Cross-national pharmacovigilance of drug-induced intestinal obstruction: disproportionality signals in FAERS and external validation in JADER

Using deduplicated FAERS reports from Q1 2004 to Q4 2024 (18,532,100 unique reports), we identified 62,919 cases of drug-induced intestinal obstruction (DIO) and prioritised the 50 suspect drugs with the highest DIO report counts. Disproportionality screening using both the reporting odds ratio (ROR) and Bayesian information component (IC) showed that signal magnitudes varied widely across the Top 50 set (ROR 0.62–5.08; IC −0.69–2.33), with 28/50 drugs meeting prespecified signal criteria by both metrics. Bevacizumab demonstrated the strongest disproportionality (ROR 5.08; IC 2.33), while adalimumab contributed the largest number of DIO reports. Time-to-onset analyses conducted at the report–drug pair level indicated that DIO frequently occurred after weeks to months of exposure (median 100 days; IQR 19–469), with substantial drug-level heterogeneity and a predominance of delayed-onset events (>30 days) for most agents. Cross-database benchmarking in JADER for 20 overlapping drugs showed moderate concordance, supporting partial transportability of leading signals while underscoring setting-specific heterogeneity. Comparison with current product information using a three-level framework (Yes/No/Unclear) suggested incomplete alignment between disproportionality signals and explicit obstruction terminology (12/50 Yes, 28/50 No, 10/50 Unclear). These findings are hypothesis-generating and support sustained clinical vigilance for high-signal therapies, particularly in patients with elevated baseline bowel risk, and confirmation in longitudinal pharmacoepidemiological studies with validated exposure windows and rigorous confounder control.