Debate: When does a seizure imply epilepsy?

Abstract

Epilepsy recently has been defined conceptually as a condition of at least one seizure, with an enduring predisposition to have seizures. It is not yet clear how to make this definition operational and practical. A diagnosis of epilepsy has potentially serious consequences for health, psychosocial well-being, and economics, and, therefore, it should be made with a high level of certainty. A definite diagnosis of epilepsy can be made with two unprovoked seizures at least 24 h apart. This method has the benefit of simplicity and consistency with past epidemiologic studies. Nevertheless, certain circumstances suggest a high likelihood of having a second seizure, as evidenced by common clinical practice of considering treatment after a first unprovoked seizure in conjunction with additional risk factors (surrogate markers). One unifying approach is an operational definition of "definite epilepsy" after two unprovoked seizures at least 24 h apart. An operational definition of "probable epilepsy" can be established with one unprovoked seizure and clinical, electroencephalography (EEG), neuroimaging, genetic, or other information to suggest greater than a 50% chance of having another seizure. "Possible epilepsy" operationally would exist with a single unprovoked seizure and insufficient evidence to predict a high likelihood of recurrence. Future clinical and epidemiologic evidence would allow refinements of the operational definitions.