Simon was at his desk by the time I got in. I called good morning as I passed his open door, but his reply was far from cheerful.
“You OK?”
“I’m not sure.”
Since early January, we’ve been self-administering lateral flow antigen testing devices (LFDs) twice-weekly – part of the drive to detect Covid in healthcare professionals without symptoms. We do them at home, and get a result within 20 minutes. Simon had tested positive on his last one, taken on a Sunday evening.
The problem with LFDs is they are unreliable. They fail to pick up 30-50 per cent of infected individuals, and they also generate false positives. So Simon had self-isolated and arranged a confirmatory PCR – the kind of test being performed in the hundreds of thousands nationally every day now, and deemed to be more accurate than LFDs. Simon’s PCR had come back negative the following day. Only he didn’t look entirely happy. “What’s going on?”
“I don’t know,” he said. “I just feel a bit achy this morning, that’s all.”
There were no other symptoms, but I suggested we check his temperature: modestly elevated at 37.7°C.
“You’d better head home.”
[see also: Emily Oster on why reopening schools is the safest option]
A weird but fascinating property of medical investigations is that the accuracy of any test varies dramatically according to the probability of the disease actually being present. The maths behind this fact is rather involved but, essentially, when the probability of a disease is high, true positive results drown out any false positives. Conversely, when the probability of a condition is low, the same test will perform poorly, positive results being more often false than true.
Ordinarily, doctors use history (the pattern of symptoms that a patient has been experiencing) together with examination findings to hone down on the likelihood of a given disease. A correctly selected test will then be highly discriminatory in either confirming or refuting the diagnosis.
LFDs are diagnostic tests – very useful in differentiating Covid from other respiratory infections – but the UK is using them as a screening tool, trying to pick up coronavirus in people without symptoms. Like all diagnostic tests, LFDs are highly fallible when used this way.
Simon’s positive LFD test occurred in January, when Covid rates in the community were sky-high, so there was a reasonable probability of coronavirus being present despite him being asymptomatic. This was borne out when he went on to develop clear-cut Covid (from which he recovered). In this instance, it was the LFD that was right; the PCR had been a false negative. But the story serves to illustrate some of the conundrums surrounding the issue of testing.
Secondary school pupils returning to class on 8 March are expected to conduct twice-weekly LFD tests. Official discourse continues to perpetuate the myth that these tests – and, indeed, the confirmatory PCRs – keep individuals “safe”. They do not, just as masks, social distancing, hand washing and ventilation are not actually about individual “safety”. What these measures do is simply dampen Covid rates by interrupting some, but by no means all, chains of transmission. There is no such thing as a “Covid-safe” or “Covid-secure” environment, only Covid-less-risky ones.
Failure to explain honestly the true purpose and the downsides of LFD testing in schools will quickly lead to cynicism and complacency. As community Covid rates fall, the proportion of LFD false positives will rise and every school will also have cases of Covid develop in people who have had negative tests. Unless all this has been explained honestly, teenagers and families will soon be asking themselves: what is the point?
[see also: From Covid Kids to Generation Grit]
Content from our partners
Related
This article appears in the 10 Mar 2021 issue of the New Statesman, Grief nation