Back to: Statistical Considerations for Premarketing Risk Assessment
🔔 Note: Stop playback at 26:05 to complete this lesson.
Presenter
Gregory Levin, PhD
Associate Director for Statistical Science and Policy
Office of Biostatistics (OB)
Office of Translational Science (OTS)
Center for Drug Evaluation and Research (CDER)
US Food and Drug Administration (FDA)
Abstract
This presentation addresses the Statistical Considerations for Premarketing Risk Assessment, outlining objectives to enhance the assessment of safety data through trial design, understand the importance of tailoring analysis to align with trial design, and identify approaches to assess causal relationships between a drug and adverse outcomes. The safety evaluation utilizes a framework distinguishing three critical aspects: the assessment of Adverse Events of Special Interest (AESIs), the descriptive assessment of general safety for signal detection, and triggered analysis for unexpected signals.
Improved planning is critical prior to initiating Phase III trials to ensure reliable evaluation of a new drug’s safety. Quality planning considers the sufficiency of the controlled period duration and the size of the proposed safety database. It is vital to determine whether routine open-ended AE reporting is sufficient or if directed data collection (e.g., supplemental case report forms, specialized instruments, or blinded adjudication) is needed for specific AESIs to improve precision and characterization. A key design element involves preventing missing data by ensuring plans are in place to follow all randomized participants through the end of the controlled period, even following treatment discontinuation. The presentation discusses the difference between on-study analyses (which preserve the integrity of randomization and are generally recommended) and on-treatment analyses (which may be subject to bias). Finally, sponsors should prospectively plan and document the principal features of safety analyses, submitting integrated summaries of safety (ISS) statistical analysis plans or program-wide documents (like PSAPs) to the FDA prior to late phase trial initiation.