Google data can help forecast coronavirus peak

University College London researchers find that internet search data could assist with predicting public health emergencies

Powered by automated translation

Internet data on coronavirus-related symptom searches through Google could help to predict a peak in infections about 17 days in advance, researchers said.

A team from University College London said that search data could complement public health surveillance tracking for the coronavirus and new infectious diseases.

The study, published in Nature Digital Medicine, included data from countries such as Australia, Italy, South Africa, the UK and US.

Lead author Dr Vasileios Lampos said that online search activity has already been used to track the spread of flu and that the research provided a new set of tools to track the coronavirus.

“We have shown that our approach works on different countries irrespective of cultural, socioeconomic and climate differences,” he said.

“Our analysis was also among the first to find an association between coronavirus incidence and searches about the symptoms of loss of sense of smell and skin rash. We are delighted that public health organisations such as Public Health England have also recognised the utility of these novel and non-traditional approaches to epidemiology.”

Search queries such as "blue face", loss of smell or appetite, "pink eye" and shortness of breath related to “the top-five most impactful symptoms with regards to estimating confirmed cases”, the study said.

Co-author Prof Michael Edelstein, of Bar-Ilan University, near Tel Aviv in Israel, said early detection offered the best chance of tackling health emergencies such as the coronavirus.

“Using innovative approaches to disease detection such as analysing internet search activity to complement established approaches is the best way to identify outbreaks early,” he said.

“We can at least use the plethora of data sets around Covid-19 for further experimentation and validation of such techniques in an attempt to complement current epidemiological approaches and be better prepared for the next pandemic.”