Google Searches as Early Warning For Disease Outbreaks
Originally Posted at The Wall Street Journal
Last summer, for a month before Canadian officials announced an outbreak of listeriosis that would kill some 20 people, Google searches seeking information on the disease were on the rise.
That finding, published this week in Canadian Medical Association Journal, is the latest sign that public-health types are trying to figure out how to mine Internet search data as a potential early-warning system for disease outbreaks.
Think of it as a wonkier example of Google Flu Trends, the tool Google.org rolled out last fall to mine Americans’ search patterns in an effort to track flu outbreaks.
There is a certain, basic logic to all this. You find out someone in your family has listeriosis, you go home and Google it. Multiply that across an outbreak, and it will add up. Interestingly, researchers found that searches for “listeria” — a less technical term for the disease, which is caused by eating tainted food — rose only after the official announcement, perhaps in response to media stories that used the term.
The authors point out some important unanswered questions about how public-health officials might actually use data from the likes of Google. There would likely be lots of false alarms, as factors other than a true disease outbreak prompted search spikes. And it’s unclear just how much of a spike from baseline levels would warrant further investigation.
What’s more, the potential false alarms might spook the public. Google already offers Google Trends, which lets anybody see how the volume of searches for a given term changes over time.
The paper was written by researchers from Harvard Medical School and the Ottawa Health Research Institute. The Harvard author has received research funding from Google.org as well as the National Institutes of Health.
