Dr. Janet Woodcock, the head of the Center for Drug Evaluation and Research (CDER) at the U.S. Food & Drug Administration (FDA), says the agency’s Sentinel drug safety surveillance system is ready for implementation. Congress passed the FDA Amendments Act in 2007. This act mandated that the FDA establish an active surveillance system to monitor drugs using electronic data from health care information holders. The FDA proceeded to develop the Sentinel program, which was intended to be a central database to monitor all FDA-regulated products, including drugs and medical devices. Sentinel was launched as a pilot program in 2008.
It appears that as of 2014, Sentinel has moved from a proposed or pilot program and is now an “integral part of routine safety surveillance,” according to Dr. Woodcock. The Sentinel program works in cooperation with existing drug and device monitoring programs such as the FDA Adverse Event Reporting System (FAERS), Manufacturer and User Facility Device Experience Database (MAUDE), and Vaccine Adverse Event Reporting System (VAERS). However, these are passive reporting systems, relying on manufacturers, patients, physicians and other health care providers to report problems with drugs and devices.
Sentinel, on the other hand, uses a rapid query system that is able to gather information from automated health care data streams such as electronic health record (EHR) systems from hospitals, insurers and universities; administrative and insurance claim databases, and registries. These sources bring to light up-to-the-minute medical product safety issues.
There are three component parts in the Sentinel program – the Active Postmarket Risk Identification and Analysis (ARIA) system, Postmarket Rapid Immunization Safety Monitoring (PRISM) system, and the Blood Surveillance Continuous Active network (BloodSCAN). The Sentinel database includes data for nearly 200 million individuals.
The program recently added two new partners, and is now able to gather data from the Medicare Virtual Research Center, which will provide more information about drug and medical device experiences for people older than 65; and the Hospital Corporation of America, which will greatly expand the amount of information about patients’ experiences during hospital stays.
It should be noted that Sentinel is not a centralized database. It is a distributed data network. Various components or organizations that are part of Sentinel retain their own data, which can then be accessed through Sentinel, or “mined,” but the data is held and owned by participating organizations.
Although Sentinel allows the FDA to explore a vast amount of information, it is not clear that the program has yet made much difference in improving drug safety. Researchers say that pooling the data about adverse events is a positive step, but that a meta-analysis of that data doesn’t necessarily expose the truth about a drug’s risk. However, they say that as the database grows and more information is collected, it could yield better analysis of a drug’s risk profile.
If data does indicate a possible risk, another problem is how the FDA should address that risk. Consumers are calling for an “early warning” system that could alert them as well as physicians to potential problems. But this opens the door to creating unnecessary alarm. It’s a fine line, and likely to be one of the most important challenges for Sentinel and the FDA moving forward. It will be interesting to see what this new technological frontier brings to the future of drug and device regulation.
Sources: FDA, HealthAffairs.org, and Regulatory Affairs Professionals Society