Diagnostic inaccuracy in children referred with "first seizure": role for a first seizure clinic.

PURPOSE: To determine (a) the range of diagnoses, and (b) the prevalence of previous seizures in children presenting to a first seizure clinic.

METHODS: One hundred twenty-seven children were seen in a tertiary care First Seizure Clinic. Inclusion criteria were age 1 month-17 years with an unprovoked event suggestive of seizure. Data collected included referring physician specialty, child's age, gender, developmental status, and clinical diagnosis of epileptologist (nonepileptic vs. epileptic). For those with epileptic events, seizure type, syndrome (if identifiable), presumed etiology (idiopathic, cryptogenic, and symptomatic), presence of prior afebrile and febrile seizures, provoking factors, family history, pre/perinatal complications and EEG results were recorded.

RESULTS: The diagnosis was epileptic in 94 (74%), nonepileptic in 31 (24%) and unclassifiable in two (2%). Pediatricians were more likely to refer true epileptic events (92%) than ED physicians (76%) or family physicians (65%). Mean age at presentation was 8 years. Fifteen percent of children were developmentally delayed and neurological examination was abnormal in 11%. For those diagnosed with epileptic events, 32 presented with generalized while 62 presented with partial onset seizures. An epilepsy syndrome was identifiable in 15 cases. Thirty-eight percent experienced a prior probable seizure which was recognized by the referring physician in only one case. An EEG was done in all children with seizures and was abnormal in 41%. Early EEG was performed in 20% of children and did not show statistical significance.

CONCLUSIONS: Diagnostic inaccuracy is common in first seizure. One quarter of children were incorrectly diagnosed as having a seizure while the diagnosis of epilepsy was missed in over one-third of children.