Medical Researchers Tune Into the Internet Buzz

Medical Researchers Tune Into the Internet Buzz – WSJ.com

From the 16 April edition of the Wall Street Journal

Looking for medical information on Internet message boards can be risky for consumers. Some of it is confusing, misleading or downright wrong. But for medical researchers, all that chatter can yield some valuable insights.

Scientists from the University of Pennsylvania, for example, are mining message boards and Twitter feeds to see what breast-cancer and prostate-cancer patients are saying about herbal and nutritional supplements—including whether they take them and why and what side effects they encounter.

“People are often hesitant to talk to their doctors about herbs and supplements. But they do talk with other people, especially in an anonymous setting like a discussion board,” says principal investigator John Holmes, an epidemiologist and medical-information specialist. Even if there is no scientific evidence to support what people post, he says, “it’s useful to identify areas that would merit further study with all scientific rigor.”…

….

Chatter on the Web also can serve as an early warning sign of adverse events linked to drugs or medical devices. “We see patient conversations on the Internet as the largest post-marketing study ever,” says Michele Bennett, chief operating officer of Wool Labs, a business-intelligence company founded in 2007. The Wayne, Pa., firm can search the entire Internet for conversations that shed light on patient beliefs, buying patterns or decision making—whatever its clients, many of them drug companies, are seeking.  Wool Labs also can search Web chatter retrospectively to see how attitudes changed over time…

…

Analyzing Web conversations does raise ethical and privacy issues; people who talk candidly about their medical problems online may not realize it is a public forum. That is why the Penn researchers mine only discussion sites that require participants to register and explicitly state in their terms of use that any information posted will become public. The programs also filter out any posts or tweets placed by “bots” that are advertisements in disguise; containing a URL to another site is a telltale sign

The team also devised an “anonymizer” program that scrubs out any names, locations or other identifiers…

Related articles

• Social media and medical research, a story from Oz (simondenegri.com)