Risk of febrile seizure associated with β-lactam antibiotic use: a case-crossover study with an active comparator

β-Lactam antibiotics are generally considered safe, but neurotoxicity has been reported. These reports mainly involve adults; however, evidence in children remains limited. Although febrile seizure (FS) is the most common type of seizure in early childhood, multiple factors contribute to FS onset, and a potential association with β-lactam exposure has not been fully examined. We therefore investigated the association between oral β-lactam antibiotics and FS using a case-crossover design with an active comparator.

Children younger than 6 years who experienced an incident FS were identified from a Japanese health insurance claims database. Exposure to oral β-lactam antibiotics was assessed on the hazard date (day -1) and the control date (day -22), while oral macrolides served as the active comparator. Conditional logistic regression was used to estimate odds ratios (ORs), and comparator-adjusted ORs were derived by dividing β-lactam ORs by macrolide ORs. The attributable fraction among the exposed (AFe) was calculated from the comparator-adjusted OR.

We identified 43,936 children. The comparator-adjusted OR for FS with oral β-lactam exposure was 1.21 (95% confidence interval [CI], 1.03-1.40). These results were generally consistent across sensitivity analyses in which the control date was shifted from day -22 to day -15 or day -29.

Oral β-lactam antibiotics were associated with a modestly increased risk of FS. This finding should be interpreted cautiously because residual time-varying confounding by infection severity and fever may remain. Clinically indicated β-lactam antibiotics should not be withheld on the basis of these findings alone.