Midazolam fails to prevent neurological damage in children with convulsive refractory febrile status epilepticus.

BACKGROUND

We conducted a retrospective study to compare the outcome of intravenous midazolam infusion without electroencephalography or targeted temperature management and barbiturate coma therapy with electroencephalography and targeted temperature management for treating convulsive refractory febrile status epilepticus.

PATIENTS: Of 49 consecutive convulsive refractory febrile status epilepticus patients admitted to the pediatric intensive care unit of our hospital, 29 were excluded because they received other treatments or because of various underlying illnesses. Thus, eight patients were treated with midazolam and 10 with barbiturate coma therapy using thiamylal. Midazolam-treated patients were intubated only when necessary, whereas barbiturate coma therapy patients were routinely intubated. Continuous electroencephalography monitoring was utilized only for the barbiturate coma group. The titration goal for anesthesia was clinical termination of status epilepticus in the midazolam group and suppression or burst-suppression patterns on electroencephalography in the barbiturate coma group. Normothermia was maintained using blankets and neuromuscular blockade in the barbiturate coma group and using antipyretics in the midazolam group. Prognoses were measured at 1 month after onset; children were classified into poor and good outcome groups.

RESULTS

Good outcome was achieved in all the barbiturate coma group patients and 50% of the midazolam group patients (P = 0.02, Fisher's exact test).

CONCLUSIONS

Although the sample size was small and our study could not determine which protocol element is essential for the neurological outcome, the findings suggest that clinical seizure control using midazolam without continuous electroencephalography monitoring or targeted temperature management is insufficient in preventing neurological damage in children with convulsive refractory febrile status epilepticus.