DIAGNOSING CHILDREN PRESENTING TO A TERTIARY CARE, FIRST SEIZURE CLINIC: HOW WELL DO EPILEPTOLOGISTS AGREE?

Objectives: To determine diagnostic agreement rates between epileptologists for children seen in a First Seizure clinic.

Methods: Between 01/01/2004 and 30/08/2005, 127 children were seen in a First Seizure Clinic at the Alberta Children's Hospital by one of two epileptologists. Inclusion criteria were age >1 month and an unprovoked event suggestive of seizure. Following review of the child's (1) clinical history and physical examination and (2) clinical data plus EEG results, each epileptologist independently gave the most likely diagnosis (epileptic vs. non-epileptic), and if epileptic, classified seizure type (generalized, partial, unknown). After review of both clinical and EEG data, epileptologists independently classified etiology (idiopathic, cryptogenic, symptomatic) and syndrome (if possible), and predicted recurrence risk (<25%, 25=50%, 50=75%, 75% or greater). Results: Based on clinical history and physical examination data alone, inter-rater agreement for the diagnosis of epileptic vs. non-epileptic events was high (98%). For the 96 children who were diagnosed as having epileptic seizures, inter-rater agreement for seizure type after review of the clinical history and physical examination data alone was 96%, and after review of clinical and EEG data was 96%. Agreement for seizure etiology was 95% and for seizure syndrome was 100%. Agreement for recurrence risk was slightly lower at 81/96 (85%).

Conclusion: Inter-rater agreement for the diagnosis of first events was high in our centre. While addition of the EEG data did not significantly affect inter-rater agreement, it allowed greater numbers of patients to be classified into generalized vs. partial onset seizures, and assisted with syndrome identification.