Boylan 2004.

Study characteristics
Methods	Randomised controlled trial
Participants	Neonates with seizures who failed to respond to first‐line phenobarbitone treatment (Quote:) "Neonates at high risk of developing seizures because of birth depression or cord blood acidosis, had abnormal movements suggesting seizures, or had meningitis. Neonates who had already received a single loading dose of phenobarbitone were not excluded from the study." Sample size was 27 neonates with EEG‐confirmed seizures, 5 were excluded because of protocol violations, 11 because they responded to phenobarbitone. 11 neonates were included in the analysis because they required second‐line treatments (3 clonazepam, 5 lignocaine, 3 midazolam). The study was performed in 2 neonatal intensive care units in London, UK. Information on study dates is not included in the publication.
Interventions	First‐line treatment (in all neonates, before randomisation): phenobarbitone in a dose of up to 40 mg/kg. (Quote:) "If this failed to abolish seizures or reduce the seizure burden by at least 80% within 12 hours of enrollment, the neonate was randomly assigned to receive midazolam or lignocaine as second‐line anticonvulsant therapy." Second‐line treatments: ‐ Midazolam bolus dose of 60 μg/kg followed by an infusion of 150 μg/kg/h, increased to either 300 μg/kg/h if midazolam failed to abolish or reduce seizure burden by at least 80% within 12 hours; ‐ Lignocaine bolus of 4 mg/kg over 20 minutes followed by an infusion of 2 mg/kg/h, increased to 4 mg/kg/h if midazolam failed to abolish or reduce seizure burden by at least 80% within 12 hours; Clonazepam was administered (quote:) "if the increased dose of either drug failed to improve the seizure burden within 48 hours of enrollment" or (quote:)"if parents were not willing for their child to be given a drug chosen randomly".
Outcomes	Primary endpoint: control of electrographic seizures, defined as (quote:) "complete absence of seizure activity on the EEG or a reduction of > 80% of pretreatment burden" Other endpoints: neurodevelopmental assessment evaluated with Amiel–Tison and Griffiths neurodevelopmental assessment at 1 year.
Notes	Response to treatment was assessed using continuous video‐EEG. (Quote:) "All neonates were monitored continuously for at least 24 hours after enrollment. If electrographic seizures were not detected during this time, recording was stopped. If seizures were present, monitoring was continued until seizure control was established or treatment was considered to have failed (at least 48 hours later)." Neonates receiving lignocaine continued to be given background midazolam at a dose of 30 to 60 μg/kg/h. Some neonates were also receiving continuous low‐dose morphine as analgesia (10 to 20 μg/kg/h). External funding sources or possible conflicts of interests were not mentioned in the publication.