Abstract
Aim: To develop an integrated framework that strengthens reproducibility and interpretability in FAERS analysis through scalable drug standardization, event harmonization, and exposure normalization. Methods: FAERS reports from 2015–2024 were deduplicated. Drug ingredients were normalized to RxNorm concepts using the OHDSI Standardized Vocabularies and mapped to Anatomical Therapeutic Chemical (ATC) classifications. Adverse event terms were harmonized across MedDRA versions. U.S. exposure estimates from the Medical Expenditure Panel Survey (MEPS) were mapped to RxNorm concepts and aligned with event year to derive annual reporting rates and assess temporal trends. The framework was demonstrated using two case studies. Results: The analytic dataset included 13.8 million cases. Exact RxNorm mapping was achieved for 95.8% of 30.5 million suspect or interacting drug records. A total of 769 legacy PTs (3.6% of event terms; affecting 4.4% of cases) were mapped to MedDRA v28.0. MEPS exposure estimates were linked to 513 RxNorm concepts representing 46.1% of U.S. FAERS drug records, with 80.8% of drug-year estimates meeting AHRQ precision standards. For bleeding, the cumulative rivaroxaban–apixaban reporting odds ratio remained elevated, while reporting rates reversed after 2020. For myocardial infarction, SGLT2 inhibitors showed higher disproportionality and reporting rates than GLP-1 receptor agonists, but drug-level analysis revealed marked within-class heterogeneity. Conclusion: Drug standardization and event harmonization reduced key sources of inconsistency in FAERS, while exposure-normalized reporting rates placed case counts in the context of drug utilization. Linked through RxNorm concepts and aligned by event year, these components work together to improve the reproducibility and interpretability of FAERS analysis.