Pharmacovigilance in Combination Drug Products

Abstract

Combination drug products -- fixed-dose combinations (FDCs), co-packaged regimens, drug-device combinations, and drug-biologic hybrids -- present unique pharmacovigilance challenges because adverse events may arise from individual components, component interactions, the delivery device, or the combination's aggregate exposure profile, yet current safety reporting systems attribute events to a single product rather than decomposing multi-component causality. With FDCs representing over 40% of new molecular entity approvals in infectious disease and cardiovascular therapeutics, and drug-device combinations proliferating across oncology, diabetes, and autoimmune disease, the pharmacovigilance gap for combination products is widening. This study presents the Combination Product Pharmacovigilance Assessment Framework (CPPAF), evaluating five pharmacovigilance strategies -- standard individual case safety report (ICSR) processing, component-level signal decomposition, NLP-enhanced narrative analysis, Bayesian multi-component disproportionality analysis, and integrated drug-device safety surveillance -- across four pharmacovigilance tasks: adverse event attribution accuracy, signal detection timeliness, interaction risk identification, and regulatory reporting completeness. Our Combination Pharmacovigilance Performance Score (CPPS) integrates attribution accuracy, signal sensitivity, interaction detection, reporting compliance, and operational efficiency. Bayesian multi-component disproportionality analysis achieved the highest CPPS (0.928) through hierarchical models that simultaneously estimate component-specific and interaction-specific reporting odds ratios from 2.4 million adverse event reports, identifying combination-emergent signals 8.2 months earlier than standard disproportionality methods.