Drug Surveillance/Epidemiology Office Updates Activities
CDER’s Office of Surveillance and Epidemiology (OSE) now has access to a broader selection of validated data sources for its safety reviewers to use. “For example, a few years ago, we couldn’t analyze disparate impacts for patients who were pregnant, or for patients by race in observational studies,” OSE says in its just-posted 2021 annual report. “Today, we often can. A few years ago, we could only access safety information that was several months old. Today, we can access a broader selection of validated data sources much more quickly, which means our staff must interpret ever-larger sets of complex safety information, to see the big picture, and take necessary action to protect patients.”
In 2021, FDA released a guidance entitled “Real World Evidence (RWE) Claims Guidance for Industry,” which was mandated by the 21st Century Cures Act. The document outlines how such data can help support or satisfy post-approval study requirements for drug products, in addition to helping support approvals of new indications for already approved drugs or biologics. The agency also issued a draft guidance that provides recommendations for evaluating the relevance and reliability of electronic health records and medical claims data that could be used in a clinical study and to eventually support a regulatory decision on effectiveness or safety, according to the report.
OSE says it also launched three new drug safety teams last year, each of which oversees the safety of a portfolio of marketed drugs. The expert multi-disciplinary teams monitor and prioritize the range of safety issues that come up in neurology, oncology, and infectious disease product areas, the report says. “Drug safety teams facilitate sharing of information about safety issues from all CDER Offices with scientific safety responsibilities and increase efficient safety evaluations,” it says.
OSE is also using artificial intelligence (AI) as a tool to efficiently extract and organize information to allow drug surveillance reviewers to “focus more on complex tasks associated with public health impact,” the report continues. “AI is being applied to data contained in individual case safety reports within FAERS [FDA Adverse Event Reporting System] to derive a visualization of the temporal relationship between the drug and adverse event, to support the identification of duplicate reports, and assist in the triaging of reports with high information quality.”
Additionally, the report said surveillance staff have begun using a computerized labeling assessment tool (CLAT) to ensure container labeling conforms with applicable statutes, regulations, standards, and FDA guidance. CLAT uses modern artificial intelligence methods, including machine learning, natural language processing, and image processing to “automate manual labeling review efforts,” the report says. “Automating the labeling reviews will create operational efficiencies and help streamline and standardize the review process to ensure consistency across different products and review teams.”