Background: Identification of outcome-predictive factors could lower risk of under- or over-treatment in status epilepticus (SE). Older age and acute symptomatic aetiology have been shown to predict mortality, but other variables are controversial and level of consciousness has received relatively little attention. The objective of this study was to assess variables predictive of mortality, particularly those available at presentation.
Methods: The discharge database (1997–2004) of two university hospitals was screened for adult patients with EEG confirmed SE, excluding cerebral anoxia. Outcome at discharge (mortality, return to baseline clinical conditions) was analysed in relation to demographics, clinical features, and aetiology. Aetiologies were also classified based on whether or not they were potentially fatal independently of SE.
Results: Mortality was 15.6% among 96 patients with a first SE episode, 10 of whom also experienced recurrent SE during the study period. Eleven other patients had only recurrent SE. Mortality was 4.8% among these 21 patients with recurrent SE. Return to baseline condition was more frequent after recurrent than incident SE (p = 0.02). For the first SE episode, death was associated with potentially fatal aetiology (p = 0.01), age ⩾65 (p = 0.02), and stupor or coma at presentation (p = 0.04), but not with gender, history of epilepsy, SE type, or time to treatment ⩾1 h.
Conclusions: At initial evaluation, older age and marked impairment of consciousness are predictive of death. Surviving a first SE episode could lower the mortality and morbidity of subsequent episodes, suggesting that underlying aetiology, rather than SE per se, is the major determinant of outcome.
- CP SE, complex partial SE
- GC SE, generalised convulsive SE
- NC SE, non-convulsive SE
- NCSEC, non-convulsive SE with coma
- PFE, potentially fatal aetiology
- SE, status epilepticus
- SP, simple partial
- consciousness impairment
- status epilepticus
Statistics from Altmetric.com
Dr Rossetti is supported by the Swiss National Science Foundation and the SICPA Foundation, Prilly, Switzerland. Statistical analyses were supported by the Biostatistics Consulting Service, Center for Clinical Investigation, Brigham and Women’s Hospital.
Competing interests: none declared
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.