I watched Prof Jason Leitch (National Clinical Director of Healthcare Quality and Strategy) on Tuesday night on the National Parent Forum of Scotland seminar discussing the current restrictions in schools. Prof Leitch gave an example of how the results from testing in schools can tell us about how SARS-CoV-2 is spreading. He gave the example of testing 2000 people, 1000 of whom have SARS-CoV-2, and 1000 who don’t. He said:
“…we’ve got 1000 people who have got the disease…650 of those people will be positive. We find about 65% positives…Now 350 have gone back to work. They’re false negatives – they have the disease, but we haven’t found them…”
Prof Leitch may have chosen those numbers for the sake of his example to make the explanation easier to follow. There’s a few points to make about it.
1. SARS-CoV-2 or COVID-19?
Throughout the seminar these were used as if they meant the same thing. They don’t. The number of people testing positive for SARS-CoV-2 is known (this is what mass testing tells us). The number of people who develop COVID-19 is more difficult to establish. Mass testing only tells us how many people test positive for SARS-CoV-2.
2. What is the sensitivity of lateral flow testing (LFT)?
From Prof Leitch’s figures, it looks like he assumes that LFT has a sensitivity of 65% (hence 350 false negatives out of 1000 people tested who truly have SARS-CoV-2). A review examining the sensitivity of LFT and other antibody tests gives a range of results that depend on how soon testing is carried out after symptoms appear (Deeks, J. J. et al., Cochrane Review). The lowest sensitivities were found in the week following the start of symptoms (30%), and highest sensitivities were found 3 weeks later (90%). There were no studies reviewed that considered asymptomatic testing. In other words, we don’t know how well LFT will work in testing people who do not have symptoms. The low sensitivity of LFT led Prof John Deeks, biostatistician at the University of Birmingham, to say:
“The poor detection rate of the test makes it entirely unsuitable for the government’s claim that it will allow safe ‘test and release’ of people from lockdown…”
Mahase, E. (2020) ‘Covid-19: Innova lateral flow test is not fit for “test and release” strategy, say experts’, BMJ, 371.
This also presumes that the test used to confirm a positive LFT result (that is, PCR via the Lighthouse Labs for Pillar 2 testing) is accurate enough to provide confirmation. Prof Leitch said “there’s almost no false positives in PCR”. We don’t know that, as the Government still do not know the sensitivity (which tells us about true positives) and specificity (which tells us about true negatives) of PCR in Pillar 2 testing. A report prepared for SAGE in June 2020 stated:
“The UK operational false positive rate is unknown…The UK operational false negative rate is unknown.”
So we’ve never tested LFT for identifying SARS-CoV-2 in asymptomatic people – we don’t know how well it works there. And we don’t know how well Pillar 2 testing works. Shouldn’t we know these basic facts before we start doing more tests on our children?
3. Why 50/50?
Prof Leitch may have used the 50% with SARS-CoV-2/50% without to make his example easy to understand. A more realistic example would reflect the actual incidence of infection based on the Community Infection Survey conducted by the Office for National Statistics. This uses repeated testing on 15,000 households across Scotland to work out how many people in the whole country would test positive for SARS-CoV-2.
Their results show that in the last few weeks, between 0.2 and 0.5% of the population would test positive for SARS-CoV-2. So instead of 1000 infected and 1000 uninfected out of 2000 people, a more realistic example would give 10 people out of 2000 infected (0.5%). If LFT was used to test them (using Prof Leitch’s assumed sensitivity of 65% and specificity of 99%) what would happen?
We would discover
• 7 infected people (65% of 10 truly infected people)
• 3 infected people would be told they weren’t infected (35% of 10 truly infected people)
• 20 uninfected people would be told they were infected (false positives) – that’s 1% of 1990 truly uninfected people
• 1970 uninfected people would be correctly identified (99% of 1990 truly uninfected people).
LFT would give almost three times as many false positives as true positives.
Prof Leitch seemed to downplay the likely number of false positives in LFT testing in schools. There aren’t one or two “inconvenienced guy[s] or girl[s]” (as he suggests) – in fact, almost 6000 high school students every week (that’s the equivalent of 7 high schools) will have to self-isolate unnecessarily until they receive a confirmatory test.
Screening tests do not tell us much about the spread of disease when there is very little disease there in the first place. We are now in a situation where mass testing in children with no symptoms in schools will create more false positives than true positives.
How can Prof Leitch tell the difference between a true pandemic and a false positive pandemic?