Artificial Intelligence: How it is supporting the pharmaceutical industry and saving lives

It is estimated that there are currently 3.78 billion active social media users (more than half the world’s population) with projections showing this will increase to 4.41 billion in the next 4 years. Many industries have turned to social media for its far reaching ability and data richness and the pharmaceutical industry is no different. The power of social media for the use of pharmacovigilance is well documented but presents many challenges. It is estimated that around 6.5% of hospital admissions in the UK are a result of adverse drug reactions (ADR). This represents a significant cost on an already stressed healthcare system and is something the pharmaceutical industry takes seriously. A study conducted by a US ‘fact tank’ found that nearly three quarters of internet users search for medical information online. This presents a significant opportunity for pharmacovigilance with the process of text mining being employed to find instances of ADR through social media posts. However, the process of text mining for the purpose of pharmacovigilance presents many issues as there can be many variations on how a particular drug is referred to by the general public. For example, some refer to it by its brand name while other refer to it by its active ingredient. Other challenges posed by the FDA are the use of generic terms such as pain killer and antibiotic which makes text mining extremely laborious and time consuming. The role of machine learning in solving this problem has already begun but has so far only been done in the context of pharmacovigilance in a supervised way. This supervised machine learning is more effective than traditional text mining process but still requires a lot of human intervention in the development and update of algorithms. Further development in artificial intelligence (AI) and deep machine learning could reduce or even replace the high level of human input in the design system leading to greater efficiencies in identifying ADRs which will help the pharmaceutical industry modify and develop products which could save valuable resource in the health care systems, prevent costly product recalls, and even more importantly, save patient lives.

AI will be impactful on the pharmaceutical industry far beyond its assistance with pharmacovigilance. A report published in January 2021 suggests that AI will be the most disruptive technology in the pharmaceutical industry and will impact everything from the supply chain, to new drug discovery and clinical trials. Traditional pharmaceutical manufacturers such as AstraZeneca, Novartis and Merck have turned to AI to test new compounds, reassess difficult to treat diseases and to help patients find the correct treatments. However, there are companies that specialise in AI that are moving into the pharmaceutical sector with the realisation that barriers to entry such as the billion dollar cost of developing a new drug, may not apply. Those who have the capabilities to use AI for drug development can reduce the time and risk associated with traditional R&D practices. For you me and millions of others, that may mean a wider range of affordable medications to keep us healthier for longer.