Skip to main content

Go Ahead, Crowdsource Your Symptoms

Online health forums actually provide decent information, according to a new study.
  • Author:
  • Updated:
(Photo: Gaudilab/Shutterstock)

(Photo: Gaudilab/Shutterstock)

It seems we fret over the unhealthy consequences of Googling our health problems almost as often as we actually Google our health problems. But when it comes to providing health tips, some corners of the Internet are less dangerous than we might think: According to a new study in the Journal of Medical Internet Research, online message forums provide safe—if incomplete—information more often than not.

"There's this complete misconception about the Internet being totally unregulated and full of inaccurate information," says Jennifer Cole, a researcher in the Royal Holloway University of London computer science department and co-author on the new research.

Cole's study stemmed from questions surrounding 2014's Ebola epidemic: Without enough medical experts around, where do people learn about disease? One obvious answer: the Internet.

"There's this complete misconception about the Internet being totally unregulated and full of inaccurate information."

To probe the quality of Web-based health discussions, Cole's team pulled advice from three forums popular in the United Kingdom: the health-specific site Patient; parenting site Mumsnet; and general interest juggernaut Reddit. The researchers chose exchanges that revolve around recognizing and treating chicken pox, diabetes, or HIV—illnesses discussed on all three sites. They then distributed the advice to be rated mostly by doctors, as well as some people who had long-term experience with HIV or diabetes.

Altogether, the doctors and other raters were four times as likely to grade discussion threads as containing high-quality information, compared to low-quality information. In those discussions where raters flagged low-quality information, the problem usually wasn't amateur tips that would spawn bad health-care choices; it was an absence of the best recommendation, like following up with a medical professional.

Sifting through the Web forums, Cole and her colleagues saw a mostly supportive environment of commenters who tried to provide accurate, or at least reasonable, information. "Particularly in these health forums, we see a much more naturalistic and real-world interaction going on than the stereotype of Internet forums full of trolls," she says. (In the future, Cole hopes to investigate the role moderators play in this respectful discourse.)

Cole plans on taking a deeper dive into her team's finding that the worst information was provided for chicken pox, which had lower stakes for bad advice compared to diabetes and HIV—suggesting that moderators, commenters, or both put forth more accurate information for more threatening illnesses. "Accuracy" can, of course, vary by cultural background: Americans consider vaccination to be the normal defense against chicken pox, whereas children are rarely vaccinated for chicken pox in Britain. Thus a call for vaccination may look alarmist to British commenters while reading as standard to their American counterparts.

Across different recommendations, the researchers found that the value of actually seeing a doctor was rarely questioned. In fact, Cole says commenters often framed their advice as complementary toward medical professionals' input. This peer-to-peer information may help fill in gaps while people wait to hear back from their doctor, or to find a doctor who can serve them at all.

In times of need, then, support systems provided by online health forums can be more friend than foe.


Quick Studies is an award-winning series that sheds light on new research and discoveries that change the way we look at the world.