What can nonclinical findings tell us about what may happen in humans? Connecting nonclinical and clinical data can provide a more complete view of safety and help researchers understand how findings observed before clinical development may translate to patients. Drawing on TransCelerate’s Enabling Translational Safety initiative, this session will discuss practical considerations for data sharing, data quality, terminology alignment, and analytical approaches that help evaluate relationships between nonclinical observations and clinical outcomes. Speakers will also explore the opportunities and limitations of emerging AI-enabled and data-driven approaches to generate clinically meaningful insights and improve confidence in safety interpretation across the development continuum.

The objective of this session is to learn how proactive benchmarking of safety event rates using a cohort of patients in indicated populations can benefit pipeline and in-line products, enabling prompt decision-making, supporting pharmacovigilance, and responding to regulatory queries.

We developed a two-stage workflow to support safety and pharmacovigilance literature surveillance by combining automated retrieval with AI-based relevance ranking and triage. The approach is designed to improve consistency, reduce manual review burden, and preserve recall by surfacing high-value articles for expert review. We will share lessons learned on model benchmarking, evaluation, and operationalizing the workflow in a regulated review environment.

As clinical trial complexity continues to grow, safety teams must move beyond fragmented reviews and disconnected data sources. This session highlights AbbVie's journey to create a unified safety intelligence ecosystem spanning the core pillars of medical monitoring, safety review, and patient profiling. Attendees will learn how integrated analytics, standardized data foundations, and purpose-built workflows enable accelerated risk identification and a holistic understanding of patient safety.

Outsourcing is commonplace but is still a complex decision. Some PV activities are more amenable to outsourcing, while others should remain in-house under the sponsor's control. The final vendor selection will always be a very company-specific decision, but there are guardrails and processes to follow that will ease the path and ensure a successful long-term partnership.

Gene therapies present distinct safety monitoring challenges across clinical development and post-marketing settings. Their complex mechanisms, individualized treatment pathways, and pre-treatment procedures can complicate causality assessment, risk characterization, and long-term follow-up. This presentation examines where traditional pharmacovigilance models may fall short and discusses practical approaches to strengthen signal detection, benefit-risk assessment, and patient protection for advanced therapeutics.

Clinical trial safety management is often fragmented across reporting systems, SOPs, signal evaluation, and communication pathways, creating variability and delays in safety oversight. This session will present a practical framework for building a unified safety ecosystem that integrates standardized SOPs, internal and external reporting, signal evaluation and communication, and frontline nurse engagement. The approach clarifies roles, escalation pathways, and documentation while leveraging nurses as essential partners in early safety signal identification and escalation.

Ultra-rare disease trials are challenged by small, dispersed patient populations and fragmented real-world data (RWD) across registries, electronic health records, clinical notes, claims, and digital sources. Using a case study, this session will outline an AI-enabled framework for integrating and harmonizing these data to develop external control arms. It will address endpoint alignment, causal inference, confounding adjustment, missing data, bias assessment, and the creation of transparent, traceable, and inspection-ready evidence for regulatory decision-making.

Electronic health record (EHR) to electronic data capture (EDC) interoperability has been proven through pilots, established standards, and advancing technology, yet adoption remains fragmented across the clinical research ecosystem. If the technology exists, what barriers remain for EHR-to-EDC to scale globally? This session explores a shared vision for a future in which research-relevant EHR data can be exchanged and reused across sponsors, sites, studies, and technology platforms without bespoke implementations or proprietary dependencies.

Drawing on TransCelerate's EHR Connectivity initiative, presenters will share insights from ecosystem assessments and stakeholder engagement, highlighting the progress achieved and the technical, operational, governance, and validation barriers that remain. Attendees will explore what it will take to move from isolated successes to broad adoption and will be challenged to join a holistic industry effort to build the roadmap, partnerships, and reusable assets needed to make scalable EHR-to-EDC interoperability a reality.

Centralized monitoring generates a significant volume of actions, but it can be difficult to know whether those actions truly reduced risk or simply closed a finding. This session shares a practical approach to measuring action effectiveness by connecting risks, signals, actions, and outcomes. I will discuss challenges, our work evaluating effectiveness, and how these insights can support more meaningful risk-based quality management.

The promise of AI-assisted statistical analysis in clinical development is significant, yet deploying AI into a highly regulated production environment remains difficult. We describe a pragmatic, two-stage strategy—recognizing that the path to trustworthy AI runs through standardization, not around it. An iterative interim platform built on validated, reusable analytical modules delivers faster, reproducible analyses today while serving as a context-engineering engine: modularizing the statistical workflow into steps that must remain deterministic and those that allow flexibility, enabling early process and SOP development, and refining modules through user feedback. Throughout, the effort empowers, not replaces, statisticians.

This presentation will explore how ATLAS uses AI-enabled capabilities to help streamline trial operations, connect information across workflows, identify insights, and support proactive decision-making. The discussion will also highlight practical considerations for responsible AI adoption, including data quality, governance, integration, and human oversight.