When Jess and her partner James arrived at the coronavirus testing centre in Preston, they were given two options. They could either swab their own noses and throats, or have a trained professional do it. Jess says that she and James, a mental health nurse, were “relieved self-testing was an option” because they “didn’t want to risk anyone else”.
But the self-test is uncomfortable, and hard to get right. Matthew Avison, professor of molecular bacteriology at the University of Bristol, describes it as “somewhere between tickling your nostril and brain surgery”.
First, you put a small stick in your mouth, push it to the back of your throat and run it over your tonsils on both sides. Then, you put the same stick far enough up your nostril that you “feel resistance” and rotate it for 10 to 15 seconds. Jess describes it as “the worst thing I’ve ever put up my nose”.
The couple’s test results took 48 hours to come back, and were negative. James had planned to return to work immediately to help his “understaffed and overworked” team – but they both still had a cough, so they decided to “follow government guidelines and add more precautions” by staying at home for a further week while their coughs subsided.
Statistics suggest they were not being over-cautious. The current coronavirus test remains, according to best estimates, around 70 per cent accurate. A positive result, backed up by symptoms, is likely to be correct, but many negative results are wrong. Even people with severe symptoms are regularly testing negative.
A false negative is more than an inconvenience. James was cautious about going back to work, but others will take the result at face value. “If they’re a supermarket worker, they infect everyone they meet that day,” says Avison.
The swabbing process is the biggest source of error. The “viral load” – the amount of the virus in a person’s body – moves and fluctuates. A single swab will often fail to pick up viral particles, especially if it has not been inserted deep enough into the subject’s mouth or nose. In a hospital, trained professionals can test and re-test, practising their technique. The average member of the public, performing an invasive test on themselves in front of a mirror, cannot.
Even doctors need practice. “Several friends and colleagues have had very clear Covid-19 symptoms, have self-swabbed and they have all had negative results,” says Lucy Pocock, a GP and co-lead of the Bristol Palliative and End of Life Care Research group. “Some of them are healthcare professionals, or the family of healthcare professionals, and the test result has been used to come out of isolation and return to work.”
One person who has self-tested says the instructions, which run to nine pages, are cryptic. “Imagine [reading] this ill, in a boiling hot car, with a huge queue behind you,” she says. “The kit they gave didn’t match the kit on the list. I couldn’t read it. I asked [my 15-year-old son] to follow it today and he asked: ‘Where do I start?’”
John Edmunds, professor at the London School of Hygiene & Tropical Medicine, is a member of the Scientific Advisory Group for Emergencies (Sage), which advises the government on coronavirus policy. Edmunds says the effect of the UK’s testing programme is a “positive one overall”, despite the inaccuracies, because it covers people who have no symptoms. Without a test, these people would carry on working, and spread the virus.
But Edmunds also acknowledges that false negatives can make testing some people with symptoms “counterproductive, if the test result is not extremely accurate”, as these people may return to work thinking they are safe.
The accuracy of the test is important not only for individuals taking it, but for ministers deciding policy. Edmunds says that the high level of false negatives means that “test, track and trace” will not control the spread of the disease on its own. “If we move to a much wider contact tracing and testing regime we will have to keep in place many social distance measures,” he says. “Self-isolation and household quarantine is highly likely to remain in place for all individuals with symptoms.”
The science of the current test is sound, but the way the virus replicates in our bodies means reliable samples are hard to take. One alternative is to test saliva instead; spitting in a cup is less invasive, and early studies show promise. “It might be the case that saliva tests are more consistent and more sensitive,” says Nicholas Grassly, professor of infectious disease and vaccine epidemiology at Imperial College London, who sits on one of the Sage sub-groups. But he stresses that there “needs to be more studies to confirm that this is the case”.
Another option is more targeted testing. On 30 April, NHS Providers, which represents hospital trusts, called the government’s target for carrying out 100,000 tests a day a “red herring”, and said ministers “may be testing for testing’s sake”. More is not necessarily better.
“The value of the 100,000 is only as good as the results, and what’s done with them”, says Danny Altmann, professor of immunology at Imperial College London, who describes the current testing approach as “scattergun”.
The government’s “test, track and trace” regime, due to arrive later this month, aims to refine this approach, by testing the contacts of anyone who has coronavirus. But simply testing those at risk, or those who are symptomatic, does not yield a lot of detail about how the virus is spreading.
If we want to understand the spread of the virus and its prevalence in the population, says Matthew Avison, we need to test a “wider, representative sample across the country”, with trained professionals testing the same groups repeatedly.
Imperial College London and Ipsos Mori launched one such programme last week. It will test 100,000 randomly selected people from 315 local authorities, which Imperial’s Grassly says will “give a snapshot of how many people are currently infected”. It will be followed up, in time, by antibody tests that will show how many people have been infected and recovered. But a reliable antibody test remains elusive. In the meantime there is a need for patience and transparency.
On 2 April the Health Secretary, Matt Hancock, pledged that the UK would “carry out 100,000 tests for coronavirus every day” by the end of the month. On 1 May, Hancock was able to announce at the daily press conference that 122,347 tests had been “provided” the previous day. But 40,000 of these tests were counted because they were posted to households or to NHS care settings; the government has no idea if they were “carried out”, and statistics obtained by the Guardian suggest that almost all the tests sent to care homes went unused.
In a disaster in which politicians are expected to take a leading role in delivering a solution, political goals will inevitably be set. Such goals introduce further risk, because they must be met in the eyes of the voting public, whether the figures are being gamed or not. Now, the government isn’t even meeting its own target: the number of tests appears to be stabilising around 84,000. And yet on 6 May, the Prime Minister, Boris Johnson, announced an even higher target of 200,000 tests per day by the end of May (No 10 later claimed he was referring to testing capacity, not tests carried out).
For self-testing to work, the public needs to know how conclusive the results are. But such transparency appears not to be in the interests of a government desperate to be seen to deliver results. Jess tells me she received no information on false negatives when she took her test. Another woman, Becky, who also took the test, says the same.
Hancock and other ministers have a daily platform where they can spell out the shortcomings of the current testing regime, giving people realistic expectations. Instead, just last week, Hancock said that testing can “help people get back to work… and there’s been a big expansion in the eligibility to get a test in order to be able to get back to work”. As the UK’s failure to act quickly enough on coronavirus becomes appallingly evident, it should be obvious that preferring good news over caution is only a temporary measure – and a dangerous one.