ICMRA Assesses AI Risks for Conducting Pharmacovigilance

Artificial intelligence-based software developed to conduct pharmacovigilance “appears suitable for the detection of safety signals,” the International Coalition of Medicines Regulatory Authorities (ICMRA) says in a new report on the use of AI for drug development.

But the ICMRA found that there are potentials risks involved in using an AI-based system. For example, the AI tool might be “difficult or impossible” for regulators to inspect or the drug sponsor might neglect to assess the effectiveness of any software updates.

To avoid such problems, the drug sponsor “would require specialist expertise in AI, data quality and pharmacovigilance signal detection to govern its use,” the ICMRA advises.

But it also lists potential benefits from using AI for pharmacovigilance. The report notes that “current signal detection and management tools have a heavy manual component that may be hard to sustain in the future,” adding that there is “potential for such AI use to discover safety signals that are more difficult to detect with current methods, such as drug-drug interactions, drug-disease interactions, medication errors, secondary malignancies, changes in frequency and severity of known events, patterns of use in medications, and misuse.”