Animation and Networks: Visualizing Relationships Among Safety Signals in Clinical Trials

• Activity Number: 54
• Type: Invited
• Date/Time: Sunday, July 31, 2011 : 4:00 PM to 5:50 PM
• Sponsor: Section on Statistical Graphics
• Author(s): William Robert Prucka*+
• Companies: Eli Lilly and Company
• Keywords: Visualization ; Network Plots ; Animation

Abstract

Efficient identification and characterization of safety signals from a clinical trial is critical to 1) ensure patient safety and 2) to optimize the benefit/risk of therapies in development. Evaluation of safety signals may be complex, requiring investigation of the interrelationships among many variables. For each patient, a time varying vector of efficacy and/or safety variables is measured throughout the course of the study. It is necessary to identify both expected and unexpected patterns in these data. Tabular summaries have an important role in identifying safety signals, but visualization offers efficient real-time exploration. This presentation focuses on visualization to understand the relationships among adverse events (AE), vital signs, and on discontinuation (DC). Network plots are used to demonstrate the connectivity (co-occurrence) of AE. Specialized plots are developed to relate the AE to DC. Animation based procedures are used explore longitudinal trajectories of vitals data and the relation to AE states. All visualization methods will be presented using realistically simulated data. The benefits and limitations of the proposed visual techniques are discussed.