Factors predictive of remission of epilepsy were evaluated in 504 patients followed for an average of 7 years at the Departments of Pediatrics and Mathematics, Dalhousie University and Children’s Hospital, Halifax, Nova Scotia, Canada. At the end of follow-up, 55% of the total cohort were off medication and in remission. Approximately 70% of those stopping medication remained seizure free. At the time of diagnosis, the best predictors of remission, determined by a multivariate analysis scoring system, were 1) age <12 years at onset, 2) normal intelligence, 3) no prior neonatal seizures, and 4) fewer than 21 seizures before treatment. Prediction was improved after 12 months follow-up, when a score for the seizure frequency between 6 and 12 months was included. Patients with absence and minor motor seizures were excluded from the study. Seizure type and the EEG were not of predictive value. 
COMMENT. During the study period, 1977-1985, the number of seizure-free years required before mediction withdrawal was attempted changed from 4 to 2 years. The authors endorse the stopping of medication after 2 seizure-free years, regardless of the predicted prognosis.
A social outcome of epilepsy study in 337 intellectually normal children included in this Nova Scotia population found school failure in 34%, mental health consultation in 22%, psychotropic medication used in 5%, unemployment 20%, and criminal conviction in 2%. Learning disorders and more than 21 seizures before treatment for epilepsy were predictive of an unfavorable social outcome, but seizure control and EEG were non-predictive.