How not to read an EEG

Misdiagnosis is a major public health issue. However, it is impossible to eliminate diagnostic error from the treatment decision-making process.¹ Neurologic conditions, such as stroke, multiple sclerosis, and epilepsy, are at risk for missed, delayed, and wrong diagnoses.² For patients with seizures, up to 30% referred to epilepsy monitoring units are ultimately discovered to have an alternate diagnosis.^3,4 A significant number of patients identified with psychogenic nonepileptic events have had at least one EEG that was misinterpreted as abnormal.⁴ Seizure-related histories are sometimes nebulous and incomplete. Therefore, over-reliance on the EEG may occur to help exclude a diagnosis, support a diagnosis, or render treatment for epilepsy.⁵ While a misinterpreted EEG is a well-known problem in epilepsy management,⁶ some electrographic patterns are difficult to discern even in expert hands.⁷ An interictal spike is the hallmark of epilepsy, yet it has been referred to as an “EEG chameleon” because it is often incompletely evident.⁸ Therefore, definitive statements regarding the epileptic origin of a “spell” are possible only when seizures are recorded on the EEG. Without recording a seizure, the EEG is only supportive when diagnosing epilepsy. However, the supportive nature of an interictal EEG is only valid if clear interictal epileptiform discharges (IED) are encountered. When overidentified IED are reported, the misinterpretation of the EEG often goes unchallenged until ictal recordings are performed.