Generalised convulsive status epilepticus in Singapore: Clinical outcomes and potential prognostic markers

Abstract

Purpose To study the characteristics, outcomes and prognostic markers of convulsive status epilepticus (SE) in Singapore. Methods 62 adult admissions to the National University Hospital Singapore from 2002 to 2005 were studied. Ethnicity, history of epilepsy, educational subnormality, neuroimaging, seizure duration, length of stay, Modified Rankin Scale (MRS) pre and post discharge, blood glucose, creatine kinase, potassium, white cell and platelet count were recorded. An MRS ≥ 3 at discharge was defined as a poor outcome. ROCs of significant variables were plotted to identify the best test cut-offs. Results Mean age was 59.2 years (range 20–94). 75.9% patients had epilepsy. Mean length of stay was 14 days (range 1–75). Univariate analyses revealed age ( p = 0.01, OR 1.075, 95% CI 1.030–1.122), length of stay in ICU ( p = 0.03, OR 1.299, 95% CI 1.014–1.665) and hospital ( p = 0.014, OR 1.203, 95% CI 1.038–1.393) and hyperglycemia ( p = 0.045, OR 1.327, 95% CI 1.007–1.750) associated with poor outcome. Test cut-off values for prognostic markers were established: age ≥ 55 years (ROC 0.790, sensitivity 72.3, specificity 85.7, PPV9 4.4%, NPV 48.8%) and serum glucose ≥ 7 mmol/L (ROC 0.737, sensitivity 72.3, specificity 80.0, PPV 93.5%, NPV 36.4%). A discriminant model using these variables was then constructed with probability scores for poor outcome. Discussion Age, hyperglycemia and length of stay in hospital influenced outcome from convulsive SE in the local population with hyperglycemia being a novel prognostic marker. Some prognostic markers cited in the literature differed, highlighting the possibility that these indicators may vary across population groups.