The Evolving Role of Artificial Intelligence in Pharmacovigilance
Artificial intelligence is increasingly being used in the science and activities relating to the detection, assessment, understanding, and prevention of adverse effects or any other medicine-related problem, known as pharmacovigilance.
Artificial intelligence in pharmacovigilance offers a way to keep pace with rising pharmaceutical innovation and fast-tracked medications that place surveillance experts under increasing pressure.
Capable of automatically translating huge swathes of patient data and formulating predictive adverse event recommendations, AI in pharmacovigilance enables a more agile, accurate, and proactive way to monitor, manage, and hinder adverse drug events for patients around the world.
Read on to discover more about:
- What pharmacovigilance is
- How AI activates agility in pharmacovigilance
- How AI in pharmacovigilance is streamlining the pharma industry
- Applying AI to your career in pharmacovigilance
What is pharmacovigilance?
Pharmacovigilance refers to the detection, assessment, monitoring, and prevention of adverse drug events (or as they’re commonly known, side effects), brought on by the consumption of drugs.
Initially founded after the thalidomide crisis in 1961, pharmacovigilance has evolved over time to include traditional medicine, medical products, blood products, and herbal medicine.
In conjunction with these new areas, the pharmaceutical industry’s consistent demand for pharmacovigilance is driving a huge amount of patient and drug data that is increasingly challenging to process.
How AI activates agility in pharmacovigilance
Artificial intelligence offers the pharmacovigilance sector a way to maximise the predictive potential of data and expand monitoring surveillance, all while minimising the burden of manual analysis to create a faster, more agile way of responding to adverse events.
Over the course of COVID-19, the criticality of an agile, and rapid approach became urgently apparent. As new medications were fast-tracked through the FDA’s new regulatory pathways, pharmacovigilance needed to find a way to evolve fast to keep pace.
Before the pandemic, pharmacovigilance relied on a multi-faceted adverse drug event process, where information on any adverse events was initially transferred to market authorisation holders and manufacturers, before being forwarded on to the Medical and Healthcare Products Regulatory Agency (MHRA).
In response to the pandemic and a growing need for agility, many regulatory bodies upgraded their approach by integrating AI.
In 2020, the UK’s MHRA paid AI software experts, GenPact, £1.5m to develop an AI tool capable of sifting through the influx of adverse event reports following the pandemic. Meanwhile, the MHRA adapted their Yellow Card programme by embedding elements of AI to enable patients to directly report to the MHRA thereby streamlining what was once a lengthy process.
Meanwhile, pre-pandemic, in 2019, the US’s Food and Drug Administration (FDA) released a 5-year plan for integrating AI into the existing pharmacovigilance framework, Sentinel.
According to official release papers, the FDA’s investments will focus “on innovations emerging from new data science disciplines, such as natural language processing and machine learning, and seek to expand its access to and use of electronic health records”. By applying AI techniques to health record data sets, Sentinel is set to evolve from being reactive, to a proactive real-time process. This means moving away from testing hypotheses against a single adverse outcome, to testing drugs against the full universe of potential health outcomes simultaneously.
How AI in pharmacovigilance is streamlining the pharma industry
Despite increased investments in AI technology over the pandemic, pharmaceutical organisations, and regulatory bodies have had a sustained interest in integrating AI into pharmacovigilance, even pre-pandemic.
In 2017, a global survey revealed that 62% of the pharmacovigilance professionals preferred using AI for adverse event reporting.
Today, pharmaceutical organisations continue to find new ways to incorporate AI across pharmacovigilance to futureproof the safety of every patient.
In an interview with Pharmaceutical Technology, Annette Williams, Vice-President and Global Head of Lifecycle Safety at IQVIA explains how public drug safety communication at the company is being enhanced with AI. Williams describes how IQVIA’s medical information call centres are adapting to respond to increasing demands for public information.
By integrating “AI-enabled agents who can handle many of the routine questions patients have”, IQVIA are able to sustain a “greater service in the after-hours scenario”, with the AI agents answering “15-25% of the live calls that humans used to have to answer”. Meanwhile, IQVIA have also integrated robotic technology and machine learning for adverse event report processing, enabling the automatic standardisation and translation of data.
Applying AI to your career in Pharmacovigilance
As digitalisation across the pharmaceutical industry continues to accelerate, AI skills are fast becoming embedded within the roles of drug safety and pharmacovigilance experts today.
While formal job descriptions seldom specify formal AI training, AI skills are often taught as part of on-the-job training, or through self-motivated learning, and are valuable indicators of a more experienced professional.
Following research investigating the implementation of AI in pharmacovigilance at Celgene’s Global Drug Safety and Risk Management (GDSRM) division, Danysz et al describe the following core competencies as indicative of professional preparedness.
Core Pharmacovigilance and AI Competencies:
Fundamental AI Skills:
- The ability to understand artificial intelligence, natural language processing, machine learning, and deep learning
- The ability to interact with, and flag issues surrounding artificial intelligence, natural language processing, machine learning, and deep learning outputs within a user interface
- An understanding of how manipulating machine learning outputs relates to deep learning
Advanced AI Skills:
- The ability to both describe and train concepts of artificial intelligence, natural language processing, machine learning and deep learning.
- The ability to alter artificial intelligence, natural language processing, machine learning and deep learning outputs within a user interface via algorithm retraining.
- The ability to resolve issues through overwriting or accepting machine learning outputs within a user interface.
Looking for your next role in pharmacovigilance or AI?
At SRG, we connect scientists across the globe with rewarding careers across STEM. Our talent consultants are on hand to help you make your next step.
Latest News, Events & Insights
Why is Diverse Leadership in STEM Important?
Jacob Midwinter, Director of Sales and Search by SRG, discusses the key success factors behind driving diversity in leadership, as well as the challenges and opportunities leaders in STEM can expect to encounter along the way, with senior leaders Julia Buckler - QIAGEN, Dr. Amy Smith - CPI, Rich McLean - GPAS, Dr. Garry Pairaudeau - Exscientia, and Professor Charlotte Deane - Exscientia.