Two thirds of cases are missed, app expert says

The British government’s data on coronavirus cases is nonsense and most cases are undiagnosed, an epidemiologist says.

Professor Tim Spector in Sydney in 2016. Picture: Damian Shaw

The British government’s data on coronavirus cases is nonsense and two thirds of cases in Britain are undiagnosed, a leading epidemiologist has claimed.

Tim Spector, professor of genetic epidemiology at King’s College London, has criticised the British government’s refusal to class common symptoms as official indicators.

The government has listed a high temperature and a new, continuous cough as the primary symptoms of COVID-19. There is evidence, however, that a loss of taste or smell is a common sign of infection. Other countries, including the US, have recognised this as a symptom on more detailed lists.

The Australian Department of Health lists fever, coughing, sore throat and shortness of breath as possible symptoms.

Professor Spector has collected data from three million Britons on his team’s symptom-tracking app.

“The reason that we got a bit stuck in this country is we took the data from China and just instantly said, ‘OK, the disease only has two symptoms: it’s fever or it’s persistent cough.’ That meant we were missing about 60 per cent of cases,” he said.

“Only people with those two symptoms got tested and ended up on the statistics. All this governmental data on confirmed cases and how many people have recovered, it’s all nonsense.”

In partnership with the British Department of Health and Social Care, 10,000 app users are being sent COVID-19 tests each week, within the first day or so of developing symptoms.

Clear patterns have emerged from the data in the six weeks since the app went live and the team are close to being able to plot how the disease will progress, depending on the symptoms someone has on day one.

According to the data the virus may have arrived in Britain at the new year as many users reported symptoms in January. At the peak one in 10 people had the virus. The team estimated that at the end of April there were more than 300,000 symptomatic cases, a fall from a peak of more than two million at the start of the month.

Professor Spector said: “We’re able to allocate people into five or six groups at the moment that follow different patterns of symptoms at different time points. It’s not random.”

In one group are those whose symptoms are a sore throat and muscle pains, which then develop to include diarrhoea, stomach pains and fatigue.

People in another group start with a headache, which progresses to a cough and fatigue, then the cough gets worse, they develop shortness of breath and may need to go to hospital. This classification is important in determining which patients are high risk.

Alan McNally, professor of microbial genomics at the University of Birmingham, agreed that there were many other possible signs of coronavirus infection. “There are myriad less common symptoms attached to COVID: things such as COVID toe [a rash on the feet] and loss of appetite,” he said.

He defended the decision by the government to concentrate on a few, however, saying: “I don’t see where on earth you draw a line on symptoms and whether or not they are COVID.”

Users of the COVID Symptom Study app in Britain log how they feel daily, even if they are healthy. The most useful loggers are people who were in good health when they first logged on to the app and have since developed symptoms.

The Times