Official statistics report around 1000 deaths due to epilepsy in the UK each year (Hanna et al 2002). Most of these deaths are un-witnessed and in many cases are believed to have been avoidable with timely assistance (Langan et al 2000). A major problem is detecting nocturnal seizures to allow body re-positioning, to maintain an open airway and to administer rescue medication. There are several seizure alarms commercially available but are often unreliable with many false alarms. The aim of this study is to investigate a novel seizure detection system with a unique algorithm.

Two duplicate novel medical devices conforming to safety regulation BS EN 60601 containing the unique algorithm will be evaluated in clinical trials at two hospitals over a 14 month period. Each device is time synchronised with an Electroencephalographic (EEG) videotelemetry system which is the gold standard in identifying different seizure types recording from scalp electrodes. The devices will be monitoring participants from 3 months and older (children and adults). The algorithm was created from results from a PhD where 527 seizures were categorised into different seizure types. Sensitivity of identifying seizures was 91% and specificity was 75% using an optimal trigger level and algorithm. This trial will investigate how reliable the algorithm is in practice and determine an optimal trigger level..

Inclusion Criteria:

• Frequent epileptic seizures 3 months and older Informed consent for participation

Exclusion Criteria:

• Infrequent seizures Attacks are of a non-epileptic nature Younger than 3 months Has an implantable device No informed consent.

Patients who are being admitted for electroencephalographic videotelemetry as part of their clinical management of their epilepsy will be invited to participate in the study.