Background Minimising exposure to factors contributing to chronic respiratory morbidity is a priority in preterm care. CPAP is established but alternatives are gaining popularity despite limited evaluation. SiPAP has not previously been compared to CPAP for first-line treatment of RDS.

Aims To compare SiPAP with CPAP as a primary mode of non-invasive respiratory support in premature infants with RDS.

Methods In this prospective two-centre trial, infants (GA 28⁺⁰ to 31⁺⁶; inborn; <6hrs old; no prior intubation; no major congenital disorders) were assigned to either SiPAP (BiPhasic Tr^©) or CPAP delivered by the Infant Flow^® SiPAP^™ device. Randomisation was stratified by centre and gestation. Crossover or use of other devices was not permitted.

The primary outcome was a pre-defined failure of non-invasive respiratory support, necessitating intubation and ventilation, in the first 72 hours of treatment. Strategies for initial settings, weaning, discontinuation and deterioration were specified. To detect a 50% reduction in failure (power 80%, α = 0.05, 2 tailed), 116 participants were required. Analyses were by intention-to-treat.

Results We assessed 368 infants at admission and recruited 120 of 149 eligible (CPAP 60, SiPAP 60). Baseline characteristics were comparable.

Failure of non-invasive respiratory support, did not differ by allocated mode of respiratory support but occurred more frequently in the lower gestational age stratum (GA < 30⁺⁰) (p = 0.004). Despite differing frequencies for some key morbidities there were no significant differences in secondary outcomes.

Conclusions For the very preterm infant, using SiPAP for first-line treatment of RDS does not confer any benefit in short-term respiratory outcome as compared to CPAP. Preterm morbidities and complications of non-invasive respiratory support were similar irrespective of allocation in this study.