Randomized controlled trials of epilepsy surgery.

Abstract

To review the following aspects of surgical randomized controlled trials (RCTs) in epilepsy: rationale, methodological issues, and state of the art. An overview of the literature and author's experience with methodological issues in surgical RCTs, contrasting them with those encountered in medical trials; a general comparison of RCTs and non-RCTs (observational studies); and a literature search for and review of existing surgical RCTs in epilepsy. Nonrandomized (observational) studies provide useful but potentially biased information about the effect of interventions. Because of their ability to deal with bias, RCTs are the optimum vehicle to obtain a true estimate of the effect of interventions, including surgery. Only seven surgical RCTs, encompassing 535 patients, have been performed in epilepsy, with variable adhesion to standard methodology for conducting RCTs. The major issues facing researchers undertaking surgical RCTs in epilepsy include timing of the study in relation to adoption of the intervention as standard practice, acceptance of randomization to an invasive procedure, timing of randomization in relation to the actual intervention, standardization of the surgical procedure, blinding, and patient recruitment. RCTs are the gold standard for evaluating surgical treatments. Despite the large number and types of surgical interventions used to treat epilepsy, only a handful has been subjected to the scientific rigour of RCTs. The challenges faced by researchers undertaking surgical RCTs in epilepsy are substantial but not insurmountable. Possible avenues to address these methodological hurdles are suggested.