METHODOLOGY: A prospective observational study on all consecutive children with head injuries at the General Hospital Kuala Lumpur between November 1993 and December 1994. The onset, type and frequency of seizures occurring within the first week of injury were documented. Using inpatients as a cohort, logistic regression analysis was used to determine clinical and radiological variables significantly associated with seizures. The outcome 6 months post-injury was assessed using the Glasgow Outcome Scale.
RESULTS: Fifty-three of 966 children (5.5%) developed seizures within the first week of trauma. Seven (13.2%) occurred within 1 h of injury, 30 (56.6%) between 1 and 24 h and 16 (30.2%) after 24 h. Factors significantly associated with early post-traumatic seizures were female sex, age less than 2 years, loss of consciousness for more than 24h and acute subdural haematoma (P<0.01). Children with seizures had a poorer outcome (death or severe disability) than inpatients without seizures (21/53 vs 19/182, P<0.001). The outcome was worst in children with recurrent partial seizures, who had a longer injury-seizure interval and were more likely to have focal neurologic deficits compared to those with sporadic or generalized seizures.
CONCLUSIONS: Anticonvulsant prophylaxis to minimize the adverse effects of early seizures in head injury should be considered for young children (less than 2 years old) with subdural haematoma and a prolonged duration of coma. Prompt and effective control of recurrent seizures is recommended.