The EpiNet Collaboration: Do Neurologists Around the World Agree when Diagnosing Epilepsy? - Results of a Multinational Study. (S14.004)

ObjectiveTo determine variation in epilepsy diagnosis when neurologists are presented with identical case scenarios.BackgroundThe difficulties in diagnosing epilepsy are under recognised. The EpiNet steering group (www.epinet.co.nz) assessed the inter-rater agreement among neurologists in diagnosing epilepsy, seizure type(s) and etiology.Methods:Epileptologists and neurologists with an interest in epilepsy were invited to participate in the study between December 2013 to September 2014.32 case scenarios describing real patients (6 children, 26 adults) with various events were evaluated online.For each patient, participants were asked: to determine how likely it was the patient had epilepsy (epilepsy, uncertain, not-epilepsy); if epilepsy, to classify the patients’ seizures using the ILAE 2010 classification; and to determine the etiology (structural/metabolic; genetic; or unknown), and epilepsy syndrome when appropriate.The “gold standard” diagnosis was determined by the EpiNet steering committee. 21patients had epileptic seizures, of whom 11 had focal seizures and 8 generalised seizures; in 2 the steering committee was uncertain if seizures were focal or generalised; 11 had non-epileptic events.In 24 cases, information was presented in 2 steps; clinical information was presented in Step 1, and after completion of this step, neuroimaging and EEG studies were provided in Step 2.Results322 neurologists commenced the study. 201 participants from 35 countries completed the 32 cases. Full data was available for the 24 cases with 2 steps from 191 investigators. Kappa values for step 1 were: diagnosis of epilepsy = 0.55; seizure type(s) = 0.36; etiology = 0.36.Kappa values increased at Step 2 for diagnosis of epilepsy (0.68), seizure type (0.48) and etiology (0.44).ConclusionConcordance between neurologists assessing identical case scenarios for the diagnosis of epilepsy is moderate when based on history alone. In contrast, concordance on seizure type and etiology is poor. Concordance improves when results of investigations are included. Disclosure: Dr. Bergin has nothing to disclose. Dr. Beghi has received personal compensation for activities with GlaxoSmithKline, UCB Pharma, and ViroPharma. Dr. D9Souza has nothing to disclose. Dr. Sadleir has nothing to disclose. Dr. Tripathi has received research support from the government of India. Dr. Richardson has nothing to disclose. Dr. Bianchi has nothing to disclose.