Identification of children with first afebrile seizure for whom neuroimaging is unnecessary

Abstract

Purpose: To find ways to identify children with first afebrile seizure for whom neuroimaging is unnecessary.

Methods: We retrospectively reviewed the clinical records of children younger than 19 years of age with the first afebrile seizure who visited the emergency department of the National Center for Child Health and Development in Japan between May 2014 and December 2020. We investigated the relationship between age, sex, focal seizure, seizure duration, seizure cluster, neurological findings, and CT and/or MRI abnormalities by univariate analysis. Furthermore, to identify children with low probability of intracranial abnormality, we performed decision tree analysis by classification and regression tree methods.

Results: Among the eligible 611 children, 14 children had a CT abnormality (4.2% of patients who underwent CT) and 26 had a CT or MRI abnormality (7.1% of patients who underwent CT or MRI). Six children had an urgent neuroimaging abnormality. In the univariate analysis, seizure cluster (P = 0.02) was significantly associated with CT abnormality, and focal seizure (P = 0.01) and seizure prolonged for more than 5 min (P = 0.04) were significantly associated with CT or MRI abnormality. The decision tree analysis identified seizure cluster, prolonged seizure, neurological disorder, and focal seizure as risk factors for CT abnormalities in that order.

Conclusion: Children without seizure cluster, seizure prolonged for more than 5 min, and neurological disorder may not require CT in the emergency room. The clinician could determine the necessity of neuroimaging by seizure cluster, prolonged seizure, focal seizure, and neurological disorder.

Keywords: Decision trees; Pediatric emergency medicine; Pediatrics; Seizures; Unprovoked seizure.