Diagnostic Yield of Neuroimaging and Electroencephalography in Children with Recurrent Headaches

Abstract

Neuroimaging and electroencephalography (EEG) are commonly used for pediatric recurrent headache evaluation, although not routinely recommended. In this study, data of 517 children with recurrent headaches were analyzed to evaluate the diagnostic yield of neuroimaging and EEG. Neuroimaging was performed in 55% (n = 283) of children, 61% with magnetic resonance imaging and 39% with computed tomography. Abnormal findings were reported in 17% (n = 48), and 5% (n = 13) were significant abnormalities altering headache management. Among children with significant imaging abnormalities, 85% had alarming signs on history, physical, or neurological examination. The diagnostic yield of imaging was 0.7% (n = 2), picking a cavernous hemangioma and a giant arachnoid cyst. EEG was performed in 69% (n = 356) of children, significant findings were reported in 8% (n = 29), with sharp waves/spikes in 4.1% (n = 15), and intermittent focal slow waves in 3.9% (n = 14). EEG was diagnostic for epilepsy in 1.6% (n = 6) of children with recurrent headaches. When there are no red flags on history or physical examination, diagnostic yield of neuroimaging is low for pediatric recurrent headaches. EEG can be helpful in selective cases, and when performed during and at headache-free periods.