Disproportionality Analysis of Drug–Drug Interactions: A Pharmacovigilance Approach

Abstract

Drug–drug interactions (DDIs) represent a significant concern in clinical practice, particularly in the context of polypharmacy and complex therapeutic regimens. This study aimed to identify and evaluate potential DDI signals using a pharmacovigilance-based disproportionality approach applied to the FDA Adverse Event Reporting System (FAERS) database. Data from the FAERS 2025 Q4 dataset were systematically processed, including deduplication, drug standardization, and identification of co-reported suspect drug pairs. Disproportionality analysis was performed using reporting odds ratio (ROR), proportional reporting ratio (PRR), and associated confidence intervals to detect statistically significant signals. A total of 66,947 reports met the inclusion criteria, yielding 226,169 unique drug pairs for analysis. Several high-frequency and high-signal drug combinations were identified, many of which were associated with clinically relevant adverse events consistent with known pharmacological mechanisms. The application of multiple statistical measures demonstrated strong concordance, supporting the robustness of the findings. Sensitivity and subgroup analyses further confirmed the stability of detected signals across different analytical conditions. The results highlight the utility of FAERS for large-scale DDI signal detection and underscore the importance of integrating pharmacovigilance data with pharmacokinetic and pharmacodynamic insights. While limitations such as reporting bias and lack of causal inference remain, this approach provides valuable real-world evidence for hypothesis generation and risk assessment. These findings support enhanced drug safety monitoring and inform clinical decision-making.