortunately it’s not that simple.
The two sets of figures vary greatly in just how accurate and useful they are. Kwiziq CEO Gruff Davies has been putting a PhD in Medical Physics from Imperial college to good use by modelling the pandemic and is unequivocal in which set we should not take too much notice of.
“Deaths is the only reliable statistic,” he told huffPost UK. “I’ve only been looking at death rates. The infection rates are basically too unreliable and too volatile.
“There’s just no way to reliably interpret that data.”
Why are infection rates unreliable?
The laboratories that are conducting the tests in the UK are obviously under great pressure and some tests are being prioritised over others.
This means the government figures for each day tell us the results of tests as they come in rather than the date when someone was tested because they were displaying symptoms of coronavirus.
This has led to large variations in the reported infection rates day-to-day that don’t reflect how the virus is actually spreading.
Sheila Bird formerly of the MRC Biostatistics Unit at the University of Cambridge told HuffPost UK she believed the government should start reporting infection rates detected in hospital-based tests sample-week by sample-week rather than daily, as this would give a far more accurate idea of the spread of the virus.
“It’s important that we know the sample week – not the week the test was reported, but the week the sample was taken from the patient,” she said.
“It’s this date that characterises the state of the epidemic.”
What do they not tell us?
A number of things, most critically the exact number of infections in the UK and therefore the “state of the epidemic” that Bird mentions.
The huge majority of tests are only being conducted in hospitals and people with mild symptoms of suspected coronavirus have been told to stay at home meaning they are not being officially recorded at all.
So how many people in the UK are actually infected?
“You probably want to multiply [the official infection numbers] by at least 20,” says Gruff.
Currently that means a total of somewhere in the region of 290,000.
What about how the infection figures are presented?
Currently the official government figures are presented in the following format:
As of 9am 8 April, 282,074 tests have concluded, with 14,682 tests on 7 April. 232,708 people have been tested of which 60,733 tested positive.
The government could make one very simple change to this that would vastly improve the usefulness of the figures. “The percentage of positives is the crucial thing,” says Bird.
As the sample size for testing is limited due to the number of tests available, the official infection figures only reflect this small section of the population.
Adding the percentage of positive cases combined with Bird’s suggestion of weekly reporting of hospital-based tests would make them significantly more informative, she believes.
What about the death figures?
There will be some discrepancy in the official death figures but nowhere near as large as the infection rates.
Like testing, deaths in hospitals are the only ones reflected in the daily government figures.
These will be highly accurate as anyone with coronavirus symptoms severe enough to be fatal will likely have made it to a hospital – but there could be a handful of cases (perhaps those in care homes, for instance) that aren’t being officially recorded in the daily figures.
The release of the weekly ONS figures should address this. Of the deaths registered in the week up up to March 27 week, 539 mentioned “novel coronavirus”, which is 4.8% of all deaths, the ONS said.
Anything else?
Currently there is no breakdown at all of who is dying yet. We have indications from other countries that certain demographics are more at risk than others.
We already know that the elderly are particularly susceptible but data from Spain shows men of all ages may also be more at risk.
So what does this all mean for me?
Stay indoors – follow the official government guidance for social distancing and self-isolating if you or anyone in your household has symptoms.
By all means, read the daily figures updates, but don’t base your behaviour on them – for instance, by relaxing social distancing or stopping self-isolating if the figures look positive from one day to the next. You could be putting yourself and vulnerable people at risk.