Ministers could face a fight to recoup their losses on orders for antibody tests amid calls to abandon pin-prick kits that can be used at home in favour of more reliable lab-based testing.
The health secretary, Matt Hancock, announced in March that 3.5m antibody tests had been bought to help establish who had acquired immunity to the coronavirus – a crucial step in getting people back to work – but later clarified that the orders were subject to the tests being approved.
Rigorous checks on a sample number of the tests have so far found that none meet the standards agreed with the Medicines and Healthcare products Regulatory Agency (MHRA), forcing government to row back on claims that the tests would be ready to send out to the public this month.
But Martin Hibberd, a professor of emerging infectious disease at the London School of Hygiene and Tropical Medicine, said finger-prick antibody tests were rarely reliable and encouraged the government to focus instead on lab-based tests.
“We need a centralised pathology lab antibody test. The government has bought these home tests but they are traditionally very poor in terms of sensitivity and specificity. Expecting those to perform exceptionally well is unlikely in my estimate because they don’t perform particularly well for any virus,” Hibberd said.
“What use is home testing going to be? What are you going to do when someone at home decides they can see a band on a strip and they go out and say I’m free, I can do what I like? I think they should drop it,” he added.
The Department of Health and Social Care said it had bought its current stock of antibody tests on the basis of “minimal initial volumes” but refused to comment on specific commercial arrangements. It said that if the tests did not work, the orders would be cancelled and costs recovered “wherever possible”.
Sir John Bell, a Regius professor of medicine at Oxford University, who is leading the testing effort, said the assessments were ongoing and that his group was talking to companies in the hope of improving the antibody tests, adding that “a lot of the tests look very similar and probably come from the same factory in China”.
The Oxford group has characterised six different antibody tests and found that most fall well short of the accuracy required. Most failed to detect antibodies half the time and the best spotted the immune cells only 70% of the time. If used at scale, the tests could leave millions of people who have immunity convinced they are still vulnerable to the infection. For the test to pass, the accuracy would have to be nearer 95%.
The tests were also found to be unreliable in spotting those who were not immune, with some declaring a blood sample positive for protective antibodies 2% of the time. That false positive is a big risk, because it could mean nurses or carers going back to work believing they are immune when they are not.
The failure of the tests and the complexity of the work to improve them means that a working antibody test could be months away. The tests may be performing badly because they were validated against blood from patients who were severely ill and produced a strong immune response to the virus, unlike most people, who become only mildly ill.
The global rush for antibody tests meant that many governments found themselves in the position where they had to weigh up the risks of buying poor-quality tests and having no tests at all. “The Department of Health and Social Care wanted to get tests quickly and felt it had to commit,” Bell said. “To their credit they put them through rigorous testing.”
“It’s still possible that we won’t get a test … We don’t want to put out a test that is poor. We want a good test that gives us the answers we want,” Bell added.
Without a home-test kit, blood samples can still be checked for antibodies using a lab-based process called an Elisa, which is what Hibberd wants to see adopted. The Elisa test at Oxford can distinguish positive and negative samples of blood with almost 100% accuracy. The downside is that blood needs to be taken from people and sent to the lab for testing.
“This is the sort of testing we need, not these home kits,” Hibberd said. “Will they get their money back on these tests? Who knows. Somebody else will buy them, I’m quite sure.”
