Abstract

he limitations of existing disproportionality analysis methods for Food and Drug Administration Adverse Event Reporting System (FAERS) data highlight the need for automated tools for efficient data mining and analysis. This study aimed to develop and validate an automated Python-based tool for FAERS data processing. The methodology included: (i) Automation development and signal detection using Python and (ii) Validation through traditional disproportionality analysis. Public FAERS quarterly extract files from 2004 (Q1) to 2021 (Q4) were accessed, matched, and deduplicated using a multi-step approach. The cleaned dataset was analysed using a contingency table to compute reporting odds ratios, PRR, Chi-square statistics, and 95% confidence intervals. Validation confirmed that results from the automated tool matched traditional analysis. Using Remdesivir as an example, we identified 12,777 adverse events and 256 safety signals. The tool offers multiple advantages: simplified coding, minimal storage requirements, cloud-based execution (Google Colab), accessibility for non-technical users, rapid processing, tailored signal detection, and automated FAERS updates. The extreme deduplication process ensures refined results, aligning with pre-defined criteria. This validated tool can significantly enhance pharmacovigilance research and regulatory decision-making, with potential for broader adoption through user-friendly desktop interfaces.

Key words: Adverse event; Disproportionality analysis; FAERS; Risk assessment