Evaluation of the accuracy of seizure descriptions by the relatives of patients with epilepsy

Abstract

The descriptions of seizures by witnesses are important in the diagnosis and classification of epileptic seizures. The aim of the study was to evaluate the accuracy of this information obtained from relatives of patients with epileptic and non-epileptic attacks. Thirty patients with epileptic or non-epileptic attacks had seizures videorecorded whilst inpatients at the Assessment and Treatment Centre of the National Society for Epilepsy and the National Hospital for Neurology and Neurosurgery, Chalfont, Buckinghamshire, UK. A relative or close friend of each patient viewed the recording and subsequently completed a structured questionnaire, testing recall of 15 separate elements of the episode. This account was compared to a definitive evaluation completed by medical staff. An accuracy of describing seizures was therefore obtained. Our results showed that there was a wide variation in the accuracy of recall with convulsive episodes described less well than non-convulsive (median accuracies 44.5 and 70%, respectively, P<0.05). Attacks in which the diagnosis had been changed, following admission, from the one obtained on history alone were particularly inaccurately described (median accuracy 26%). Elements of the questionnaire that were most inaccurately recalled were description of limb movement and post-ictal behaviour. Those most accurately described were facial appearance and vocalization. Features only present in non-epileptic attacks included reactivity to eyelash stimulation, opisthotonic posturing, flailing or protective limb movements. In conclusion, our results confirm that inaccuracy exists when witnesses describe attacks and that this may lead to errors in diagnosis and subsequent treatment.

Introduction

The diagnosis of epilepsy relies primarily upon a detailed eyewitness account of a seizure. The description of seizures by witnesses is often incomplete or inaccurate. Descriptions can occasionally be misleading to the extent that psychogenic episodes may be incorrectly diagnosed as epileptic, leading to treatment with anti-epileptic medication for several years that is ineffective, costly and has adverse effects (Aird and Tsukaki, 1958, Meierkord et al., 1991). Further, seizures may be incorrectly classified and therefore treated inappropriately. Videorecording and, in particular, simultaneous video/EEG recording has been shown to increase the accuracy of diagnosis, importantly identifying those patients with non-epileptic attack disorder (Luther et al., 1982, Meierkord et al., 1991, Samuel and Duncan, 1994).

The in-patient Assessment and Treatment Centre of the National Society for Epilepsy and the National Hospital for Neurology and Neurosurgery at Chalfont receives tertiary referrals of patients with epilepsy from all over the UK. The seizures of all patients in the assessment centre are videotaped and these recordings are reviewed by medical staff to aid seizure classification. In addition, structured questionnaires are used routinely to quickly assess and classify patients’ attacks. For the purposes of this study, the seizures of 30 patients were videotaped and their relatives or close friends completed a questionnaire after watching the recorded seizure. We report the accuracy of these completed questionnaires in describing the videotaped event.