Objectives To assess the clinical usefulness of HVT over the first 2 years.

Design Cohort observational.

Subjects 60 patients (49 F) with epilepsies or non-epileptic paroxysmal clinical events.

Methods 48–72 hour continuous video EEG at patients’ own environment.

Results HVT answered the primary clinical question in 45/60 patients (75%), and provided additional clinical information in 5 patients [2 with unsuspected coexistent psychogenic non-epileptic seizures (PNES) and 3 with unsuspected sleep disorders (SD)]. Of the 12 patients with Idiopathic Generalized Epilepsy, absences had been overestimated in 6 and underestimated in 4, while absence status was recorded in 1 of the 2 patients in whom it had been suspected. Valproate was possible to drastically reduce or stop in 3/6 women. Focal seizures were recorded in 19 of 28 patients with focal epilepsies, PNES were the habitual seizures in further 2 patients, while syndrome classification changed in one. In all 4 patients referred for differentiation between SD and epilepsy, HVT confirmed parasomnias in 2, daytime naps in 1 and idiopathic hypersomnia in 1. The diagnosis of PNES was confirmed in 8 of 13 suspected patients. HVT was unhelpful in the 3 patients referred for not witnessed, poorly understood episodes of loss of consciousness. Three patients switched off the video and 2 failed to change battery on day 2.

Conclusions HVT is a useful diagnostic test provided that diagnostic hypothesis and clinical question are appropriate.