The success of EEG in confirming epilepsy--revisited

This study was designed to re-examine the question of how sensitive the first EEG is in confirming a definite seizure disorder. Patients (358) with a diagnosis of "epilepsy" were randomly selected and their EEGs reviewed. An abnormal EEG was found in 84% (epileptiform paroxysms in 77-82%), while 16% had a normal tracing. However, after careful review with up to 7 years of follow-up, slightly more than 1/2 of those with a normal EEG did not actually have epilepsy, 1/4 did and in the remaining 1/4 insufficient information was available for a definite diagnosis. The incidence of normal records among patients with definite epilepsy then fell to 4.7%. With the evaluation of additional patients with "epilepsy" and a normal EEG (200), a similar distribution was found. The final diagnosis for the group without seizures was a behavioral-psychiatric condition in 1/4, syncope in 1/5 and pseudoseizure in 1/6.