As a result of a slower inclusion rate than expected, competence of medical staff to work with the AVEA ventilator could no longer be guaranteed.

Premature infants often receive respiratory support and supplemental oxygen for a prolonged period of time during their admission in the NICU. While maintaining the oxygen saturation within a narrow target range is important to prevent morbidity, manual oxygen titration can be very challenging. Automatic titration by a controller has been proven to be more effective. However, to date the performance of different controllers has not been compared. The proposed randomized crossover trial Comparing Oxygen Controllers in Preterm InfanTs (COCkPIT) is designed to compare the effect on time spent within target range. The results of this trial will help determining which algorithm is most successful in controlling oxygen, improve future developments in automated oxygen control and ultimately reduce the morbidity associated with hypoxemia and hyperoxemia.

Both hypoxemia as hyperoxemia can potentially be harmful to premature infants. Oxygen titration during respiratory support is vital to prevent these conditions but is very challenging. In the investigator's neonatal intensive care unit preterm infants routinely receive automatic oxygen titration performed by a controller. The currently used controllers are both proven to be more effective than manual titration, however which of the two controllers is most effective in keeping oxygen saturation within target range remains unclear.

This randomized crossover trial tests tests both controllers within every study patient to determine which controller is most effective and thereby would hopefully reduce morbidity associated with hypoxemia and hyperoxemia the most. The primary outcome measure is the proportion of time spent within target range, each controller will be tested for 24 hours within the same study subject. This is excluding a 1-hour wash-out period after a change in ventilator.

Eligible infants are randomized to start with either the Oxygenie algorithm or CLiO2 algorithm and will switch to the other study arm after 24 hours of measurement.

Inclusion Criteria:

• Preterm infants with a gestational age (GA) at birth of 24 - 29+6/7 weeks
• Receiving invasive mechanical ventilation or non-invasive respiratory support (NCPAP or NIPPV)
• Receiving supplemental oxygen (defined as FiO2 ≥ 0.25) at the time of enrollment and for at least 18 hours during the previous 24 hours; Or a coefficient of variation in supplemental oxygen of ≥ 0.1 in the previous 24 hours.
• Expected to complete the 49-hour or 50-hour study period in the current form of respiratory support, i.e. invasive mechanical ventilation or non-invasive respiratory support
• Written informed parental consent must be present.

Exclusion Criteria:

• Major congenital anomalies
• Arterial hypotension requiring vasopressor therapy within 48 hours prior to enrollment.
• If the attending physician considers the infant not stable enough for a switch to another ventilator.

The outcome assessor/investigator will be blinded for treatment allocation