Introduction First seizure diagnosis may be delayed due to financial, geographical or social barriers to healthcare, or misdiagnosis with differentials including syncope or stroke. Seizures may recur until correct diagnosis and appropriate treatment is instituted; meanwhile, patients may experience increased seizure-related morbidity and mortality. We compare patient and seizure characteristics between a first-ever ‘new-onset’ seizure (NOS) cohort, and a recurrent-untreated seizure (RUS) cohort.

Method Medical charts were reviewed to extract information on patient demographics and clinical characteristics using a standardised proforma. Inclusion criteria were patients aged 18 or over who attended a tertiary-level Melbourne hospital between 1 ^{January 2008} and 30 November 2016 with discharge codes ICD-10 G40-Epilepsy, G41-Status epilepticus, or R56.9-Unspecified convulsions.

Results 367 episodes were identified. 151 episodes met inclusion criteria: new-onset seizures (115) and recurrent-untreated seizures.³⁶ 216 excluded cases included pre-existing epilepsy (186), and non-seizure events.³⁰ RUS-cohort experienced a median of two seizures prior to coming to medical attention, most commonly focal impaired awareness seizures (50.00%). Considering the index seizure, focal seizures were more common in RUS-cohort (36.11 vs 24.35%) while primary generalised seizures predominated in NOS-cohort (62.61% vs 50.00%). Compared to NOS-cohort, RUS-cohort was more likely to have unprovoked seizures (72.22% vs 55.65%), identifiable remote risk factors (41.67% vs 26.09%), younger age (69 vs 76), normal MRI and EEG, and be discharged on antiepileptic drugs (86.11% vs 73.91%). RUS-cohort was more likely to receive Neurology outpatient follow-up (72.22% vs 39.99%), and in a more timely manner compared to NOS-cohort (30.56% vs 11.31% saw a Neurologist within a month of discharge).

Conclusion Recurrent-untreated seizures often have subtler semiology and are more likely to have normal MRI and EEG results than patients presenting immediately following new-onset seizures. RUS-cohort tend to receive more inpatient investigations and AED prescriptions, and are offered more timely neurology follow-up than NOS-cohort.