AI technologies, such as machine learning (ML) and natural language processing (NLP), continue to advance at a rapid rate. They can offer significant support to quality, regulatory and safety professionals with their daily activities, as well as enhancing commercial operations, through accelerating innovation, data insights, and improved process efficiency and effectiveness. However, constraints on how these AI technologies can be applied to the healthcare industry still exist: global regulatory requirements must be met; data availability, volume, and congruence can provide limitations; and the resultant product solution needs to be commercially viable for broad industry uptake.
To successfully navigate the deployment of AI technologies in a QMS, a company needs to consider the strategic utilization of technology in targeted use cases, guided by human expertise and supported by well-designed processes. For example, intelligent, data-driven insights provided by a digital, AI-powered QMS can accelerate time-to-action and identify potential risks throughout the product lifecycle. With AI providing targeted insights, this “electronic eye” can support the quality, regulatory and safety teams to become true, high-performance, augmented professionals.
Increasing quality management complexity
The growing complexity of quality management, driven by the evolution of global regulations and standards and the addition of regulation around new technologies (for example, the EU AI Act and other global publications on AI that are in draft/under consultation) drives challenges in how a company implements its quality management system. Within this environment, quality, regulatory and safety professionals are navigating the design, manufacture, distribution, and lifecycle activities of technologically advancing product solutions under continued economic constraints.
Identifying targeted use cases for the deployment of AI in a QMS, such as the ability to identify potential adverse events and product quality issues across a range of structured and unstructured data sources, could offer significant value and alleviate some of the resource burden. In this use case, industry professionals would spend less time involved in the manual identification and entry of cases and more time focused on the review of reported events and strategic decision-making of how such cases could drive further product improvements. With AI supporting potential case identification, the timeliness, quality and volume of case intake could increase, providing data of a higher quality, congruence and volume on which to gain targeted insights to drive improvements in product performance. The utilization of an “electronic eye” also reduces the risk of human fatigue and drives enhanced process consistency. As AI algorithms “learn” from human verification, the operations of such a system increase, driving improved decision-making and increased efficiency alongside enhanced monitoring of product performance.
However, technology alone is not a silver bullet. Augmenting a “human-in-the-loop” professional with AI technologies allows a company to leverage their deep domain expertise and critical thinking abilities to interpret the outputs of the AI tools and to contextualize the information.
