AI AND CAUSAL ML DRIVEN SITE SELECTION AND REAL TIME ANALYTICS FOR CLINICAL TRIAL EXECUTION IN INFLAMMATORY BOWEL DISEASE

Abstract Conducting clinical trials in inflammatory bowel disease (IBD) presents ongoing challenges that limit efficiency and trial outcomes. Key barriers include slow and unpredictable patient recruitment, variability in clinical practice across sites, and limited agility in responding to protocol changes. Advances in AI and machine learning now make it possible to transform patient- and site-level data into actionable, precision-guided insights that address longstanding challenges in development. At PhaseV, we are operationalizing this promise through our ClinOps Platform—a next-generation platform that harnesses the combined power of AI, causal machine learning, and real-time analytics to deliver trials that are adaptive, inclusive, and patient-centered from the outset. PhaseV’s ClinOps platform reframes site selection as a covariate-based optimization opportunity. Patient- and site-level covariates—including age, body mass index (BMI), ethnicity, genomic markers, comorbidities, and treatment history—are modeled using causal ML to estimate each site’s likelihood of achieving recruitment targets and data quality outcomes, aligned with protocol inclusion/exclusion (I/E) criteria and therapeutic mechanism of action. This enables precision-guided selection that enhances population diversity and efficiency. Once trials are underway, an interactive dashboard provides continuous oversight of recruitment, visit adherence, screening rates, protocol deviations, and data timeliness. This innovation is particularly critical in IBD, where heterogeneous disease presentation, fluctuating patient conditions, and evolving standards of care demand adaptive and precisely targeted trial strategies to achieve meaningful outcomes. PhaseV’s offering is further enhanced by a strategic collaboration with the Crohn’s & Colitis Foundation. By integrating insights from the Foundation’s IBD Plexus® Research Accelerator, a globally comprehensive repository of IBD patient data and biosamples, with PhaseV’s causal machine learning solutions, we amplify the power and precision of our analyses, enabling more patient-centric trials and advancing IBD research with adaptability. PhaseV’s approach transforms clinical operations in IBD trials from reactive tracking to intelligent orchestration. Every data point becomes an operational lever, every site an active contributor, and every trial a self-optimizing system—shifting from monitoring to optimization, and from hindsight to foresight. Most importantly, this means focusing on outcomes that matter to patients—timely access to therapies, better quality of life, and inclusive participation that reflects the diverse needs of the IBD community. By applying our clinical operations approach, we create trial environments that recruit faster, adapt smarter, and deliver more successful, patient-relevant results. Figure 1: