More Information
Speaker and Supports:
• Frank Bretz (Novartis)
• Laura Flight (NICE)
• Rieke van der Graaf (University of Utrecht)
• Thomas Jaki (MRC Biostatistics Unit)
• Jonathan Kimmelman (McGill University)
• Alex John London (Carnegie Mellon University)
• Richard Milne (University of Cambridge)
• Philip Pallmann (University of Cardiff)
• Mark Sheehan (Oxford Population Health)
• Jerome Singh (University of KwaZulu-Natal)
• Sue-Jane Wang (Food and Drug Administration)
• Sofia S. Villar (MRC Biostatistics Unit)
• Haiyan Zheng (University of Bath
There will also be speakers from the US, UK and EU regulatory authorities.
Why is this workshop important at this time?
Clinical trials play a crucial role in advancing medical knowledge and managing uncertainty by systematically comparing the effects of various interventions. In a confirmatory setting, they provide a structured approach that enables reliable conclusions about the relative safety and efficacy of treatments. However, these trials are complex, require significant time and resources and raise ethical concerns, particularly due to practices such as randomizing participants to different interventions and implementing additional procedures like blood draws or biopsies. If the evidence produced is flawed, it can have serious consequences, potentially leading to the discontinuation of effective therapies or the adoption of interventions that are ineffective or even harmful. Therefore, while the urgency to gain insights and accelerate learning is important, it must be carefully balanced with the need for reliability, generalizability, and the ethical treatment of participants.
Recently, innovative statistical methods have been proposed to allow clinical trials to adapt in various ways. Some forms of adaptation are already widely accepted and commonly used, such as stopping a trial early when treatments are statistically effective or ineffective. However, other forms of adaptation are more controversial. This includes complex computational methods for dynamically adjusting the probability that individuals are randomized to different interventions, aiming to maximize the proportion of participants receiving the most effective treatment, if one exists.
The controversy surrounding these methods raises both ethical and methodological questions. There is ongoing debate about whether, when and which of such methods are more efficient, meaning they can answer research questions with fewer participants without compromising reliability or generalizability. This debate is fueled by competing simulation studies that present conflicting assessments of these designs, as well as concerns that certain approaches, particularly Bayesian designs, may introduce unacceptable bias. Adaptive trials played a crucial role in addressing the COVID-19 pandemic, leading to increased public interest in this methodology and highlighting the need to resolve outstanding uncertainties about their statistical and ethical implications.
Event web page:
Ethics and Innovative Clinical Trial Designs: A Multi-Stakeholder Perspective on Adaptive Designs - K&L Gates Initiative in Ethics & Computational Technologies - Carnegie Mellon University
Special thanks to the International Centre for Mathematical Sciences for kindly supporting this event.
