Article Text
Abstract
Aims To obtain epidemiological data on the prevalence, aetiology, management and outcome of refractory convulsive status epilepticus (RCSE).
Methods Data on children admitted with RCSE to eight paediatric intensive care units (PICUs) were collected retrospectively using a standard proforma designed and co-ordinated by PICAnet. Data were collected between 31/12/2007 and 31/12/2009.
Results Data were collected on 245 (male, 179) patients aged <1 month to 16.5 years (mean 3.8, median 2.8 years). Causes included acute symptomatic (12.4%), remote symptomatic (18.4%), epilepsy-related (22.4%), progressive encephalopathy (5.3%) and febrile (14.7%); no cause was identified in 23.7%.
Thirty nine patients received treatment (midazolam > diazepam) at home and 73 received treatment from paramedics (diazepam > midazolam).
In the Emergency Department (ED), 219 patients (89%) received at least one benzodiazepine dose, 197 (80.4%) received phenytoin and 23 (9.3%) received phenobarbital. Subsequent anticonvulsants in ED included thiopentone (105 patients, 42.9%), midazolam (15, 6.12%) and midazolam and propofol (4, 1.63%).
In ED the APLS guidelines were followed precisely in 90 patients (36.7%); 88 patients (35.9%) received an inappropriate dose of benzodiazepine (above guideline dose in 62, below guideline dose in 26). Thirty seven patients (15.1%) received anticonvulsants in the wrong sequence.
The mean duration of admission to PICU was 3.7 days (range 1–45, median 2). The mean length of days ventilated (on PICU) was 3.2 days ventilated (range 1–31, median 2).
Nine patients died (3.7%). Twenty seven patients (11%) demonstrated a new neurological deficit on discharge from PICU, of whom 10 (4% of the entire cohort) continued to show this deficit at 30 days.
Conclusions Approximately one third of the patients admitted to PICU with RCSE had been treated in the ED appropriately using the APLS algorithm. Thiopentone was the most commonly used first-choice anticonvulsant to treat RCSE on admission to PICU. Mortality was low and approximately 1 in 25 showed a new neurological deficit at the 30-day follow-up.