Abstracts

Rationale: Prevalence of acute confusional state (ACS) is about 25 % in patients over 65 years old admitted to the emergency room (ER). Epilepsy and ACS share many risk factors and are both associated with confusion and fluctuating consciousness. Nevertheless, in clinical practice, a link between epilepsy and ACS is seldom sought in the elderly. The aim of this work was 1) to study when a link between epileptic activity (EA) and ACS was searched in an ER of an academic hospital and 2) when it was done, the prevalence of EA in elderly patients with ACS. Methods: Charts of patients over 65 years old admitted for ACS in the ER between and were retrospectively reviewed. Risk factors for ACS and epilepsy were studied, including medical history, medications, biological and iconographic work-up, EEG results if done. Those characteristics were compared in patients who had EEG and those who hadn't. Results of EEG were classified in normal, diffuse slowing, focal slowing and the presence of EA. Clinical characteristics of patients with EA on EEG were compared to the patients without EA on EEG. Results: Hundred and seventy seven charts were analysed. Mean age was 78 years old. There was no difference between patients who had EEG and those who hadn't. Fifty six percent (98/177) of patients had an EEG. Eighty four percent of EEGs were abnormal. Seventy four percent of EEG showed diffuse slowing, 44 % focal slowing. EA were present in 15 patients : 12 with interictal epileptiform discharges (IED) and 3 with non convulsive status epilepticus (NCSE). There was no statistical differences between patients with EA on EEG and patients without EA on EEG. Conclusions: No clinical, biological or iconographic anomalies can distinguish a priori patients in ACS who will have EA on EEG and therefore EEGs should be performed in all elderly patients with ACS. EAs probably have a role in ACS since their prevalence is ten times higher than expected in an hospitalized population (ref). This high prevalence of EA in ACS has not been reported before and should be confirmed by prospective studies, ideally with continuous EEG monitoring since 30 minutes routine EEG may underestimates EA.