Detecting Safety Signals in Pharmacovigilance With Dataiku

Post-market drug safety surveillance is critical for discovering and addressing unforeseen adverse drug events in diverse real-world populations, but the quality and volume/velocity/variety of data from spontaneously reported systems create barriers to operational analytic pipelines. In this session, we’ll discuss some of these challenges and show you how to use Dataiku to prepare data from the FDA Adverse Event Reporting System, calculate disproportionality metrics, and apply modern ML algorithms for enhanced detection and insights into safety signals in pharmacovigilance. We will also discuss how other data sources such as drug labels, population metrics, medical literature, and social media could be incorporated for more powerful applications of AI for global drug safety.