Are prodromes preictal events? A prospective PDA-based study

Up to 29% of patients with epilepsy report “prodromal” sensations more than 30 minutes prior to seizures. We developed and implemented an objective methodology to prospectively assess the sensitivity and specificity of these subjective experiences using personal digital assistants (PDAs). The key property, in contrast to paper-based diaries, is the internal recording of the patient's entering time of prodromes and seizures. Of 500 patients with epilepsy interviewed, 31 claimed to sense prodromal symptoms at least 30 minutes before seizure onset. Eleven of them agreed to participate in a 4-week study to objectively measure their prospective prediction performance. In 9 patients returning data, the majority of prodrome entries were not followed by seizures or were identified only retrospectively. Statistical analysis revealed that no patient could outperform a nonspecific random predictor when predicting seizures based on the occurrence of prodromes, and that the group performance matched precisely the expected result for a by-chance prediction. These results question the predictive value of “prodromes” and the specificity of their occurrence in the preictal period.

Research Highlights
► The predictive value of subjectively experienced “prodromes” was prospectively assessed.
► Patients were selected based on their conviction that they could predict seizures from experiences occurring more than 30 minutes prior to a seizure.
► Patients entered prodromes and seizures into electronic diaries for statistical evaluation of prediction performance.
► No patient outperformed an unspecific random predictor; sometimes, prodromes were retrospectively classified after occurrence of a seizue.
► The results question the predictive value of “prodromes” and the specificity of their occurrence in the preictal period.