Bridging Real-World Data and Regulatory Decision-Making: The Role of AI in External Control Arm Development


Discussion Leader: Paul Arora, PhD, Division of Epidemiology. Dalla Lana School of Public Health, University of Toronto, Toronto, ON, Canada
Discussants: Xiaoyan Wang, PhD, Intelligent Medical Objects, Westport, CT, USA; Mei Yang, Ph.D., Happy Life Technology, Short Hills, NJ, USA


The workshop aims to critically examine recent guidelines on external control arms (ECA), with a specific focus on integrating AI to enhance the utilization and efficacy of Real-World Data (RWD) in ECA development. The discussion intends to offer insightful perspectives and practical applications to advance the methodologies in clinical research settings.


NICE has pioneered among European HTA organizations by introducing guidelines for implementing ECAs within its 2022 RWE framework. In early 2023, The US FDA released guidance for the utilization of ECA, emphasizing the use of patient-level data. Concurrently, China's CDE unveiled a draft outlining the design and framework for real-world drug research, with extensive discussions in ECA development. In parallel, the rapidly advancing generative AI, epitomized by Large language models (LMMs) and ChatGPT, has initiated a paradigm shift in the healthcare landscape. In this workshop, Dr. Arora will provide an overview of the guidelines, address common challenges of utilizing RWD for ECA in the US, and review a case study highlighting potential benefits of AI in ECAs (15’). Dr. Yang will explore the guidelines and challenges associated with utilizing RWE for ECA in China, including a case study in rare diseases (15’). Dr. Wang will delve into the advancements in generative AI and LMMs, discussing her recent research on ECAs created from case reports using generative AI in the fields of rare disease and oncology (15’). All panelists will analyze the areas of alignment and divergence based on the audience poll, steering the discussion of opportunities, challenges and future trajectories of integrating AI and RWD for regulatory submissions (15’)

The workshop is designed to be particularly relevant for professionals in RWE research, clinical development, and regulatory sectors. It aims to elucidate the significant potential contribution of AI in enhancing the generation and application of RWE.




Study Approaches