Use of artificial intelligence in monitoring adverse reactions

Abstract

Background. Pharmacovigilance, the cornerstone of public health, is being transformed by technological advances, in which artificial intelligence (AI) plays an important role in minimizing conventional approaches, characterized mainly by human errors and scalability issues in monitoring adverse reactions. Objective(s). Researching the prospects for using artificial intelligence in the field of pharmacovigilance and the techniques used in monitoring and evaluating adverse drug reactions to guarantee patient safety. Materials and methods. Systematic, analytical-descriptive study of EMA, WHO, FAERS directives and reports, laws, orders of the Ministry of Health, AMED, journals listed in electronic databases (Pubmed, EMBASE, SCOPUS). The search employed a range of keywords pertinent to the research topic, including but not limited to "artificial intelligence" and "adverse reactions." Results. Globally, the VigiBase platform, developed by WHO, integrates advanced tools such as VigiRank, VigiMatch and VigiGrade, which use specific algorithms to prioritize and evaluate reports on adverse drug reactions, transmitted by over 130 countries. In parallel, the Sentinel system, developed by the U.S. FDA, leverages real-world data and applies modern machine learning and natural language processing techniques to monitor drug safety. Within the pharmaceutical industry, IBM Watson for Drug Safety is a reference AI solution, contributing to compliance with the requirements for compliance with Good Pharmacovigilance Practices imposed by EMA. Conclusion(s). Realizing the full potential of AI in pharmacovigilance demands close collaboration among regulatory bodies, healthcare professionals, and AI developers to ensure process validation, ethical compliance, and continuous human oversight in monitoring adverse drug reactions.