- Adverse Event reporting has increased substantially in the last few years.
- Clients and CRO’s are both looking to reduce costs and increase efficiencies.
- How AI is helping shape this revolution in the industry.
For drug manufacturers globally, introduction of new drugs in their respective markets has resulted in huge increase of AE (adverse events) reporting in the last few years. And going by the trend and the rate of introduction of new drugs (and for molecules already existing in the market), the reporting of AE’s is set to increase manifold. Regulators now not only depend on the traditional channels of reporting AE’s, they have, in fact, increased the stakes for the MAH’s by including websites, social media, insurance claims, online forums etc. for data collection. In spite of all these checks and balances in place, according to an estimate, more than 85%- 90% of AE’s go unreported every year.
From a regulators view point, they want to spend more time addressing complex issues rather than spending long hours glued to moving around data from various projects they work on. They are also open to looking at solutions that addresses these issues scientifically without missing out AE’s getting reported.
In many organizations, current IT infrastructure and applications are capable of automating activities like case processing and basic level of reporting, though it still requires considerable manual effort. There are multiple tiers of automation which involve machine learning with robotics and natural language processing (NLP) and the final tier involves AI (artificial intelligence), system goes through pattern based self-learning to make the correct assessments and the requirement for human intervention is minimal. Also, the AI system gets ‘intelligent’ over a period of time as more and more data is crunched through it as it’s ‘productivity’ improves…now both quality and quantity of output is far more than what humans can ever match. The AI systems can take the form of multi lingual, multi format case extraction and adverse event classification. They also have the ability to classify ICSR’s whether they are valid, duplicate or invalid on the basis of case characterization. They also deploy bots and mine literature through cognitive search processes and in the process keep producing better and better quality results. Though AI systems are improving rapidly, so far these haven’t ‘evolved’ so much that they can completely replace the medical judgements of highly experienced professionals as far as high end pharmacovigilance activities like medical writing goes. Of course, as these systems keep learning, the day is not too far off in the future when they will be able to match the medical judgment of a highly skilled pharmacovigilance professional.
The development of AI in pharmacovigilance may also lead to possible changes in pharmacovigilance professionals’ job responsibilities. Their future roles in a workplace assisted by AI and potential skillsets required when working with AI are likely to undergo a paradigm shift. AI would allow for pharmacovigilance professionals to shift their work focus from volume-based to a value-based strategy. Machine learning algorithms will assist and enhance the drug safety professionals’ decision-making abilities and support them in generating more efficient and accurate results while performing their job responsibilities.
AI adaption is now a fast growing trend in the industry. Ranging from in-house AI process implementation to outsourcing for AI systems based solutions, pharmaceutical companies or MAH’s are looking at all possible available solutions. The 3rd party pharmacovigilance audit service providers (CRO’s) are also fast gearing up for to implement similar systems and process to meet these requirements from clients who outsource their pharmacovigilance activities. Overall, very exciting times for the Pharmacovigilance industry ahead.