Advertisement

Greek fiasco highlights the problem with false positives

Cruise lines are relying on testing to prove they are safe - getty
Cruise lines are relying on testing to prove they are safe - getty

Picture the scene. For months you’ve been looking forward to this holiday. The tantalising prospect of a week in the Med has kept you sane throughout the doom and gloom of the Covid saga.

It wasn’t cheap – you even plundered your precious savings to pay for it. But it will be worth every penny.

The day before you’re due to depart, however, that dream holiday turns into a nightmare. Your destination demands evidence of a negative Covid test, taken no more than 72 hours before you arrive, and the results are in. You’re positive. Your holiday is cancelled, with little hope of a refund, and you’re obliged to self-isolate for 10 days or face a hefty fine.

This scenario would be devastating – but how much more frustrating would it be if you didn’t actually have the virus at all, and your “positive” result was a false positive?

A cautionary tale in the Aegean

Our scenario is imaginary, but surprisingly plausible. This week hundreds of passengers on board a Tui cruise, the first sailing for the company in the post-lockdown era, had their holidays interrupted after a dozen crew members tested positive for Covid. The captain halted the voyage, ordered all passengers to self-isolate in their cabins, and moored the vessel off the Greek island of Milos.

The 12 staff members, however, all of whom were asymptomatic, were retested. Low and behold, all turned out to be free from the coronavirus after two further checks, a PCR (Polymerase Chain Reaction) test performed by Tui and an antigen rapid test carried out by Greek authorities.

One of the Tui Cruises fleet
One of the Tui Cruises fleet

It isn’t an isolated example for the cruise industry. Last month a SeaDream passenger with no symptoms was tested on arrival in Denmark and told they had Covid. When tested a second time, however, the result was negative. A “positive” test during an UnCruise sailing, meanwhile, also appeared to be a false alarm.

Beyond cruises, a single weekend last month saw 77 NFL players and team personnel handed positive results that turned out to be false.

Following the high-profile outbreak on Diamond Princess, holidays at sea are under more scrutiny than most, so cruise lines are relying on testing to prove they are safe. Last week US authorities said testing will be compulsory for passengers prior to boarding if major cruise lines want to restart operations.

But it isn’t just cruise lines. An increasing number of countries, including Antigua, Barbados, Cyprus and St Lucia, are now demanding evidence of a negative test before they’ll permit travellers to enter. So what if the problems encountered by Tui and SeaDream are commonplace? Will ordinary travellers be retested, or only crew? How many holidays will be ruined by false positives?

The false positive problem

The issue with false positives has not been widely reported, but it is starting to gain traction. Matt Hancock, the Health Secretary, was quizzed by Talk Radio presenter Julia Hartley-Brewer about it earlier this month, with Mr Hancock appearing to suggest the problem was negligible because the false positive rate is “less than one per cent”. However, what he didn’t appear to understand was that the false positive rate does not refer to errors among all positive results, but to errors among all tests carried out.

For the sake of simplicity, let’s say the false positive rate is 0.5%. Currently, the ONS reckons around one in 500 people in England (or 0.2% of us) have Covid. This means that if you test 1,000 people at random (on, say, a cruise ship departing from Southampton – were such a thing possible) you will (hopefully – false negatives are also an issue) catch those two real positives, but also five false positives. In other words, 71 percent of your “positives” would actually be false. When explained in such terms, the problem with using testing to unlock travel is obvious.

The issue goes well beyond holidays

Our scenario considers false positives and travel, but the issue is far bigger than that. It seems we’re relying on testing to “beat” the pandemic and get life back to “normal”. Boris Johnson wants to drop £100bn – more than half the annual NHS budget – on his ambitious ‘Project Moonshot’, which would see 10 million Britons tested every day. But with a false positive rate of 0.5% that would mean 50,000 of us might be told to self-isolate on a daily basis when we don’t even have Covid.

Accuracy of 99.5% sounds good when you’re checking sick people with a high chance of having the disease, but when you’re testing hordes of symptomless people at random, it isn’t so hot.

Writing for The Spectator, Professor Carl Heneghan, director of Oxford’s Centre for Evidence-Based Medicine, explains: “At very low prevalence, the proportion of people with infection falls and the numbers of falsely misdiagnosed increases. If Covid-19 completely disappears, then no one will be infected. However, people would still be wrongly diagnosed as positive and the official data would show a national Covid-19 prevalence. This is why understanding the accuracy of tests in the population that they are applied to matters: going off current testing practices and results, Covid-19 might never be shown to disappear.”

The solution for travel

Suggesting we follow the lead of Turkey (which merely screens arriving passengers for obvious Covid symptoms, and tests those who exhibit them) or Sweden (which has no restrictions in place – but has banned entry to non-EU and non-EEA citizens arriving from outside of the EU) would be wishful thinking of our safety-first government.

Therefore The Telegraph is urging Boris and co to replace quarantine restrictions with testing for arrivals from high-risk countries, a measure that could negate the need to self-isolate – or at least reduce the quarantine period to a more manageable five days. Given the mounting evidence that false positives are a serious problem, however, it seems logical that governments, cruise lines and other holiday companies that require a negative test result for travel ought to guarantee a follow-up test if the first result is positive. That would begin to address the problem for travellers. Making Project Moonshot feasible might be a trickier task.