Controlled trials in epilepsy: a review

A comprehensive review, evaluating 51 randomized double-blind controlled studies, covering different aspects of epileptology, is presented. Trials were grouped according to the investigated topic and for each group an attempt was made to derive an overall conclusion. The majority of studies investigated antiepileptic drug treatments. Other topics were: psychotropic effect of antiepileptic drugs, folic acid and vitamin D administration in epilepsy, and EEG investigations. A cross-sectional analysis of items such as designs, patient sampling principles, recording of effect parameters and side effects, concomitant treatments, and statistical evaluations demonstrated that cross-over designs, investigating fixed dosage schedules, were extensively used. Less than half of these studies included a washout period between treatments, complicating the interpretation of the obtained results. The vast majority of studies involved only chronic patients; and marked heterogeneity in patient selection with respect to age, seizure type, and mental status, and severity of epilepsy was observed. Classifications of seizures varied between the studies. The most prominent effect parameter was seizure frequency. The use of heterogeneous patient samples frequently necessitated equalization of widely different seizure types in order to perform statistical analyses. The mean duration of trials was 6 months, precluding evaluation of chronic toxicity. The majority of studies recorded side effects, but data collection was rather unsystematic and statistical evaluation was seldom applied. Most studies were add-on trials, and since concomitant treatment was frequently changed during the investigations, it was difficult to evaluate the influence of this variable. A correlation analysis across trials demonstrated, among other things, that the common assumption that short controlled trials provide too optimistic results, could not be substantiated. This survey provides no firm indication of which drug is more suitable for which seizure type.