Google, Twitter and the internet might be helping researchers predict where there might be a coronavirus hot spot in the near future.
Researchers at Harvard have created a new algorithm that uses Twitter posts, Google searches and new location data to predict an increase where there might be an increase in COVID-19 symptoms with a 14-day notice.
In a paper, the researchers said this method could work “as a thermostat, in a cooling or heating system, to guide intermittent activation or relaxation of public health interventions.”
“In most infectious-disease modeling, you project different scenarios based on assumptions made up front. What we’re doing here is observing, without making assumptions. The difference is that our methods are responsive to immediate changes in behavior and we can incorporate those,” said Mauricio Santillana, director of the Machine Intelligence Lab at Boston Children’s Hospital and an assistant professor of pediatrics and epidemiology at Harvard.
Separately, University College London used new research to focus on people’s loss of smell — a common symptom for the novel coronavirus, according to OneZero. The model will look at search phrases such as “can’t smell” and “lost my sense of smell” to determine where there might be an increase in these symptoms, allowing the researchers to identify potential hot spots early.
For this second study, there wa an uptick in searches in New York and Chicago — two of the hardest-hit cities in the beginning days of the COVID-19 pandemic, according to One Zero.
Houston and Dallas, Texas, saw spikes in June. Those cities became hot spots as well in recent days, according to One Zero.
According to The New York Times, Google started estimating doctor visits for the flu back in 2008 by monitoring how often phrases such as “feeling exhausted,” “joints aching,” “Tamiflu dosage” and many others were searched.
The algorithm didn’t really work since “it continually overestimated doctor visits, later evaluations found, because of limitations of the data and the influence of outside factors such as media attention, which can drive up searches that are unrelated to actual illness.”
Now, researchers use real-time analysis to help identify these habits, painting a path for a better understanding of what people are looking for in terms of medical care.