Can Google Predict Who Catches the Flu?

In the last ten years, the Internet has drastically transformed the way people interact, learn, and seek information. Each and every day millions of people search for health-related information on the web. According to Pew Research, 72% of internet users say they looked online for health information within the past year and 77% say they used a search engine like Google to conduct their search.

With the flu epidemic racing across the United States, it’s likely you have turned to Google to search for information about your symptoms or advice on how to avoid getting sick. This year’s flu season has been particularly brutal. High levels of influenza-like symptoms have been spotted in over 24 states, with an additional 13 reporting moderate flu activity. The elderly and young children have been especially hard hit: 53 pediatric deaths have been reported and over 26,00 patients have been hospitalized. The Internet is reflecting the epidemic. Get this: Mashable reports 626,379 social media users reported having the flu in the last two months, a 2,336% increase from previous months.

As the world becomes increasingly mobile and connected, millions more will turn to search engines when they feel unwell. To piggyback this trend, in 2008 Google launched “Flu Trends” which provides up-to-date estimates of influenza activity in nearly 30 countries across the world.  By aggregating data on web searches and queries conducted about influenza, the web service uses an algorithm to make predictions about flu activity and influenza outbreaks.

So how does Google Flu Trends work?

Google found a close relationship between the number of searches for flu-related topics, such as flu symptoms, flu prevention, and flu treatments. By tracking the frequency and location of flu-related search queries, the search engine giant is able to put a pattern to the presence of flu-like symptoms in the population and track how the flu circulates through different regions.

Findings are then compared against historic baseline public health data of influenza activity and estimated as minimal, low, moderate, high, or intense. The 2012-2013 flu season reveals that activity is “intense”, the worst in the tracker’s seven-year history.

Technology and disease forecasting 

Google has been at the forefront of predicting just how bad this flu season is and preparing the public and health officials to respond. Flu Trends is a new and novel form of surveillance data for the modern era that has been remarkably accurate in predicting flu outbreaks in the US since its launch, however debate as to its accuracy does exist. Some argue that Google’s algorithm is faulty and may not be as reliable as health agency data, however other reports show Google Flu Trends can predict regional outbreaks of the flu up to 10 days before they are reported by the Centers for Disease Control and Prevention.

Having the ability to identify influenza activity early is a key benefit Google’s tool provides. On an individual level, knowing that the flu is active in your area can help you take preventative actions to avoid getting sick, such as frequent hand-washing and applying a homeopathic nasal solution daily. Early prevention and control measures can reduce the spread of the outbreak and reduce the number of people affected.