Long-term prognosis after childhood convulsive status epilepticus: a prospective cohort study

Abstract

Background: The prognosis of convulsive status epilepticus (CSE), a common childhood medical neurological emergency, is not well characterised. We aimed to investigate the long-term outcomes in a cohort of participants who previously had CSE.

Methods: In this prospective study, we followed up a population-based childhood CSE cohort from north London, UK (the north London convulsive status epilepticus surveillance study cohort; NLSTEPSS). We collected data from structured clinical neurological assessment, neurocognitive assessment (Wechsler Abbreviated Scale of Intelligence), brain MRI, medical records, and structured interviews with participants and their parents to determine neurological outcomes, with adverse outcome defined as presence of one or more of epilepsy (active or in remission), motor disability, intellectual disability, or statement of special educational needs. We applied multiple imputation to address missing data and performed binary logistic regression analyses on complete-case and imputed datasets to investigate sociodemographic and CSE factors associated with adverse outcomes.

Findings: Of 203 survivors (90% of inception cohort), 134 (66%) were assessed at a median follow-up of 8·9 years (IQR 8·2-9·5). The cumulative incidence of epilepsy was 24·7% (95% CI 16·2-35·6), with most (89%) emerging within 18 months after CSE. The cumulative incidence of epilepsy was lower in patients with prolonged febrile seizures (14·3%, 6·3-29·4) and survivors of acute symptomatic CSE (13·3%, 3·7-37·9) than in those of remote symptomatic CSE (45·5%, 21·3-72·0) and unclassified CSE (50·0%, 25·4-74·6). One participant (2·9%, 0·5-14·5) in the prolonged febrile seizures group developed temporal lobe epilepsy with mesial temporal sclerosis. The absence of fever at CSE was the only predictor of incident epilepsy (odds ratio [OR] 7·5, 95% CI 2·25-25·1). Motor and intellectual disability was seen predominantly in participants who had idiopathic and cryptogenic CSE (seven [36·8%, 95% CI 19·1-59·0] and 16 [84·2%, 62·4-94·5] of 19, respectively) and remote symptomatic CSE (33 [62·3%, 48·8-74·1] and 40 [75·5%, 62·4-85·1] of 53), and most of these participants had pre-existing disabilities. Pre-existing epilepsy was the only predictor of intellectual disability (OR 8·0, 95% CI 1·1-59·6). 51·5% (95% CI 43·1-59·8) of those followed up had a statement of special educational needs.

Interpretation: Childhood CSE is associated with substantial long-term neurological morbidity, but primarily in those who have epilepsy, neurological abnormalities, or both before the episode of CSE. Survivors without neurological abnormalities before CSE have favourable outcomes.

Funding: BUPA Foundation, The Academy of Medical Sciences, Wellcome Trust, National Institute for Health Research, and Young Epilepsy.