Pediatric new‐onset seizure‐like events: Yield of diagnostic work‐up

OBJECTIVE: To assess the diagnostic yield of ancillary investigations in children presenting with new-onset seizure-like events (SLEs) at a tertiary first seizure clinic (FSC). METHODS: We conducted a retrospective cohort study including 1213 children evaluated for new-onset SLEs at a tertiary FSC. As part of the FSC consultation, all children underwent a routine EEG recording. Data on demographics, diagnoses, and additional investigations after FSC consultation were collected via chart review. Diagnostic tests (e.g., repeated EEG, magnetic resonance imaging (MRI), or genetic testing) were defined as those performed after FSC consultation with the intention to establish a final epilepsy or no-epilepsy diagnosis. Descriptive statistics summarized the results and yields of ancillary investigations. RESULTS: At FSC consultation, the initial FSC diagnosis was epilepsy in 23.0%, no epilepsy in 43.4%, and undefined in 33.6% (including single unprovoked seizure in 5.9% and unclear in 27.8%). At follow-up, 33.6% were diagnosed with epilepsy, 60.8% with no epilepsy, and in 5.7% the final diagnosis remained unclear. Only 23 children (1.9%) were misclassified at FSC consultation, corresponding to 2.9% of children with a defined diagnosis (epilepsy or no epilepsy). Additional investigations were mainly performed in children with an undefined FSC diagnosis. Of 934 children without an epilepsy FSC diagnosis, repeated EEG contributed to a final epilepsy diagnosis in 7.7%, MRI in 1.2%, and genetic testing in 0.4%. Sleep-deprived EEG was the most performed repeated EEG and showed interictal epileptiform discharges in up to 71.2% of children with an unclear FSC diagnosis and final epilepsy diagnosis. SIGNIFICANCE: This study provides valuable insights into the diagnostic yield of ancillary investigations performed after initial evaluation for pediatric new-onset SLEs. In children with an initially undefined diagnosis, repeated EEG recordings, particularly sleep-deprived EEG, contribute to a final epilepsy diagnosis more frequently than MRI or genetic testing.