We all want to forget about the grief and destruction caused by COVID-19. But, sadly, public health concerns such as major pandemics could be even more frequent in the future: A 2021 study analyzed data on previous disease outbreaks and found that the rate of novel pandemics breaking loose to infect humans is rising. What's more, the risk of outbreaks will increase three-fold over the coming decades.
“One may erroneously presume that one can afford to wait another 100 years before experiencing another such event,” said Gabriel Katul, a researcher at Duke University and one of the study's authors, in a statement. “This impression is false.”
The need to boost pandemic preparedness is clear. Fortunately, data scientists have taken up the mantle. They’re delving into the untapped potential of social media platforms and search engines to predict the trajectories of infectious diseases, which might help authorities more effectively curb the spread of future pandemics, and also better manage seasonal health problems such as influenza.
Studies are highlighting how aggregated data from platforms like Google and X, formerly Twitter, can reveal that the volume of search queries related to flu-like symptoms (or keywords associated with viral illnesses) and foreshadow regional outbreaks. The aim is to exploit these digital footprints as early indicators of impending health crises.
A 2023 study, for example, examined 1.3 million posts on X using machine learning techniques to analyze the content for potential mentions of Lyme disease in the U.S. The computer algorithm was able to determine with 90 percent accuracy which tweets were talking about Lyme disease.
How Are Scientists Leveraging That Data?
It’s this kind of data that scientists want to use to forecast potential disease spread. In fact, they already are, but leveraging that data is more complicated than simply showing that lots of people are tweeting about — or searching Google for — certain relevant keywords.
Computer models need to factor in all sorts of additional variables, such as where people tend to move around within a city or a country, the incubation time of a given disease — how long it takes to become infectious after being exposed to the pathogen — and vaccination rates. It’s a complex picture, but studies are beginning to show results.
In a study published in 2022 in Applied Soft Computing, researchers retrospectively predicted the spread of COVID-19 from social media posts with more than a 90 percent accuracy rate.
Read More: How AI Will Change Medicine Forever
Can This Work Help Prevent Future Disease Outbreaks?
The key challenge is now taking these models from proof of concept — where scientists analyze previous outbreaks to show the models work — and applying them to predict and prevent future diseases. That’s much harder, because we often don’t know all we’d like to know about a novel disease, such as its incubation time, before the outbreak.
The predictive power of these models therefore hinges on sophisticated algorithms and machine learning techniques that sift through massive amounts of data to provide a best guess. By discerning patterns and nuances within these datasets, researchers hope to unlock the potential to forecast health issues before they show up in official data.
Read More: What If We Tolerated Diseases?
Still, decision makers will need to accept these new technologies first. Even then, ethical considerations regarding user privacy pose serious challenges. Several studies have highlighted the need for responsible data usage and stern ethical frameworks to ensure that user privacy is respected in the pursuit of harnessing of big data for the benefit of public health surveillance.
As technology continues to evolve and shape our world, the fusion of big data analytics and public health offers a promising avenue to revolutionize disease surveillance and response strategies. So, there is reason for optimism, even if pandemics do indeed be more common in the future.
Necessity is the mother of invention, after all.