Mass testing means Britain may never reach zero Covid deaths

When will Britain's coronavirus epidemic be deemed officially over?

At a recent select committee, Prof Dame Angela McLean, the chief scientific adviser at the Ministry of Defence, said experts were "crying out" for some clarity on how many deaths would be considered acceptable so life could get back to normal.

There is certainly a feeling among several vocal government advisors that nobody should ever be allowed to die from Covid, and that we will need to keep restrictions in place until there are no deaths at all.

Yet current mass testing may mean the Government’s daily figures will never reach zero - even if people are no longer dying of Covid.

The problem lies in the huge numbers currently being tested.

Professor Chris Whitty, the Chief Medical Officer, has warned that coronavirus is likely to become endemic, so there will always be some cases spreading in the community. The more testing we do, the more of those cases we will find.

The Government's coronavirus dashboard measures daily coronavirus deaths by any death which happens within 28 days of a positive Covid test. So potentially, someone could be hit by a car or succumb to cancer four weeks after testing positive and would still be included in the daily figures.

Watch: What is the UK's future on testing, travel and vaccine passports?

When a high number of people are testing positive for Covid, a certain proportion of those people will naturally die within 28 days.

Between April 7 and April 15, some 17,061 people tested positive for coronavirus - of which 11 (0.066 per cent) would be expected to die naturally within 28 days - or 1.5 people a day, according to death rates from the Office of National Statistics for this time of year.

It may be higher than that, because the chances of death increases with age, rising to 0.8 per cent for the over 80s.

While there are currently more under-60s testing positive for Covid than over-60s (30.4 per 100,000 versus 11 per 100,000), the chance of an older person dying is more than ten times greater than the average.

Given the mass testing regularly taking place in care homes and hospitals, it is almost inevitable that some people who test positive for Covid will also die naturally within the next month.

Dr Jason Oke, of the Nuffield Department of Primary Care Health Sciences, at the University of Oxford, said: "I wouldn’t want to name a figure, but I think there could be a number of people dying just from natural causes, unrelated to Covid, who would have tested positive and our daily figure might be getting close to it now.

"We might be reaching that point where we can’t go any lower, and it’s not necessarily an indication that the virus is not decreasing.

"There are still people in older age groups becoming positive so some of them will die of other causes."

The Medical and Healthcare products Regulatory Agency (MHRA) has made this point repeatedly when warning the public not to worry so much about deaths following vaccination.

The MHRA argues that when such a huge population is being vaccinated a certain number would naturally be expected to die in the following weeks.

Yet no such allowance is currently being made for Covid deaths and the problem is likely to get worse after the Government made tests available to anyone who wants one, and encouraged businesses to regularly monitor staff.

While the ONS publishes more accurate death registration figures - separating out people who died "due to" Covid from those that died "with" an infection - that data is published 11 days after the Government's figures, and it is the daily dashboard data that has largely been driving real-time policy responses to the pandemic.

In recent weeks data from the ONS has shown that growing numbers of people included in the official data died "with" Covid rather than "from" the infection, and it is currently around 23 per cent.

Professor Carl Heneghan, director of the Centre of Evidence Based Medicine at Oxford, said: "This is an issue of when we have mass testing and low thresholds, you are going to pick up background people who are going to die. We’ll have difficulty interpreting the figures on a daily basis.

"The advantage epidemiologically, is that now that numbers are low you could go in depth and look at the records and assign a causation to the deaths, and answer some of these questions about whether people are actually dying from Covid.

"We have a real opportunity to clean up the data and make it more accurate."

Some experts believe that the large numbers of young people currently testing positive may help us reach zero deaths for a sustained period, because only a tiny number would be likely to die within 28 days.

Watch: How England will leave lockdown

Prof Kevin McConway, Emeritus Professor of Applied Statistics at The Open University, said: "In theory there would still be people dying within 28 days of a positive test, even if Covid had entirely gone away and all or almost all of those people who tested positive were false positives. But we aren’t remotely near that situation yet.

"The rolling rate of testing positive per 100,000 people is a lot higher for under 60s than over 60s.

"This means that, in fact, the 17,000 weekly new cases are a good bit younger on average than the population as a whole, so the number of deaths in 28 days in that 17,000 would be less than the 1.5 a day. But of course, it would still be greater than zero."

But Professor McConway warned it may be better to start using the ONS data as cases get very low to avoid the problem.

"If the indications are that deaths actually caused by Covid are getting very low, I’d hope that people would concentrate much more on the data from death registrations that ONS publish every week," he said.

"They are a bit in arrears – when they come out each week the latest figures are for the week that finished 10 days before – but with these things there’s always a trade-off between accuracy and speed, and if we want accurate figures rather than figures for yesterday, we’ll have to learn to wait a bit."

Dr Stephen Burgess, Group leader in the Medical Research Council Biostatistics Unit, at the University of Cambridge, warned that calculating the true number of people dying from coronavirus is difficult, as some people will die from the virus without testing positive, while others will be included wrongly in the official figure.

He added: “If Covid-19 becomes endemic in the population in the same way as other coronaviruses are endemic, then a few individuals would be counted in the category of "died within 28 days of a positive test" even if no-one in the population dies of Covid-19.

"While it is true that in such circumstances it wouldn’t be possible for this measure to reduce to zero over a prolonged period so long as the virus is still circulating, the number of such individuals counted in this way is likely to be very small compared to those with serious long-term consequences of Covid-19, several of whom may die prematurely due to Covid-19, but will never be counted in Covid-19 death statistics.”

Since the first wave, deaths in Britain are yet to fall below three a day, even last summer when there was very little virus circulating. While deaths fell to just four in Britain on Monday, they shot back up to 30 on Tuesday.

With excess death levels now way below average and with no signs that the virus is increasing, it is maybe now time to ditch the dashboard and realise that once numbers fall this low, they can no longer be trusted.