Emergency room access for recurring seizures: when and why

BACKGROUND AND PURPOSE: To develop a hypothetical model identifying potentially modifiable predictive factors of Emergency Room (ER) visits by patients suffering from drug resistant epilepsy. METHODS: During a 1-year period, all adult drug resistant patients followed by the same epileptologist were recruited after the occurrence of one or more epileptic attacks. They were divided into two groups based on whether they went to the ER after seizures. A prospective comparative analysis of the clinical and social characteristics of the two groups was performed in order to identify independent predictors of ER visits. Logistic regression analysis was used to confirm the potential predictive role of the evaluated variables. RESULTS: Logistic regression analysis confirmed the potential role in predicting ER visits for these variables: foreign nationality, current psychiatric therapy, current antiepileptic drug polytherapy, comorbidities, more than one episode in the same day and changes in usual seizure pattern. A relevant association was also found between the frequency of ER neuroimaging use and the following variables: occurrence of episodes on holidays or weekends, current antiepileptic drug monotherapy, multiple comorbidities and brain injury after seizure. CONCLUSIONS: The present study evaluated factors, some potentially amenable to change, related to drug resistant epileptic patients' ER visits following a seizure. This information may serve to improve the clinical and therapeutic management of patients, decrease the need for urgent care and reduce subsequent patient stress and related costs.