A new kind of coronavirus test that is being widely used to screen people without symptoms had very low levels of accuracy at the University of Birmingham, UK, one of the few places where it was directly compared with a more accurate kind of test in a real-world setting.
Among Birmingham students, only 3.2 per cent of those infected with the virus were correctly given a positive result from the lateral flow tests being used there, according to preliminary data from the university.
This is much lower than previously reported sensitivity levels for this type of test. The sensitivity was 57 per cent when it was used in a mass-screening pilot in Liverpool, UK, and more than 70 per cent when it was checked in UK government laboratories, according to a spokesperson for the UK’s Department of Health and Social Care.
The first kind of assay to be developed for the coronavirus, called a PCR test, is more accurate. PCR tests indicate if the virus’s genes are present and are usually offered to people who have symptoms of covid-19. The tests normally have to be sent to a laboratory for processing and can take a day or two to produce results.
Lateral flow tests, also called antigen tests, are a more recent development. They indicate if certain viral proteins are present. Similar in design to home-use pregnancy tests, they can produce results in as little as 15 minutes.
Many organisations have begun screening people without coronavirus symptoms to try to stop infections being passed on unknowingly. They often use lateral flow tests because of their rapid results. But concerns have been raised about their lower accuracy, especially around the rate of “false negatives”, when someone is told that no coronavirus can be detected, when actually they are incubating the virus.
The University of Birmingham was one of several UK educational institutions that offered lateral flow tests to all its students before they returned home for the Christmas break. About 7200 took up the offer during the first week in December, and one-tenth of them, chosen at random, were also given a PCR test to verify the accuracy of the lateral flow tests.
The lateral flow tests indicated that 0.03 per cent of the students had the coronavirus. But in the smaller fraction who had a PCR test, the prevalence was much higher, at 0.86 per cent, according to figures put online this week.
It is unknown why the lateral flow test’s sensitivity was so low, says the university’s Jon Deeks, who was involved in the screening programme. One factor could be that students took their own nasal and throat swabs, and may have done this improperly, although they did do this under supervision, and many other bodies offering screening also use self-swabbing.
It also could be an artefact of the small numbers involved, says Louise Kenny at the University of Liverpool. The PCR positivity level came from 6 positive results among 710 students. “Statistically that’s meaningless,” says Kenny.
The tests are made by US firm Innova Group. A spokesperson for the company wrote by email that they are unaware of the data reported by the University of Birmingham. “The low level of efficacy being suggested is not something we recognise, suggesting a careful and considered examination of the methodology used would be advisable. When used properly, the Innova test is a highly effective tool in detecting infectious people and allowing for proper response to assist in the reducing of the spread of the SARS-CoV-2 virus.”
Birmingham’s coronavirus prevalence figure from the lateral flow tests was much lower than that seen at some other UK colleges. For instance, the University of Portsmouth found a prevalence of 0.2 per cent among its asymptomatic students, and the University of Reading, 0.4 per cent, according to a BBC report. But Edge Hill University in Lancashire found no positive cases from about 2100 students, according to the same report. It is unclear whether these other colleges asked students to swab themselves.
Mass screening using lateral flow tests is set to be introduced among all secondary school students in England, starting in January. “They certainly should not be used like this,” says Deeks.
Allyson Pollock at Newcastle University in the UK says mass screening using lateral flow tests shouldn’t have been widely introduced before being tested for accuracy in the community. “They have not been evaluated outside of a laboratory setting.”
A spokesperson for the UK’s Department of Health and Social Care said by email: “The country’s leading scientists rigorously evaluated the lateral flow test and confirmed the accuracy. Our evaluation work and the ongoing pilots are helping us to understand how lateral flow tests work in the field and how we may use them to help stop the spread of the virus.”
More on these topics: