BAHAMA

BAHAMA implements Bayesian hierarchical analysis of MedDRA®-coded adverse events from randomized controlled trials to estimate rate ratios (RRs) and improve detection of rare safety signals.

Key Features:

• Hierarchical Bayesian Framework: Uses a multi-stage Bayesian hierarchical Poisson model to estimate rate ratios (RRs) for MedDRA®-coded adverse events.
• Handling Rare Adverse Events: Implements a data aggregation strategy at higher MedDRA® levels based on predefined incidence thresholds and the MedDRA® hierarchical structure to capture infrequent events.
• Simulation Studies: Validated through extensive simulation studies assessing model performance across scenarios and the benefits of using the full MedDRA® structure and aggregated data.
• Real-World Application: Applied to real-world datasets with practical comparisons to existing methods for adverse event analysis.
• R Package Implementation: The modeling approach and pre-processing steps are implemented in an R package.

Scientific Applications:

• Safety analysis in randomized controlled trials: Estimates adverse event incidence and compares treatment arms using MedDRA®-coded data.
• Detection of rare adverse events: Aggregates hierarchical MedDRA® information to improve identification of infrequent safety signals.
• Regulatory safety assessment: Provides analyses suitable for informing regulatory submissions regarding adverse events.
• Post-market surveillance: Supports pharmacovigilance and post-market safety monitoring using MedDRA®-coded AE data.

Methodology:

Multi-stage Bayesian hierarchical Poisson model to estimate RRs; data aggregation at higher MedDRA® levels using predefined incidence thresholds and the MedDRA® hierarchy; validation via simulation studies; application to real-world datasets; implementation and pre-processing provided in an R package.