The purpose of this study is to determine if in preterm infants < 34 weeks' gestation at birth receiving respiratory support with continuous positive airway pressure (CPAP) or nasal cannula (NC), CPAP compared with NC will decrease the number of episodes with oxygen saturations less than 85% of ≥10 seconds in a 24-hour randomized controlled trial. This will be a randomized controlled trial with a 1:1 parallel allocation of infants to CPAP or NC oxygen using stratified permuted block design.

We hope to determine whether in preterm infants < 34 weeks' gestation at birth receiving respiratory support with continuous positive airway pressure (CPAP) or nasal cannula (NC), CPAP compared with NC will decrease the number of episodes with oxygen saturations less than 85% of ≥10 seconds in a 24-hour randomized controlled trial.

This study will include preterm infants < 34 weeks' gestation on respiratory support via CPAP with a PEEP ≤ 5 and FiO2 ≤ 30%. There will be two randomization strata [≥ 22+0/7 to ≤ 27+0/7 weeks, and ≥ 27+1/7 to ≤ 33+6/7 weeks. The purpose of stratification is to ensure an appropriate distribution of risk between study arms. This study will not be powered to detect outcome differences within or between strata.

Following informed consent, randomization, stratified by gestational age at delivery, will be performed using sequentially numbered sealed opaque envelopes. Each envelope will indicate either Treatment group (CPAP group) or Control group (NC group). The envelope will only be opened after informed consent has been obtained and just before starting the study on each infant.

This will be a single center randomized controlled trial with a 1:1 parallel allocation of infants to CPAP or NC oxygen using stratified permuted block design. 15-30 minutes will be provided as a washout period at the beginning of the intervention, followed by 24 hours on the intervention. Infants enrolled must meet CPAP stability criteria that are based on prior randomized clinical trials of weaning from CPAP to NC.

All infants enrolled in the study will have routine monitoring, uniform target saturation ranges of 91-95% with alarm limits set at 88-95%, and standard care for the duration of the study. The high alarm limit will be increased to 100% if an infant is weaned to 21% FiO2. Supplemental FiO2 will be titrated per unit routine to achieve goal target saturations. Pulse oximetry recordings will be downloaded using ixTrend (iexcellence, Wildau, Germany) software to a secure computer system or via Bedmaster software for later data analysis. The target oxygen saturations (91 to 95%) are based on data from the meta-analysis of randomized controlled trials of oxygen saturation targets which included data on 4911 infants from the SUPPORT, COT, and BOOST II trials.

Primary and secondary outcomes are described below. Other safety outcomes include recordings of episodes of bradycardia and circumstances surrounding the event. Pulse oximetry recordings will be downloaded using ixTrend software to a secure computer system for later data analysis. Abdominal and cerebral NIRS will also be performed with subsequent data analysis..](streamdown:incomplete-link)

Inclusion Criteria:

• Infants must be currently utilizing CPAP (without a rate) with a PEEP ≤ 5 and FiO2 ≤ 30%
• Infants must meet CPAP stability criteria as follow:
• If previously intubated, must be extubated ≥ 72 hours
• < 3 self-resolving apneas (≤ 20 s) and/or bradycardia (< 100 bpm) in any hour over previous 6 hours)
• Gestational age < 34 weeks' gestation at birth
• Informed consent by parents/legal guardians

Exclusion Criteria:

• presence of a major malformation
• a neuromuscular condition that affects respiration
• a terminal illness or decision to withhold or limit support
• currently being treated for sepsis
• enrollment in a competing trial