FDA Officials, Key Stakeholders Reflect on Sentinel’s Evolution
Ten years after the US Food and Drug Administration (FDA) launched its Sentinel Initiative, agency officials and other key stakeholders reflect on the continuing evolution of the system in an article appearing in the New England Journal of Medicine on Thursday.First launched as a pilot program in 2008, the Sentinel Initiative encompasses FDA’s effort to meet obligations set by Congress in the 2007 Food and Drug Administration Amendments Act to develop a system for active postmarket risk identification and analysis for medical products.
Since then, the agency has committed to further expanding Sentinel’s capabilities as part of the Prescription Drug User Fee Amendments (PDUFA VI) negotiations.
Earlier this year, FDA Commissioner Scott Gottlieb said that FDA has conducted more than 250 analyses using Sentinel since it became fully operational in 2016, which Harvard Professor and Sentinel Principle Investigator Richard Platt pointed out would have been too expensive for the agency to perform without Sentinel.
In the NEJM article, Platt and others write that “Sentinel data have informed many regulatory decisions made by the Center for Drug Evaluation and Research and, in the past 2 years, have eliminated the need for postmarketing studies on nine potential safety issues associated with five products (e.g., ustekinumab and serious infections).”
The authors also write that FDA has been able to use Sentinel to better understand drug usage patterns, including for opioids, and to investigate whether drugs are being used according to their label or during pregnancy.
Looking forward, the authors predict that in five to 10 years Sentinel will be much more powerful than it is today as new data sources are added and new analytic methods are developed.
“Examples of such approaches include distributed regression methods that preserve privacy and enable analysis of individual patients’ data when they are distributed throughout multiple organizations, as well as adoption of machine learning, natural language processing, and other technologies that enable improved use of electronic health records or other sources of real-world data,” the authors write, noting that the agency is building separate programs to meet the data needs of its other product centers.
The authors also see an opportunity to expand Sentinel’s use for other public health needs, such as chronic and infectious disease surveillance, or to track quality-improvement efforts at healthcare delivery systems.
NEJM