Agent-Specific Clinical Profiles of Antibiotic-Associated Encephalopathy: A Dual-Database Pharmacovigilance Study Using JADER and FAERS
TL;DR
AAE is reported not as a single uniform event but with profiles specific to the causative antibiotic, which may support earlier recognition, although confirmation in analytic studies with exposure denominators is warranted.
Abstract
Background/Objectives: Antibiotic-associated encephalopathy (AAE) is a serious adverse reaction whose presentation is thought to differ between antibiotics; whether these agent-specific profiles are consistently reflected in spontaneous reports across independent data sources is unknown. Methods: Using the Japanese Adverse Drug Event Report (JADER) database and the US FDA Adverse Event Reporting System (FAERS, 2004–2026), we studied metronidazole, cefepime, ceftriaxone, ceftazidime, and levofloxacin. Encephalopathy was defined by the MedDRA terms Encephalopathy or Toxic encephalopathy, and reporting odds ratios (RORs) were calculated in each database; co-reported neurological manifestations, time to onset, patient age, and renal-associated terms were then characterized. Results: All five antibiotics showed significant reporting disproportionality in both databases, with metronidazole and cefepime showing the largest RORs in each (JADER: metronidazole ROR 186.70, cefepime 55.33; FAERS: cefepime 69.29, metronidazole 29.11), consistent with established adverse reactions. In FAERS, metronidazole showed a cerebellar profile (ataxia OR 12.48, dysarthria OR 11.11) and cefepime a renal-associated altered-mental-status/myoclonus profile (depressed level of consciousness OR 12.88, myoclonus OR 11.76; renal-associated terms OR 2.88; median age 71 years); event-specific JADER dates indicated a delayed course for metronidazole (median 28 days). The symptom-level contrast was FAERS-based and was not reproduced when the same analysis was repeated in JADER, a discrepancy that may reflect differences in coding practice. Conclusions: AAE is reported not as a single uniform event but with profiles specific to the causative antibiotic. Awareness of these agent-specific patterns may support earlier recognition, although confirmation in analytic studies with exposure denominators is warranted.