Data Mining Tracks Drugs' Side Effects; Social Media Can Identify Problems with Drugs, Researchers Say

Article excerpt

RICHMOND, VA. - Experiencing nausea, headaches or other side effects from prescriptions or over-the-counter medicines?

Researchers say tweeting about it or posting your concerns online could one day help alert drug companies and federal regulators to problems more quickly - potentially saving lives and money.

The project at the University of Virginia and West Virginia University, still in its infancy, capitalizes on the idea that many companies - pharmaceutical and otherwise - already use the Internet to get consumer product and service feedback.

Sifting through innumerable posts on Twitter, Facebook, online message boards and blogs, the researchers will search for early warning signs of adverse drug reactions and interactions normally reported to the Food and Drug Administration and pharmaceutical companies through official channels by consumers and doctors.

People are increasingly turning to the Internet to find out what's ailing them, complain about their symptoms or read up on personal health issues, said Ahmed Abbasi, a professor of information technology at U.Va.'s McIntire School of Commerce. He says using data from social media could help modernize drug surveillance and have major public health, safety and business impacts.

Funded by a $130,000 grant from the National Science Foundation, the project builds on earlier work analyzing online posts from 2000 to early 2012 for mentions of 20 drugs. Researchers say the earlier effort detected adverse drug reactions - in some cases years earlier than current methods.

Abbasi said their data mining of online sources was able to identify patients experiencing tendon ruptures after using the popular antibiotic Cipro at least two years before the FDA issued its most urgent "black box" warning for the drug.

Despite the rigorous testing and clinical trials before new drugs are on the market, sometimes side effects don't show up until they are used by a large number of people. …


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.