Wikipedia 'foresees virus outbreaks'
Wikipedia page views can predict disease outbreaks nearly a month before official health advice, a team of US scientists says.
The Los Alamos National Laboratory team says people are searching online before seeking medical help.
They managed to forecast tuberculosis and influenza outbreaks four weeks in advance.
Other experts said they were "wary" about the value of using online searches to predict outbreaks.
Traditional disease surveillance involves collecting data from laboratory tests, calls to doctor's surgeries and tracking the number of people who visit health facilities.
The US team, writing in the journal PLoS Computational Biology, argued these methods were accurate, but slow and expensive.
Previous attempts at using the internet to predict disease outbreaks include Google Flu Trends.
The US scientists instead turned to Wikipedia and tracked page views between 2010 and 2013 on visits to disease-related Wikipedia pages.
They mapped the languages the information was written in, using this as an approximate measure for people's locations.
Their data was then compared with disease outbreak information provided by national health surveillance teams.
In eight out of 14 cases, there was a clear increase in page views four weeks before health officials declared an outbreak.
Their statistical technique allowed them to predict emerging influenza outbreaks in the United States, Poland, Japan and Thailand, dengue fever spikes in Brazil and Thailand, and a rise in tuberculosis cases in Thailand.
Lead researcher Dr Sara Del Valle said: "A global disease-forecasting system will change the way we respond to epidemics.
"In the same way we check the weather each morning, individuals and public health officials can monitor disease incidence and plan for the future based on today's forecast.
"The goal of this research is to build an operational disease monitoring and forecasting system with open data and open source code.
"This paper shows we can achieve that goal."
But it is not clear from the present study whether the model will be useable in countries where access to the internet is poor.
Dr Heidi Larson, from the London School of Hygiene and Tropical Medicine, said the Wikipedia research data was "compelling" but she would be "wary" about using it as a tool for predicting outbreaks of all disease.
She said that people needed to have internet access, be literate, and be familiar with Wikipedia and with the disease itself.
Dr Larson added: "There are different things that drive people to Wikipedia, sometimes a new piece of research can drive people to go online."
While online trends could provide valuable signals about disease, questions remained as to what extent you used the data to inform policy or for intervention, she added.
"I'm not sure how much Wikipedia is used in Africa," she said. "For issues like Ebola, I don't think people at the beginning of the outbreak in West Africa would have [been searching], because they wouldn't have had it [Ebola] before."