A reappraisal of the value of interictal EEG findings in diagnosing epilepsy plus a critical review of controversial "normal variants", utilising Long-Term Ambulatory EEG recordings

Utilising long-term ambulatory EEG recordings over a period of 72 hours interictal epileptiform potentials (IED’s, interictal epileptiform discharges) could be recorded in 311 from 316 patients with a confirmed epilepsy diagnosis (seizures recorded). These figures allow us to calculate a PPV of 98.4%. These results allow us to predict the chance of an unclear seizure being epileptic if we record IED’s as being over 98%. Perhaps of more use in the daily clinical practice are our results in patients with non-epileptic seizures where the chance of an unclear attack being epileptic is only 1.5% (NPV 98.5%) if there were no IED’s in a 72-hour recording. Although it still holds true that spikes without seizures are not diagnostic of epilepsy and a positive diagnosis of epileptic seizures is still based on a combination of clinical and EEG findings we can say that with unclear attacks and IED’S the chances of the attack being epileptic are very high. Alternatively, if no spikes are recorded the chance of the attack being epileptic is negligible. We identified, in three of the more controversial “normal variants” (BETS, wicket spikes and 6/s spike-and-wave) two distinct differences which only in long-term EEG recordings could be defined. Depending on morphology and topography they could be considered either to be true normal variants or pathological patterns associated with epilepsy. For practical purposes we suggest that in patients with the variations associated with epilepsy that seizures should be recorded before the final diagnosis is made.