22 May 2020
University of Liverpool epidemiologist urges authorities to exercise caution when interpreting antibody tests due to potential implications of false positive results.
James Westgate
Job Title
One of the UK’s leading veterinary epidemiologists has warned against the risk of overconfidence in new coronavirus antibody tests.
Earlier this month, two pharmaceutical firms – Roche and Abbott – both had antibody tests approved by the UK Government as it looks to ease lockdown and get society moving again.
Both tests are highly accurate, with specificity rates of between 99.7% and 99.8%, and it is hoped their increased roll-out will give health officials a clearer picture of how the COVID-19 infection has spread.
This information will guide how and where lockdown restrictions can be phased back as Boris Johnson and his ministers attempt to kick-start the mothballed economy while doing all they can to minimise the risk of another spike in infections.
But Matthew Baylis, chairman of veterinary epidemiology at the University of Liverpool, believes authorities must exercise caution when interpreting diagnostic tests as the number of “false positive” results could have big implications in the fight against such a newly emerged disease.
He said: “It [the Roche test] is undoubtedly a very good test. But it is not a perfect test, although its specificity is very high – 99.8%.
“That leaves a small window that would mean two out of every thousand negative samples would show up as positive. That is a small number, but it is not zero.”
Prof Baylis added: “If many of us have already had COVID-19 – and when I say many, I mean even a few per cent of us – then the vast majority of positive test results will be those true positives and these false positives will be negligible.
“But it could be possible that not many of us have had COVID-19 at all – we just don’t know yet for sure. We have this problem that when a disease is rare then those false positives become really significant.”
