A meta-analysis of 2303 patients with a diagnosis of absence epilepsy (AE), derived from 26 publications on 23 study cohorts, was conducted at Leiden University Hospital, The Netherlands. Age at onset of AE, stated for 60%, was as follows: 73% before puberty, 18% between 12 and 17 years, and 8% in adulthood, not conforming to strict AE criteria. Despite application of the 1989 diagnostic classification criteria of the International League against Epilepsy, the outcome definitions differed substantially due to heterogeneity in inclusion criteria and length of follow-up. Remission rates varied from 0.21 to 0.89, the poorest outcomes occurring in patients who developed generalized tonic-clonic seizures (GTCS) and in studies with longer follow-up periods. In the 50 percent with AE and GTCS, the proportion seizure free at follow-up was only 0.35, whereas in the 50 percent with absence seizures alone, 0.78 were seizure free. The prognosis for AE suggested by this meta-analysis was worse than previously stated, and the results would not permit an early prediction of outcome in individual patients presenting with absence seizures. 
COMMENT. The main purpose of this study was to determine if the outcome of absence epilepsy could be predicted with certainty at the time of diagnosis in the individual patient. The authors conclude that early prognostication is not feasible because of the extensive heterogeneity of calculated remission rates. Generally, the outcome should be more pessimistic than that currently accepted. If the AE is “pure” and uncomplicated by tonic clonic seizures (GTCS), the outcome may be good and remission rates favorable. In long-term follow-up, however, a 50% chance of developing GTCS is accompanied by a poorer outcome and lower remission rates.
In a previous study of childhood epilepsies which showed an overall remission rate of 0.71 after AED withdrawal, predictors of relapse were adolescent age at onset, symptomatic epilepsies, and an abnormal interictal EEG . An identical remission rate was reported by Camfield and colleagues in a study in which patients with absence and minor motor seizures were excluded and seizure type was not of predictive value.(see Progress in Pediatric Neurology II, PNB Publ, 1994, for further discussion of outcome studies in childhood epilepsies). Greater attention to EEG characteristics, especially form of spike-wave complexes and duration of paroxysms, at the time of diagnosis and later at AED withdrawal could be more revealing in future outcome studies.
Accidental injuries, especially bicycle accidents, pose a 27% risk during absence seizures in children, according to a study of 59 patients at the IWK-Grace Health Centre and Dalhousie University, Halifax, Nova Scotia.