PS53: Designing Clinical Trials in Neuroscience: When a Statistician, a Machine Learning Researcher, and a Neuroscientist Walk into a Bar

Design and success of clinical trials in neuroscience encounter numerous and exceptionally complex challenges. Such challenges include the heterogeneity of the expression of the diseases, the uncertainty of the outcome assessments especially with softer outcomes or patient reported outcomes, high placebo response, overlapping symptoms of diseases and disorders making diagnoses difficult, variability of the patient population, lack of adherence to medication etc. Experiencing one or more of these challenges can often lead to inconclusive or uninterpretable study results leading to a remarkably high rate of failure of new compounds. Thus, there is a need to develop and implement innovative statistical methodologies and study designs to enhance the efficiency as well as the robustness of CNS studies. Fortunately for us, we are in a new age of technology and digital which when seamlessly leveraged and integrated in clinical trial design, can offer remarkable benefits in enhancing the probability of success and making future trials more patient centric than ever before. In this session, we will hear from perspectives from sponsors, regulators, and clinical investigators. These expert speakers and panelists have spent decades encountering the challenges in designing neuroscience trials, but more importantly innovating leveraging the opportunities to leverage statistical and machine learning methods in addressing and mitigating the risks in such studies to make them more patient centric. We will discuss and highlight how AI/ML enabled methods, tools and approaches are changing the game in neuroscience and pointing to new ways of understanding the disease and the patients, thus helping us design more resilient patient centered clinical trials in mental health.