COVID-19: Could your earliest symptoms predict how ill you’ll get?

By Sean Fleming

July 22, 2020


The symptoms experienced in the first few days of a COVID-19 infection could help predict the course a patient’s illness will follow.

That’s one of the initial conclusions from a team of researchers from King’s College London in the UK. Their research has identified six possible sub-divisions of COVID-19, using machine learning to analyse data from a symptom-tracking app.

These findings – from a pre-print paper not yet peer-reviewed – reflect how the thinking regarding the disease is evolving and how technologies are being leveraged to fight the disease.

Symptom clusters

Most people can name some key COVID-19 symptoms, thanks to media reports or — unfortunately — personal experience. The most well-known symptoms include fever, shortness of breath, fatigue, muscle aches, headaches and the infamous persistent cough.

As more has been learned about the illness, the list of symptoms has grown. Loss of sense of taste and smell (ageusia and anosmia, respectively) was detected anecdotally before being accepted by medical bodies around the world.

The percentage of confirmed coronavirus cases reporting olfactory dysfunction.
Image: Centre for Evidence-Based Medicine

The King’s College team recently scrutinised data from around 1,600 people with confirmed COVID-19 infections in the UK and US. Each had logged symptoms using an app during March and April. A second dataset of around 1,000 app users from the UK, US and Sweden who had logged their symptoms during May was also examined.

The researchers say they have been able to group symptoms into six divisions, which they say could indicate how unwell a patient could become.

Their six suggested clusters are:

1. Flu-like with no fever: Headache, loss of smell, muscle pains, cough, sore throat, chest pain, no fever.

2. Flu-like with fever: Headache, loss of smell, cough, sore throat, hoarseness, fever, loss of appetite.

3. Gastrointestinal: Headache, loss of smell, loss of appetite, diarrhoea, sore throat, chest pain, no cough.

4. Severe level one, fatigue: Headache, loss of smell, cough, fever, hoarseness, chest pain, fatigue.

5. Severe level two, confusion: Headache, loss of smell, loss of appetite, cough, fever, hoarseness, sore throat, chest pain, fatigue, confusion, muscle pain.

6. Severe level three, abdominal and respiratory: Headache, loss of smell, loss of appetite, cough, fever, hoarseness, sore throat, chest pain, fatigue, confusion, muscle pain, shortness of breath, diarrhoea, abdominal pain.

Severity indicators

The six categories represent a spectrum of breathing difficulties. Understanding this range could help with clinical management and matching patients with the right care effectively and efficiently.

Analysing patient-provided information on symptoms and their outcomes, the research team says that while 16% of Group 1 patients were admitted to hospital, almost half of those in Group 6 were.

They also found that patients in 4, 5 and 6 tended to be older, and were more likely to have pre-existing conditions ranging from diabetes to obesity.

The specific combination of symptoms reported by a patient was found to be a potential indicator of whether they would become severely ill.

While more research must be done, the work underscores the importance of institutions around the world contributing to the wider pool of global COVID-19 knowledge.

The coronavirus crisis has sparked a range of global collaborations including the COVID Action Platform from the World Economic Forum. So far, more than 350 public bodies and over 850 private organisations around the world have joined the Platform, collaborating on 35 projects ranging from healthcare delivery and vaccines to supply chains and economic support.

The research also demonstrates the power of new technologies in aiding medical research. Said Sebastien Ourselin, a professor of healthcare engineering at King’s College London and a senior author of the study: “Being able to gather big datasets through the app and apply machine learning to them is having a profound impact on our understanding of the extent and impact of COVID-19, and human health more widely.”

This article is curated from the World Economic Forum.

About the author
Inline Feedbacks
View all comments
The Mandarin Premium

Insights & analysis that matter to you

Subscribe for only $5 a week


Get Premium Today