There is a "very strange" lack of positive results coming from rapid Covid tests conducted in schools, an expert in biostatistics has warned.
Of the 300,000 children tested using lateral flow devices (LFDs) last week, only 189 were found to be positive for Covid-19, according to Professor Jon Deeks, from the University of Birmingham.
Professor Deeks, who heads up the university's biostatistics, evidence synthesis and test evaluation research group, warned that "probably more than half" of positive results from these LFD tests are actually "false".
And negative results are also imperfect – as "you can't be sure that you haven't missed Covid", he said.
"Anybody who thinks that using these tests is going to make the schools completely safe is being misled," he added.
Professor Deeks pointed to research showing that the LFD tests give false positives at a rate of around one in 1,000, arguing that this suggests "most of the positive results we're seeing at the moment are probably false positives and not true positives".
Covid testing in schools: Doubts over accuracy of lateral flow devices
He was speaking after the Association of School and College Leaders (ASCL) warned yesterday that a quarter of a million students could be forced to self-isolate "unnecessarily" because of the government's failure to amend its controversial stance on Covid tests in schools.
The Department for Education has decided that students who get a positive result from a LFD test in school should be sent home for 10 days even if they then go on to get a negative PCR test result.
Speaking this morning on BBC Radio 5 Live, Professor Deeks said: "These [LFD] tests, we know, give false positive results. They're very rare, and the government keeps on telling us that they're very rare. But the problem is true positives are even rarer.
"At the moment we have data from the people who've had these tests in secondary schools for the past week, so these have been the children of the key workers and the staff at schools. And the last week's data we had, they tested nearly 300,000 children, and they only had 189 positive test results
"That means that we've got a very strange thing happening with these tests – that very few are coming back positive. And we know that these tests give false positives in about one in 1,000. This means that, actually, most of the positive results we're seeing at the moment are probably false positives and not true positives."
The Department of Health and Social Care yesterday used the same statistic – that there is "fewer than one false positive in every 1,000 lateral flow tests carried out" – to claim that "lateral flow tests are accurate and reliable and have extremely low false positive rates".
Health secretary Matt Hancock said: "This new data further confirms what we know – these rapid tests are extremely accurate and are helping to dramatically curb Covid cases.
"Rapid testing can help detect asymptomatic cases quickly, preventing the virus from entering schools, colleges or workplaces, and stopping outbreaks before they occur."
Asked about the proportion of tests showing false negatives, Professor Deeks said: "Well, we've got a big issue that the government has not evaluated the accuracy of these tests in children. They've tried them, they've piloted them, but they never actually looked at the accuracy of them, and there are many of us who are absolutely despairing that the opportunity to do this has been missed.
"But the fact that this test has found so few positives in the last week's data suggests to us that actually they're not very good at finding the disease when it's there in children. And we know that children respond to this disease in different ways. We desperately need some data on this, and the way we get data is by doing the PCR.
"So we've got negative results, which, when you get them you can't be sure that you haven't missed Covid, so anybody who thinks that using these tests is going to make the schools completely safe is being misled. And we have positive results, probably more than half of which are false positives.
"So, this isn't really an attractive testing proposition for us to look at doing, but it could be so much better if the government...took notice of these PCR tests on verifying, confirming these positives, like they will do next week [when tests will be taken at home and positive results will need to be confirmed via a PCR test]."
Other academics have previously countered claims that lateral flow tests are "inaccurate and potentially harmful".
A group from the University of Liverpool, including Iain Buchan, Louise Kenny and Calum Semple, said in January that reports suggesting the tests would cause harm were "flawed and confused".
In response to Professor Deeks' claims, the DfE pointed Tes to its recent blog on Covid testing in schools.