Department of Biomedical Informatics

Home / Degree Programs / Academic Units / Department of Biomedical Informatics / Can Tweets and Posts Tell Us About Adverse Drug Events?

Can Tweets and Posts Tell Us About Adverse Drug Events?

Research Topic: 
Clinical Informatics
Graciela Gonzalez
Daniel Malone (University of Arizona)
Karen Smith (Regis University)

Research focuses on how to best use information about adverse drug reactions that is posted by patients and their caregivers on web sites and social networks such as, and

Drugs undergo extensive testing in animals and clinical trials in humans before they are marketed for widespread use. Pre-market testing produces reasonably high quality information about the efficacy of the drug as a treatment for the condition for which it was approved, but gives a very incomplete picture of the drug’s safety. Clinical trials done during the pre-market phase of drug testing are limited in both duration and scope involving a sample-size patient group selected by inclusion/exclusion criteria reflecting specific subject characteristics related to demographics, specific diagnosis, and other factors. It is only after a drug is marketed and used on a more widespread population and over longer periods of time that it is possible to identify other more serious side effects., Post-marketing surveillance currently relies on voluntary reporting to the FDA by health care professionals and, more recently, by patients themselves through MedWatch, the FDA’s adverse event reporting program. The FDA is also exploring proactive means to identify adverse drug reactions (ADRs) through its Sentinel Initiative which queries electronic health record systems and administrative and insurance claims databases for signals of potential adverse effects.

Self-reported patient information offers a valuable perspective that might not be captured in a doctor’s office, clinical trial, or the most sophisticated surveillance software, and has been found to be of similar quality to that provided by health professionals (Blenkinsopp et al., 2007). The goal of this research is to deploy the infrastructure needed to explore the value of Web and social network postings as a source of “signals” of potential adverse drug reactions soon after the drugs hit the market.