Predictive value of electroencephalography and computed tomography in childhood non-traumatic coma

Abstract

To study value of electroencephalogram (EEG) and computed tomography (CT Scan) in predicting outcome of non-traumatic coma in children. 100 consecutive children, between 2 months to 12 years, with non-traumatic coma, (Glasgow Coma Scale score < 8). Demographic and clinical data was recorded at admission. EEG and CT scan were done within 24 hours of admission. Etiologic diagnosis was assigned on basis of clinical data and relevant laboratory investigations. The outcome was recorded as survived and died. Among survivors it was graded as no disability, or mild, moderate, or severe disability. Odds ratio and/or relative risk (RR) with 95% confidence interval (C I) were calculated. EEG could be done in 60 patients (43 survived; 7 were normal, 8 had mild, 17 moderate and 11 severe disability) CT scan in 93 patients (60 survived; 11 were normal, 14 had mild, 21 moderate and 14 severe disability). A normal/borderline EEG was associated with good outcome (P = 0.001); 11 of 12 survived and of survivors 55% had no or mild disability. Electrocerebral silence on EEG was a predictor of death (OR = 44 -sub .95% Cl - 1.5-7372; P = 0.01). An abnormal EEG was associated with significant increase in risk of disability among survivors (RR = 2.6, 95% Cl = 1.2-5.4, P = 0.03). Among CT abnormalities intracranial bleed suggested increased risk of death (RR = 2.1; 95% Cl - 0.8-5.3; P = 0.058), while, hydrocephalus was associated with better survival (RR = 0.7; 95% CI - 0.5 to 0.96; P = 0.029). However, hydrocephalus when compared with other abnormal CT scan findings, was associated with higher risk of moderate and severe disability among survivors (P = 0.046) A normal CT scan and EEG, and some of the specific findings could be helpful in predicting outcome in children with non-traumatic coma. EEG and CT scan should be done at admission in all patients with non-traumatic coma if feasible.