McPherson 2015.

Study characteristics
Methods	Study design: RCT Location: USA Setting: NICU (St. Louis Children’s Hospital, Missouri) Duration: November 2008 to June 2010
Participants	Inclusion criteria: 74 infants ≤ 32 weeks' gestational age admitted to the NICU. Exclusion criteria: infants who had a known congenital anomaly, were moribund or in respiratory failure (defined as requiring > 80% FiO2 for 6 hours and/or having more than 2 inotropic drugs excluding hydrocortisone), or had severe brain injury (grade 3 to 4 intraventricular hemorrhage (IVH)) in the first 24 hours of life. Infants who were not expected to survive past 72 hours of life were also excluded.
Interventions	High dose: loading dose of 40 mg/kg followed by 20 mg/kg 12 hours later, then 10 mg/kg at 24 and 36 hours after the initial dose (80 mg/kg total over 36 hours). Low dose: 20 mg/kg followed by 10 mg/kg 24 hours after the initial dose (30 mg/kg total over 36 hours).
Outcomes	Outcomes reported that are considered for this review: mortality during first admission; chronic lung disease at 36 weeks of corrected age; severe IVH ≥ grade 3; IVH, any grade; cerebellar hemorrhage; periventricular leukomalacia (PVL); lesions indicative of brain injury (detected by US and MRI): IVH, any grade; PVL; white matter injury, deep grey matter injury; cerebellar hemorrhage; duration of MV; retinopathy of prematurity ≥ grade 3; cognitive delay (at 2 years of age): assessed with Bayley‐III scores for cognitive development; seizure before discharge; need for treatment of PDA; necrotizing enterocolitis (NEC); and at 5 years of age: standardized neurodevelopmental tests; parent reports of child socioemotional problems.
Notes	Caffeine initiation time: within 24 hours of birth Baseline imbalances: in the high‐dose caffeine group the maternal age was higher than in the low dose group (P = 0.03) This study was reported as two different publications: ‐ McPherson 2015: A pilot randomised trial of high‐dose caffeine therapy in preterm infants ‐ Vesoulis 2016: Early high‐dose caffeine increases seizure burden in extremely preterm neonates: a preliminary study
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Group assignment was performed by a parallel 1:1 blocked randomization, generated by the dispensing pharmacist who was not involved in clinical care.
Allocation concealment (selection bias)	Unclear risk	Not described
Blinding of participants and personnel (performance bias) All outcomes	Low risk	The clinical and research team remained blinded to each infant's randomization until completion of developmental assessment at two years of age.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	All testers were blinded to study assignment and past medical history, including imaging findings.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Out of 37 infants allocated to each group: 28 versus 30 underwent MRI and 28 and 31 underwent neurodevelopmental testing at term‐equivalent age, in the high‐ and low‐dose groups respectively. 24 and 22 underwent two year developmental assessment in the high‐ and low‐dose groups respectively. 8 infants in the high‐dose group and 7 in the low‐dose group were excluded due to corrupt data files or recordings equal or less than five hours in length from the EEG recording (for the assessment of seizures) through the 72 first hours of postnatal life
Selective reporting (reporting bias)	Low risk	The trial protocol is available at www.clinicaltrials.gov/ct2/show/NCT00809055 (25 April 2018). All pre‐specified outcomes were reported in the manuscript.
Other bias	Unclear risk	In the high‐dose caffeine group, the maternal age was higher than in the standard‐dose group (P = 0.03).