Artificial-Intelligence based Machine-Learning in Pharmacovigilance

Abstract

Introduction:

Adverse drug responses (ADRs) pose a serious threat to healthcare, increasing the risk of death, morbidity, and medical expenses. The growing complexity and volume of healthcare data of ADR is driving the field of pharmacovigilance to evolve and integrating artificial intelligence (AI) techniques as Machine Learning (ML) has emerged as a potential answer. This article collectively addresses the evolving landscape of AI implementation on drug safety monitoring, emphasizing advancements, challenges, and opportunities

Objective:

The objective is to comprehensively examine the utilization of AI and ML techniques in pharmacovigilance, spanning topics such as distributed data networks, drug–drug interactions, ADR, real-time contextual intel, content formation

Method:

The review incorporates articles that were obtained from the databases of PubMed, Embase, Web of Science, and IEEE Xplore between the years 2000 and 2023 using keywords “artificial intelligence”; “machine learning”; “pharmacovigilance”

Result:

There is a significant shift towards advanced ML techniques, particularly deep learning, in pharmacovigilance. AI can predict and assess drug–drug interactions, emphasizing their intricate nature. It also structures data for pharmacovigilance from well-coordinated multi-databases but issues have been identified in distributed data networks. Although pharmacovigilance tasks and data sources now in use may not have been specifically created for causal inference, there is great potential for integrating machine learning with causal paradigms

Conclusion:

The collective findings underscore the promising advancements, persistent challenges, and future potential of AI and ML in enhancing pharmacovigilance practices. Standardization, interdisciplinary collaboration, and ongoing research efforts are crucial for realizing the full benefits of these technologies in ensuring drug safety and mitigating adverse events