Doctors develop web tool to predict seizure in kids

The model uses five key pieces of information from first consultation

Clinicians from the Netherlands have developed and validated a model to help determine the risk of epilepsy in children who have one or more paroxysmal events.

Epileptic seizures in children may be underrecognised because of the heterogeneous clinical symptoms and limited sensitivity of the initial interictal EEG, note Dr Eric van Diessen of University Medical Center in Utrecht and colleagues in Pediatrics.

The prediction model (see link below) combines information available at the first consultation, including clinical characteristics (sex, age of first seizure, event description, medical history) and the initial EEG findings.

Researchers tested the tool on retrospective data from 451 children who visited their outpatient department for diagnostic workup related to one or more paroxysmal events.

They had a least one year of follow-up data for all children who were either diagnosed with