Abstracts

Rationale: The gold standard for diagnosis of psychogenic non-epileptic seizures (PNES) is EEG-video monitoring (vEEG), but there is no confirmatory test such as pathology. In the absence of a pathophysiological comparator, interrater agreement may be the best measure of diagnostic reliability for vEEG. In this pilot study, we sought to evaluate interrater reliability using vEEG for patients commonly seen in monitoring. Methods: We prepared a DVD of monitoring samples from 22 unselected consecutive patients who underwent vEEG at our EMU and who had at least an episode recorded. The data were collected in a standard fashion on XLTEK that included interictal samples and all recorded episodes. Results of other tests (e.g., imaging) and extensive histories were not provided, as the aim was to assess the reliability of the diagnosis based on vEEG data for interpretation. The data discs were sent to 25 reviewers and 15 completed the survey. Reviewers were board-certified neurologists and epileptologists at tertiary epilepsy centers. Survey choices were: 1) PNES; 2) Epilepsy; 3) Non-epileptic but not psychogenic ('physiologic'). Interrater agreement was measured using a kappa coefficient. Results: For the diagnosis of PNES, there was fair agreement (kappa=0.49, 95% CI: 0.45 - 0.53). For the diagnosis of epilepsy, there was good agreement (kappa=0.65, 95% CI: 0.61 - 0.69). For physiologic nonepileptic episodes, the agreement was low (kappa=0.09, 95% CI: 0.05 - 0.13). Including all three categories the kappa-statistic was 0.49 (95% CI: 0.46 - 0.52). Four patient samples had considerable variability in diagnosis. When these samples were excluded, the kappa-statistics were PNES=0.6, Epilepsy=0.75 and Physiologic=0.36. Conclusions: In this pilot study, we found good agreement for the diagnosis of epilepsy. As noted in the literature where seizure descriptions are given to epileptologists, the interrater reliabilty of identifying PNES by EEG-video was only fair, further demonstrating the challenge in making this diagnosis. Diagnosis of PNES by vEEG alone poses a diagnostic challenge and incorporation of history and neurpsychiatric examination may increase the diagnosistic accuracy. Low agreement among physiologic nonepileptic episodes may be attributed to small sample size. A fully powered study across epilepsy centers may provide a representative picture of interrater agreement.