Testing has been at the forefront of the international response to combat coronavirus, but for some countries, notably the UK, it has been the source of much frustration and confusion.
Speculation surrounds how effective the coronavirus tests are.
Researchers from the UK’s University of Bristol explain the accuracy of the two main types of coronavirus test, and why it matters for policymakers.
Coronavirus testing has been the subject of huge interest, frustration and confusion. The UK has come under worldwide criticism for its lack of mass testing, despite the director general of WHO encouraging countries to “test, test, test”. Health Secretary Matt Hancock announced that the UK now aims to test 100,000 people a dayfor COVID-19 by the end of April.
There are two main types of COVID-19 tests. Swab tests, which usually take a sample from the throat or nose, to detect viral RNA. These determine if you currently have COVID-19. Blood tests, which detect antibodies, can determine if you have had COVID-19, and are therefore immune.
No test is 100% accurate. Although tests can perform well in ideal laboratory conditions, in real life lots of other factors affect accuracy including the timing of the test, how the swab was taken, and the handling of the specimen.
Early on in the novel coronavirus outbreak, doctors started reporting cases of people who had coronavirus which had been missed by swab tests – also known as “false negatives”. We don’t know for sure how often these false negatives occur in the UK, but evidence from China suggests up to 30 out of every 100 people with coronavirus might test negative.
The meaning of a test result for an person depends not only on the accuracy of the test, but also on the estimated risk of disease before testing. This was described mathematically by Thomas Bayes and later explained by Siddhartha Mukherjee as the law that “a strong intuition is much more powerful than a weak test”.
Let’s explain this with an example. Jane works for the NHS as a receptionist in a GP surgery in London, in an area of high rates of coronavirus infection. After noticing a loss of smell for a few days, she wakes up one night feeling shivery, with a dry cough. She checks her temperature to see it’s 38.5°C. However, after getting a swab test, the result comes back negative for COVID-19. Great news. Or is it?
Doctors use their experience to recognise patterns in symptoms, risk factors, and signs to estimate the likelihood of infection before testing. This is known as “pre-test probability”. Based on her symptoms, the probability that Jane has COVID-19 will be high – perhaps 80%.
However, let’s say 100 people experiencing symptoms like Jane have a swab test. Of these people, 80 actually have COVID-19. A positive test means we can be pretty certain someone has COVID-19. But if the test misses 30% of those 80 people with COVID-19, this means an estimated 24 out of 100 people will have a “false negative” test result. This means these people might go back to work and unknowingly spread coronavirus to others.