Abstracts

Rationale: We wanted to determine how much variability there is in neurologists' use of EEG when discontinuing anti-epileptic drugs (AEDs) in patients with epilepsy who are seizure-free. Methods: An Internet-based questionnaire was sent to adult and paediatric neurologists with an interest in epilepsy in New Zealand, Australia, Pakistan, Malaysia, Great Britain, Italy, Belgium and Canada. The questionnaire addressed physician's use of EEG in deciding whether to discontinue AEDs in seizure-free patients. The questionnaire contained 7 common clinical scenarios allowing open ended and closed responses. We sought to determine: 1) in what circumstances neurologists obtain an EEG when considering withdrawal of AEDs in patients who are seizure-free; 2) how the results of the EEG influence the neurologist's decision; 3) what other clinical factors affect AED withdrawal; 4) the rate of AED withdrawal. Results: Neurologists from all 8 countries responded. There was substantial variability in the use of EEG, both between and within countries. Of the initial 100 respondents, 45% always obtain an EEG before withdrawing an AED in a seizure-free patient taking a single drug, 37% sometimes obtain an EEG, and approximately 18% never do so. Approximately 30% of doctors report that they never obtain an EEG when discontinuing one of several AEDs. A routine 20-40 minute recording is most widely used. The most important factors determining whether an EEG is obtained are the epilepsy syndrome, the seizure type, the presence of epileptiform discharges in an earlier EEG, and uncertainty regarding the original diagnosis. In only 3 of the 7 scenarios did more than half the epileptologists agree on the approach. Approximately 2/3 would be guided by the EEG when considering withdrawal of AEDs in a child with absence epilepsy. In the other 2 scenarios where more than half agreed on the management, the majority of neurologists stated they would advise the patient not to discontinue the AED, regardless of what the EEG showed. Fewer than 20% of neurologists would be guided by an EEG in other clinical scenarios. Most neurologists withdraw AEDs over 3 to 6 months, but significant numbers withdraw more rapidly or more slowly. Conclusions: There is considerable variability in the use of EEG when withdrawing AEDs in seizure-free patients. There was little agreement regarding its usefulness in several common clinical scenarios. A multicentre trial should be considered to guide rational decision-making in this common clinical setting.