‘Mass testing gives a skewed picture of the scale of the Covid pandemic
Telegraph analysis shows community prevalence is now far lower than December last year, when we had similar case numbers
By Sarah Knapton, SCIENCE EDITOR
29 June 2021 • 4:13pm
Mass testing is giving a skewed picture of the coronavirus pandemic, with community prevalence currently five times lower than when the country had similar case numbers last year, analysis by The Telegraph shows.
On Tuesday, Downing Street said it would continue to publish daily coronavirus figures, even after restrictions are lifted, claiming they "provide an important level of transparency".
Yet critics have previously called for the daily cases data to be scrapped, with focus shifted to admissions and deaths, because vaccinations have broken the link between infections and healthcare needs.
On Tuesday, Britain recorded 20,479 cases, with the seven-day total increasing by 72 per cent. Looking at the daily case data, it might be assumed that the country is now in a similar predicament to mid-December, when around 20,000 daily infections were reported.
Yet on December 13, when cases hit 20,263 average of 340,285 tests were being carried out each day, compared to the current seven-day rolling rates of 922,622.
The chance of testing positive now is just 2.2 per cent in comparison to 5.9 per cent when we had similar daily rates last winter.
It means that community prevalence must have been higher the last time that the country reported similar figures.
This is confirmed by latest infection figures for the Office for National Statistics (ONS) which has been monitoring community prevalence separately. Weekly data shows that just 1 in 440 people in the community in England currently has a coronavirus infection.
When we were experiencing similar daily figures last winter, community prevalence was 1 in 85, five times higher than it is now.
The data suggests that if we were testing the same number of people as last year, we would only be picking up around 7,536 cases a day.
Prof Carl Heneghan, director of the Centre for Evidence-Based Medicine, University of Oxford, said even if lateral flow testing is removed from the data, the country is still in a very different position to last December.
“The seven-day positivity rate by PCR (Polymerase chain reaction) testing on June 23 was 3.6 per cent, whereas it was 9.3 per cent in mid-December and at its height it hit 18.2 per cent,” he said.
“The fundamental problem with testing is it depends on who turns up, so you get selection bias, and the increase in testing only adds to problems with the data. You put all this together and following the case data becomes a problem for us in interpreting what to do next.
“The ONS infection survey is a better judge of what’s going on because you have a random selection of people.
“Singapore is now accepting that this virus is endemic and has said that daily case reporting is becoming problematic for society. It’s instigating a sort of anxiety, and policies that are uncalled for.
“The mass testing also skews where Britain sits in world rankings. A lot of asymptomatic cases are being picked up here through blanket testing, but you’ll get underestimates in countries not doing that.”
Prof Kevin McConway, emeritus professor of applied statistics, The Open University, said the daily case figures can be subject to biases depending on who is being tested, the general level of concern, whether surge testing is happening, and what kinds of tests are being used.
“I'm not sure it is as extreme, but I think it’s certainly an issue,” he said.
“It’s always been the case that the new confirmed case counts on the dashboard can be subject to biases, in that they don’t compare like with like.
“Other things being equal (which of course they never are, quite), the more people you test, the more cases you find. It’s important to find cases, in order to keep control of outbreaks and local surges, but that’s bound to lead to changes in who’s being tested and why, and in turn that means that the trends in the numbers of cases that are found might not mean what it looks as if they mean.
“Recently, the number of infections measured in the ONS survey has certainly been going up, but it hasn’t been going up as fast as the count of new cases on the dashboard.”
Death data from the ONS for the week ending June 18, also shows that the country is experiencing very few Covid-19 fatalities despite the apparent surge in infections, with a weekly rise of just 19 deaths. Wales recorded no Covid-19 deaths for the first time since the start of the pandemic.
Deaths from Covid-19 now make up just 1.1 per cent of all deaths, up from 0.8 per cent last week but experts said the figures were still small and not yet cause for concern.