Earlier Public Health Scotland published a set of numbers from the pre-Christmas student asymptomatic testing programme. As around the rest of the UK, this was a key component of the “get students home for Christmas” thing that dominated the discourse last term.
Lateral flow device testing began on the 30th November and finished on the 19th December 2020 – and the rules were that students were offered two tests, spaced three to five days apart, and if either of the LFD tests returned a positive result, the student was asked to self-isolate and arrange for a confirmatory polymerase chain reaction (PCR) test.
We now have results from that exercise (the only nation to have published the numbers is Scotland), the first real world results from student lateral flow testing, and they are concerning. The first thing to note is that all of Dundee University’s results (and associated follow-up PCR confirmatory tests) are scrubbed because of evidence of a flawed LFD test batch at this test site which is currently under review. Let’s hope that’s an isolated incident.
More concerning are the overall results. Approx 28,000 students took part in the programme, just shy of 10 per cent of Scotland’s students. Even if you juggle around justifications like many opting out because they weren’t “going home” and the Dundee thing, it’s a far lower number than I’ve heard anyone dare predict out loud.
Of that 28k, about 10 per cent didn’t bother to get a second test. And no – that’s not because anything like 10 per cent had a positive on their first test. The good news is that just 0.1 per cent – 35 cases – tested positive. The problem is what happened next.
Of those 35 that tested positive, all went on to get a confirmatory PCR. But only 25 (71.4%) tested positive. Now maybe there are real complexities here I’m missing – but if you were to take PCR as some gold standard, that’s a “real world” false positive rate of 28.5%.
We’ve looked before at questions over whether LF testing in the real world generates false negatives – which may lead to students believing they’re safe when they’re not. If there’s also a potential false positive problem – causing unnecessary self-isolation, that’s a problem too. In any even, we’re back to a bigger debate about whether LF tests should be used to screen a population in general, or to inform individual decisions or reassure individuals about infection or otherwise. The former looks fine – the latter there’s much more debate over.
The potential for false positives also matters because of what’s been announced today south of the border.
As we signaled in Friday’s Wonkhe Daily, DfE has announced that due to the high prevalence of infections in England, confirmatory PCR testing has been temporarily removed for LFD tests taken at higher education asymptomatic test sites until prevalence lowers again. The notification of a positive LFD test now triggers the legal duty to self-isolate, self-isolation payments and contact tracing.
DfE says that confirmatory PCR testing was introduced to reduce the number of people self-isolating unnecessarily after receiving a false positive result from LFD tests. This, it says, is because in areas of prevalence lower than 1 per cent, there is a higher likelihood of false positive results.
As a result, it now says that because prevalence is above 1 per cent, the performance (positive predictive value) of Innova LFDs and PCRs are “broadly comparable” for measuring viral loads of infectiousness, and expected to be “indistinguishable” when used at test sites – significantly reducing the need for routine PCR confirmatory testing.
Cleverer people than me will be able to test whether that assertion from DfE is true, and will also be able to reflect on whether the apparently high “false positives” from the Scottish pre-Xmas exercise are as falsely positive as they look, and whether there’s any read-across to the supposed reliability of LF tests now.
In any case, there are serious questions to answer here – not least because today’s ONS figures suggest that thousands more students are in university towns and cities than ministers care to admit. Are they getting a test, and if so, how many of those could end up false positives?