Researchers Track Drug Reactions Using Social Media

Researchers are turning to Twitter, blogs, and online discussions to track adverse drug reactions.

According to West Virginia University’s Research Highlights and a September 21, 2012, posting on ScienceDaily, the study results could determine whether a reaction database could be established for medical professionals and pharmaceutical companies. In addition, it could also be useful for establishing drug-drug interactions that cannot be pinpointed through testing or clinical trials.

Researchers from the University of Virginia and West Virginia University’s information technology departments lead the project. It will use computer algorithms to detect patterns and relationships within social media that indicate adverse drug reactions.

The system allows researchers to catalogue and monitor complaints until there are enough to warrant FDA scrutiny. A professor at West Virginia School of Pharmacy who is also director of the West Virginia Center for Drug and Health Information, Marie Abate, will handle newly published adverse drug reaction reports, which will help determine the similarity of reports made through social media.

“The idea is for drug manufacturers, federal regulatory agencies, health care professionals and the public to be aware of such potential problem s for a given drug, and perhaps to take a deeper look at the drug,” Donald Adjeroh, professor of computer science at West Virginia University, told the university’s press office.

The monitoring eliminates the reaction reports typically given to physicians and imparted to the FDA, or the consumer-issued reports collected through the FDA website. The study, in particular, could help determine whether reaction reporting could occur more quickly by using social media.

According to West Virginia University’s research update, a preliminary study with 20 drugs detected adverse reactions 80% of the time, and much earlier than FDA-issued warnings. More than 50% of adverse reaction cases were detected 3 years before FDA warnings, Adjeroh told the university.

The preliminary research analyzed websites, blogs, Web forums and social media sites from 2000 to 2012, and identified hundreds of thousands of adverse drug reactions, ScienceDaily reported.

A $130,000 National Science Foundation grant will fund the research.

Although researchers note the potential for over or misreported side effects, they believe the system will generate enough information to reduce mistakenly reported information.

“We have used control cases to check how often the method reported signals when none was expected,” Adjeroh told West Virginia University. “Ultimately, the system aggregates various pieces of information from online sources to generate hypotheses about potential adverse drug reactions. These are expected to be further studied by interested parties, such as regulatory agencies or the drug manufacturers.”