Preterm neonates born at less than 30 weeks' gestation are commonly maintained on invasive or non-invasive respiratory support to facilitate gas exchange. While non-invasive respiratory support (NIS) can be gradually reduced over time as the infant grows, most weaning strategies often lead to weaning failure. This failure is evidenced by an increase in significant events such as apneas, desaturations, and/or bradycardias, increased work of breathing, or an inability to oxygenate or ventilate, resulting in escalated respiratory support. Although the optimal approach to weaning NIS remains uncertain, neonatal units that delay Continuous Positive Airway Pressure (CPAP) weaning until 32-34 weeks corrected gestational age exhibit lower rates of chronic lung disease. Therefore, the investigators aim to compare the duration on respiratory support and oxygen exposure in infants born at less than 30 weeks' gestational age who undergo a structured weaning protocol that includes remaining on CPAP until at least 32-34 weeks corrected gestational age (CGA). The hypothesis posits that preterm infants following a structured weaning protocol, including maintaining CPAP until a specific gestational age, will demonstrate lower rates of weaning failure off CPAP (defined as requiring more support and/or experiencing increased stimulation events 72 hours after CPAP weaning) than those managed according to the medical team's discretion.

This is a multicenter, non-blinded, randomized control trial involving premature neonates born between 23 0/7 and 29 6/7 weeks gestational age. The trial will take place in five Neonatal Intensive Care Units (NICUs) within the Rady Children's/University of California, San Diego network, including Rady Children's Hospital - San Diego, Jacobs Medical Center, Scripps La Jolla (Rady NICU), Rancho Springs (Rady NICU), and Palomar (Rady NICU). The investigators aim to recruit 130 infants, a target sample size determined based on retrospective data from all of the participating units.

Examination of CPAP failure rates in babies < 28 weeks who were weaned off CPAP before 34 weeks CGA revealed a 62.5% failure rate, whereas those who remained on CPAP at or beyond 34 weeks exhibited a 26.7% failure rate. Thus, the investigators selected 34 weeks CGA as the time point for maintaining CPAP in babies < 28 weeks GA. For infants > 28 weeks, the retrospective review demonstrated a 76% failure rate if weaned off CPAP before 32 weeks, while those who remained on CPAP at or beyond 32 weeks showed an 11% failure rate. Consequently, the investigators chose 32 weeks CGA as the designated time point for continuing CPAP in babies with a GA of 28-30 weeks.

The sample size calculation employed a two-independent- study-group design with a primary endpoint of a binomial outcome (failed CPAP wean, yes/no). The investigators set the alpha error rate at 0.05 and power at 80%. To achieve a 50% reduction in the CPAP weaning failure rate, they aimed to enroll a total of 80 infants < 28 weeks and 50 infants 28-30 weeks (130 infants in total).

Consent will be obtained after the eligible infant has been extubated or has been stable on NIS (defined as CPAP/NIMV/NIPPV- all modes of pressure reliant respiratory support) for over 72 hours. NIS is delivered via occlusive (Fischer & Paykel [F&P]) or non-occlusive (RAM TM/Nioflo TM) interfaces at any pressure and oxygen need.

A standardized maintenance/weaning protocol will be implemented for the treatment group (standardized NIS wean) while the control group (routine care) will undergo weaning based on unit-specific practices. All infants in the treatment group will remain on CPAP until either 32 or 34 weeks CGA, depending on their GA age at birth. Infants born at 27 6/7 weeks or less will continue on CPAP until at least 34 weeks CGA if they are in the treatment group, whereas infants born at 28 0/7 to 29 6/7 will stay on CPAP until at least 32 weeks in the treatment group. The weaning protocol in the treatment group will incorporate algorithms outlining stability criteria, failure criteria, and algorithms for registered nurses (RN) and respiratory therapists (RT), including steps to take in such situations. The control group will be weaned according to the unit's or medical team's practices. According to the retrospective chart review, no standardized weaning practices have been identified at any of the sites, with decisions primarily driven by the medical team caring for the infant.

The treatment group algorithm will contain the following features:

◦ The algorithm will specify the type of NIS to use, outline how to assess the infant every 24 hours, and provide guidance on whether to wean, maintain, or increase support based on the following 3 questions:

Within the last 24 hours:

1. Has the FiO2 been less than 30%?
2. Has there been weight gain?
3. Have there been no significant events necessitating stimulation unrelated to feeding?

If the answer is 'yes' to all 3 questions, the provider can begin to wean the infant according to the algorithm's recommendations. If the answer to any of the questions is 'no', the NIS will be maintained. In the event of clinical instability, the support can be increased.

• The FiO2 should be titrated based on CGA parameters
• All babies should initially be extubated/maintained on occlusive CPAP (F&P). Proper placement of the interface is essential, and a video demonstrating accurate interface placement is provided. It is crucial to ensure that the prongs are positioned 2mm from the septum and the mask is the proper size. They should be alternated every 6 hours if that's the unit policy and the nose should be monitored for breakdown. Appropriate barriers should be applied. If the FiO2 is more than 10-20% above the baseline, if the number of stimulation events increase, and/or there is increased work of breathing, the MD/NNP should be notified. The RT and RN should refer to the Keep the PEEP algorithm (algorithms will be available at the bedside for staff reference).
• For non-occlusive CPAP, the RAM or Nioflo cannula, or equivalent can be used. The team may transition the infant to non-occlusive NIS if there is a pressure wound despite changes in the F&P interface, or if the occlusive CPAP is set at a PEEP of 7 or lower. Consider transitioning infants to non-occlusive NIS if they are over 30 weeks and not on NIMV/NIPPV. Increase the PEEP by 1-2 from the transition from occlusive to non-occlusive NIS.
• In cases of nasal breakdown, if breakdown is identified, the occlusive CPAP interface should be changed (mask to mask, mask to prongs) to reduce pressure on the wound as per unit policy. If necessary, the wound team should be contacted, and a barrier and therapeutic cream should be applied.
• To address issues with chin straps, pacifiers, or hands, the chin strap or hands should be placed under the chin, or the pacifier should be inserted into the mouth if it is open, resulting in loss of PEEP and/or worsening oxygenation/ventilation. The need for the chin strap should be reassessed every 6 hours. Ensure that the chin strap is appropriately positioned from the parietal-occipital part of scalp to the mandible (a video demonstrating proper placement will be provided), securing it so that the mouth is passively closed but the infant can still yawn and cry.

Inclusion Criteria:

1. All infants admitted to the NICU at Jacobs, Rancho Springs, Scripps La Jolla, Rady Children's Hospital, or Palomar born at < 30 weeks CGA
2. Informed parental consent obtained

Exclusion Criteria:

1. Declined or unable to give informed consent
2. Infants with known congenital anomalies or complications that require long term support (pulmonary hypoplasia, airway defects, genetic syndromes, necrotizing enterocolitis (NEC), spontaneous intestinal perforation (SIP), anything surgical)
3. Intubated for over 4 weeks of life (28 days)

Individual participant data (IPD) will be made available upon study completion per International Committee of Medical Journal Editors (ICMJE) policy. The data sharing plan will include all IPD that underlie the reported results, after deidentification (text, tables, figures, appendices). Additional documents to be made available will be the study protocol, statistical analysis plan, and informed consent form. Data will be shared with investigators whose proposed use of the data has been approved by an independent review committee identified for this purpose.

Immediately following publication and ending 24 months following article publication.

Proposals for access may be submitted up to 24 months following article publication. After 24 months the data will be available in our University's data warehouse but without investigator support other than deposited metadata.