Predictors of unfavourable seizure outcome in patients with epilepsy in Nepal

Despite optimal medical therapy, a sizeable number of patients continue to have persistent seizures. We evaluated the association of pretreatment and treatment variables with unfavorable seizure outcome.

Patients with follow-up over 12 years in the Nepal Epilepsy Association were evaluated. Patients having seizures for at least a year and already on polytherapy after failure of two monotherapy trials were considered having unfavourable outcome. Variables under study were: age, sex, duration and frequency of seizures prior to treatment, type of seizure, neurological status, Computed Tomography (CT) finding, and failure of first anti-epileptic drug (AED). Bivariate analysis was done with Chi-square and Fisher exact tests. Potential interaction between variables was studied with a logistic regression analysis.

Out of a total 529 consecutive patients, 490 were included in the study. Unfavorable seizure outcome was seen in 26.8% of patients. Among 284 patients who remained viable for analysis, bivariate analysis showed significant association of unfavorable outcome with frequency of seizure (p 0.01), abnormal neurological status (p 0.01) and failure of first AED (p 0.00), while no significant association was seen with age at onset (p 0.45), sex (p 0.47), duration of seizure (p 0.43), type of seizure (p 0.12), and presence of CT abnormality (p 0.46). The fitted regression model portended an unfavorable prognosis with failure of first AED and abnormal neurological status, however, failed to show significant association with frequency of seizure.

Failure of first AED trial and associated neurological deficits are significant predictors of unfavorable seizure outcome.