Enhancing Patient Safety:

A Paradigm Shift Through Early AI Adoption in Pharmacovigilance Workflows

Traditional PV operations typically start at Case Processing, but the persistent challenge of underreported adverse drug reactions (ADRs) demands a paradigm shift. US FDA’s estimate that only 1%–10% of ADRs reach the FDA Adverse Event Reporting System underscores a need for a more proactive strategy. This paper stems from the FDA Discussion paper on Using Artificial Intelligence and Machine Learning in the Development of Drug and Biological Products, which highlighted the applications for PV beginning at the Case Processing stage. However, the emphasis on identifying adverse events remains a pivotal precursor to effective processing. Vast unstructured data, often beyond a company’s control, poses a considerable hurdle in this endeavor.