A study review does not support the roll out of coronavirus antibody tests.
Antibodies are released by the immune system in response to an infection. If picked up by a blood test, it implies an individual has fought off the coronavirus and may have some protection against it should they be exposed again.
Boris Johnson described the tests as potentially “game changing”, while health secretary Matt Hancock said an accurate kit may be an “important milestone”, safely freeing survivors of the infection from the confines of social distancing.
The government bought 10 million of the kits from the pharmaceutical giants Roche and Abbott after Public Health England (PHE) studies suggested their effectiveness.
NHS England and NHS Improvement wrote to trusts on 25 May asking them to “offer antibody testing at short notice and ramp up capacity to thousands of samples a day”.
Experts have warned, however, the coronavirus is novel. It is therefore unclear how long immunity lasts or even if it exists at all.
After analysing studies on the tests, scientists have warned their “confidence in the evidence is limited for a number of reasons”, with timing seemingly being everything.
‘We can’t trust any of the data’
Once an infection has passed, memory antibodies circulate at low levels in the blood.
If the infection is identified again, the immune system ramps up production of these proteins, preventing the virus from taking hold.
Government officials have hailed antibody tests a way back to normality, even floating the idea of “immunity passports”.
The scientists behind the review, however – led by Professor Jon Deeks from the University of Birmingham – were unsure of their accuracy.
“We don’t have much data and we can’t trust any of it”, he previously said.
To learn more, the team looked at 54 studies with nearly 16,000 blood samples. Of these studies, 46 were made up of people in hospital with suspected or confirmed coronavirus.
The findings “come mainly from 38 studies that provided results based on the time since people first noticed symptoms”, wrote the scientists.
Data were only available for 27 tests, a small fraction of the more than 200 kits on the market.
The studies included in the review were carried out before mid-April and did not include PHE’s trials into Roche’s or Abbot’s kits.
‘Antibody testing early in the disease is unreliable’
Overall, the scientists wrote their “confidence in the evidence is limited for a number of reasons”.
“In general, studies were small, did not use the most reliable methods and did not report their results fully,” they added.
Most of the participants in the studies were in hospital and therefore likely had severe disease.
“This means we don’t know how accurate antibody tests are for people with milder disease or no symptoms”, wrote the scientists.
The results were often released preliminarily and therefore not subject to the rigorous analysis that comes from being in a peer-reviewed journal, they added.
Overall, the scientists found the antibody tests carried out one week after symptoms emerged detected only 30% of people who had the infection.
Accuracy increased in week two to 70% and peaked in week three at more than 90%. Little evidence was available beyond that.
“The review shows antibody tests could have a useful role in detecting if someone has had COVID-19, but the timing of when the tests are used is important,” wrote the scientists.
COVID-19 is the respiratory disease that can be triggered by the coronavirus.
To put the results into context, imagine a sample of 1,000 people of whom 200 (20%) have had the coronavirus, which is typical of workers in a hospital where infectious patients were treated.
Of these, 193 people would receive a positive test result, but 10 (5%) of those would not have had the coronavirus – a false positive.
Eight hundred and seven people would get a negative result, of whom 17 (2%) would have had the coronavirus – a false negative.
“Although this might sound impressive, this could lead to a large number of false positive results if such a test was applied to entire populations”, said Dr Hayley Jones from the University of Bristol.
Professor Eleanor Riley from the University of Edinburgh added: “It is very clear from this analysis that antibody testing early in the course of disease is unreliable and thus antibody tests cannot (and should not) replace virus detection for diagnosis of acute cases except where the time-course of disease is already well advanced.
“This is a useful and important clarification”.
The review authors will update their results over the next few months as new evidence comes in.
“We should look at the totality of the data and not get excited by small studies,” Professor Deek said during a Science Media Centre press briefing.
Antibodies ‘do not guarantee immunity’
This is not the first time the validity of coronavirus antibody tests has been questioned, with a separate group of scientists previously casting doubt on how “game changing” they really are.
Antibodies do “not guarantee immunity” against hepatitis C, flu or the virus responsible for bronchiolitis – inflammation and congestion of the small airways.
“What people really want to know from these tests is, ‘am I safe from infection?’,” Dr Al Edwards from Reading University previously said. “These tests, at the moment, can’t answer that.”
Assuming coronavirus antibodies do ward off a second infection, how long that lasts is unclear.
A measles vaccine offers lifelong protection, while the most effective jab against cholera leads to antibodies being produced for just five years, hence the need for boosters.
The coronavirus is one of seven strains of a class of viruses that are known to infect humans.
Others include Severe acute respiratory syndrome (Sars) and Middle East respiratory syndrome (Mers), which killed 774 and 858 people in their 2002/3 and 2012 outbreaks, respectively.
“People who were infected with Sars and Mers produced antibodies against these illnesses following infection, for up to three years in some cases”, Dr Tom Wingfield from the Liverpool School of Tropical Medicine previously said.
“However, it is not clear whether the presence of these antibodies means that a person is immune to a repeat infection.
“I share the concerns that antibody testing is being rolled out before we fully understand the immune response to [the coronavirus], what the results of these tests may mean and what actions to take based on test results.
“Another critical unknown is what advice to give someone who has a positive or negative antibody test and how the result and such advice might influence their behaviours, including social distancing, and thereby impact upon [the coronavirus’] transmission”.
Coronavirus: what happened today
Read more about COVID-19
How to get a coronavirus test if you have symptoms
How easing of lockdown rules affects you
In pictures: How UK school classrooms could look in new normal
How public transport could look after lockdown
How our public spaces will change in the future