On February 16, 2012, Harvard Medical School held an International Conference on Digital Disease Detection. The conference discussed ways that the internet can aid in disease surveillance: the science of epidemiology.
Epidemiology is the study of the pattern and distribution of health events and disease outbreaks. In the original Greek, it means “the study of what is upon people”, even though it’s also used to study animal populations.
How Will Social Media be Used?
That discipline is one tool doctors and scientists use to establish foodborne illness outbreaks. Before the internet age, epidemiology was labor-intensive and slow moving, using surveillance, data, and statistics to discover cause-and-affect relationships. Now the power of the internet can be used to track epidemics in real time. This conference put social media in the spotlight, advocating internet use to help pinpoint and track disease outbreaks.
Fred Pritzker, an attorney who specializes in foodborne illness cases who attended the conference, said, “Anything that allows for easier and faster detection of foodborne illness outbreaks is a valuable tool. Even a few days advance notice could make the difference between life and death.”
Scientists are using Twitter, Facebook, and Google News to uncover foodborne illness outbreaks. Food Poisoning Bulletin has told you about using Twitter as a epidemiologic tool. And phone apps may be another tool in the dissemination of information.
Twitter was used during the January 14, 2012 norovirus outbreak at the Canadian University Press National Conference. More than 60 people were sickened in a few hours. Students used Twitter to report illnesses and help those who were sick, telling them where to get help.
Turning Data into Predictions
Considering that foodborne illnesses are very under reported, and most people affected don’t even see a doctor, the internet can boost public health efforts and communicate outbreaks and risks to the public. Researchers are studying social media to see if it can be used to predict disease outbreaks and if it’s reliable enough to be used as an epidemiological tool.
For instance, Google Flu Trends (GFT) uses data from search engines to estimate flu outbreaks. A recent study conducted in Baltimore and published in the Oxford Journal of Clinical Infectious Diseases showed a strong correlation between GFT cases and influenza emergency room visits. This tool can help emergency rooms plan for a sudden influx of patients.
This is interesting, but correlation only shows disease patterns; extrapolating to predictions is less certain. There is a lot of “background noise” on the internet. Useful tools will have to filter out that data and focus on true early warning signals.
That’s why researchers are planning more studies using Google Trends and Twitter to see if data from those applications can be used to forecast outbreaks. Considering that many populations have more access to the internet than they do healthcare, interactive maps such as HealthMap help inform the public about potential danger, giving people time to make changes and seek medical help.
For now, social media can be a valuable resource to decrease the time between an actual outbreak and a formal government announcement, which will help alert the public to potential danger. Mr. Pritzker added, “It’s important that we understand how data from social media can be mined for the benefit of public health rather than just selling products. If information gleaned from it allows for faster identification of public health threats, it will be an incredibly valuable tool.”
