BACKGROUND: The incidence of recurrent febrile seizures during the same febrile illness (RFS) is 14-24%. A pilot study found that body temperature and male sex were predictors of RFS. This study sought to validate body temperature as a predictor of RFS, calculate the optimal cut-off body temperature for predicting RFS, and identify the other predictors of RFS.
METHODS: This prospective cohort study enrolled children with febrile seizures aged 6-60 months who visited the emergency department at Atsugi City Hospital, Japan, between March 1, 2019, and February 29, 2020. Children who had multiple seizures, diazepam administration before the emergency department visit, seizures lasting >15 min, underlying diseases, or who could not be followed up were excluded. The optimal cut-off body temperature was determined using a receiver-operating characteristic curve.
RESULTS: A total of 109 children were enrolled, of whom 13 (11.9%) had RFS. A lower body temperature was significantly associated with RFS (P = 0.02). The optimal cut-off body temperature for predicting RFS was 39.2 °C. Children with RFS also had significantly lower C-reactive protein and blood glucose levels (P = 0.01 and 0.047, respectively), but none of the other factors considered were significantly associated with RFS.
CONCLUSIONS: This large prospective study confirmed that body temperature is a predictor of RFS. The optimal cut-off body temperature for predicting RFS was 39.2 °C. Low C-reactive protein level and blood glucose level might be predictors of RFS, but this needs to be confirmed in prospective multicenter studies.