Evidence-based medicine and antiepileptic drugs.

Evidence based health care uses systematic literature reviews with statistical strategies like meta-analysis to aid decision-making. This information can help clinicians by organizing data and providing up-to-date quantitative summaries of efficacy and adverse effects of treatments. Limitations of meta-analysis include problems inherent in combining data from trials of somewhat different design, choice of appropriate dosages, and summarizing complex questions as a single odds ratios. I summarize the results of a meta-analysis of the following antiepileptic treatments for partial seizures in adults: gabapentin, lamotrigine, topiramate, tiagabine, valproate and the vagal nerve stimulator. Each treatment was significantly more efficacious than placebo, and there were nonsignificant trends toward differences among the treatments in efficacy and tolerability. Quantitative analysis of adverse effects is presented. Absent the availability of a comprehensive randomized controlled trial for comparison, a rigorously conducted meta-analysis provides some useful information.