Sentinel: Harnessing the Power of Databases to Evaluate Medical Products

Posted on March 18, 2014 by FDA Voice

By: Michael D. Nguyen, MD

Clinical trials are designed to evaluate the safety and effectiveness of medications. But the number of participants in clinical trials represents only a fraction of the number of people who will ultimately use the product after FDA approves it. Occasionally, rare and unexpected health problems (which we call “adverse events”) become apparent only after a medical product is on the market and many more people use it. That’s why it’s so important for FDA to continue to monitor the safety of medical products, a practice called surveillance.

Most safety surveillance systems are passive: They rely on health care professionals and consumers to notice and report adverse events. Although these passive systems remain essential, they have certain limitations. For example, health care professionals and consumers might not recognize that the product is the cause of an adverse effect and not report it to FDA. Or, they might report a suspected adverse event that’s not truly the cause of a problem the consumer experiences.

Now imagine if we could actively search more than 100 million health insurance records to uncover possible adverse events, rather than relying on doctors and patients to report them. Such a system would enable us to get continuous feedback on the use of medicines under real-world conditions – feedback that might help us to discover unexpected patient reactions or unexpected drug interactions.

FDA scientists have partnered with the Harvard Pilgrim Healthcare Institute to create such a surveillance system, called Sentinel. Within Sentinel, FDA has supported the development of software that analyzes information from health insurance and health record databases to search for evidence that certain products are linked to specific adverse effects. Although these data are protected behind tight firewalls and remain under the control of the original health insurance plans that created them, the software makes it possible to analyze the information without disclosing identifying information in order to strictly maintain patient privacy.

FDA and Harvard Pilgrim Healthcare Institute are using this surveillance system to determine whether a certain type of immune therapy is associated with heart attacks or strokes, and to better define the true rate of acute lung injury after transfusions of certain blood components.

More recently, FDA completed its first study using the Sentinel system, which evaluated the safety of the two current vaccines (RotaTeq and Rotarix) that prevent rotavirus infection (the leading cause of severe diarrhea and dehydration in infants). The new study revealed that these rotavirus vaccines slightly raise the risk of a rare bowel problem (intussusception) that previously caused a prior rotavirus vaccine (Rotashield ) to be voluntarily withdrawn from the market by its manufacturer. But the Sentinel study showed that the newer vaccines have a much lower rate of this bowel problem and are safer, with the benefits outweighing the risks, including the risk of intussusception, associated with vaccination.

The Sentinel rotavirus vaccine and immune therapy studies are examples of how FDA scientists are harnessing the power of big electronic databases to ensure the safety of the medical products we use every day.

Michael D. Nguyen, MD, is the Acting Director of the Division of Epidemiology in FDA’s Center for Biologics Evaluation and Research