Risk factors and scoring system as a prognostic tool for epilepsy after neonatal seizures

Neonatal seizures may cause irreversible changes to the immature brain and. A scoring system for early prognostic information could be a useful clinical tool. The aim of the study was to analyze risk factors for epilepsy after neonatal seizures, to validate Garfinkle's scoring system, and to analyze whether a new scoring system is feasible.

A retrospective study of 176 newborns (59.1% boys, 40.9% girls, 70.5% term, 29.5% preterm; mean birth weight 2820 g), admitted to the Department of Neonatology, Division of Pediatrics, University Medical Centre, Ljubljana, because of neonatal seizures (clinical and/or neurophysiological), was performed. Epilepsy rate between 2 and 12 years of follow-up was 18.1%. Five independent predictors from Garfinkle's study and other known predictors were entered into hierarchical binary logistic regression models and analyzed through four steps to identify independent predictors of epilepsy. We tested whether any of the predictors was an effect modifier.

Of five potential predictors from Garfinkle's score, electroencephalograph background findings and etiology were predictive. Etiologies, gestation, mode of delivery, duration of seizures, and other risk factors at birth were found to be independent predictors. Duration of seizures has a different effect on prognosis depending on the gestational age.

Gestational age determines the association between duration of seizures and epilepsy. Scoring systems to predict development of epilepsy after neonatal seizures need to limit interaction between important predictor variables.