Where is RBQM Heading in The Future With AI?

Richard Young, Chief Strategy Officer at CluePoints, shares insights on AI and risk-based quality management (RBQM) in clinical trials. With experience at both CluePoints and Veeva, Young discusses AI’s role, transparency, and the future of risk management in clinical trials.

Moe: Many companies tout AI for efficiency. How does CluePoints ensure AI-driven RBQM enhances trial quality and integrity, not just cut costs?

Moe: With increased scrutiny on AI decision-making, how do you ensure CluePoints’ AI models are transparent and innovative?

Richard Young: AI is a very broad term, covering automation and machine learning. At CluePoints, we use statistical models and augment them with automated large language models (amongst other things). It’s about choosing the right technology for the job and ensuring transparency. Educating stakeholders builds trust in AI, as regulatory guidance can sometimes be based on misunderstanding. For example, the FDA’s guidance against retraining models during a phase three study reflects a lack of understanding of continuous learning models. Trusting the machine and the validation process is crucial, as is understanding that AI is here to stay. We aim to bridge this gap through transparency and education, ensuring our models are not just black boxes but tools stakeholders can trust and understand.

Moe: AI can introduce biases in trial data. What steps is CluePoints taking to mitigate biases and ensure AI-driven insights are equitable?

Richard Young: Managing noise is a major concern, as it can lead to mistrust and misdirection. At CluePoints, we focus on closing the loop between analytics, operations, and outcomes. By identifying signals and outcomes, we can trace actions and ensure biases are minimized. For instance, by analyzing past data and outcomes, we adjust models to avoid false positives and negatives, which are less concerning than noise that can distract and mislead. This approach builds trust and ensures AI-driven insights are reliable and equitable, leading to better patient outcomes and more efficient trials.

Moe: The industry’s perspective on risk is changing. Is this shift improving trial oversight, or is there a danger of over-relying on AI?

Richard Young: Risk is often misunderstood, leading to overcompensation. I believe we need to accept risk as part of informed data management. The challenge is balancing the scientific push for more data with expanding regulatory requirements. These two opposing forces are bridged by operations, the “how” connecting the what we want with what we need  By focusing on how we bring these worlds together, we can improve trial oversight without over-relying on AI. For example, shifting from a risk-based to an informed data management approach helps manage risks effectively. It’s about understanding that risk is inherent in innovation and using AI to manage it intelligently, rather than fearing it.

Moe: AI can streamline trials but risk patient distrust. How are you addressing transparency in AI processes to build trust?

Richard Young: Transparency is key. By integrating AI with human oversight, we ensure the right people are involved in decision-making. This approach builds confidence among patients and regulators. Continuous data lock and rolling submissions require total transparency and audit trails, which AI can facilitate, expediting decisions and ensuring patient safety. For instance, by ensuring AI-driven decisions are traceable and accountable, we can defend our decisions and build trust with patients. This transparency is crucial as we move towards more dynamic and continuous trial processes.

Moe: You moved from Veeva to CluePoints, suggesting a belief in risk-based approaches. What gaps in risk management can CluePoints solve better than Veeva?

Richard Young: Veeva and CluePoints serve different demands. Veeva focuses on operational execution, while CluePoints turns data into a product, enabling better patient management. The opportunity at CluePoints lies in automating data reuse and improving decision-making. This is the next significant barrier in data management, and solving it is my goal at CluePoints. By transforming data into actionable insights, we can manage patients more holistically and improve trial outcomes. This focus on data as a product is what sets CluePoints apart and is why I made the transition. Solving this challenge will be a significant step forward for the industry and patient care.