The main goal of this trial is to test if:

automated adjustment of supplemental oxygen to preterm infants in noninvasive respiratory support based on feedback from a measurement of blood-oxygen saturation

results in more stable blood-oxygenation compared to routine nurse controlled adjustment of oxygen

Due to immature lungs and immature regulation of breathing, preterm infants require long periods of respiratory support including supplemental oxygen. Noninvasive monitoring of oxygenation using pulse oximetry (SpO2, blood oxygen saturation) has become standard of care despite not always accurately reflecting paO2 (partial pressure of oxygen). Several clinical trials have aimed to determine the optimal range for SpO2. Although subject to an ongoing debate currently most neonatal units aim to keep SpO2 within a target-range (TR) of 90-95% based on consensus guidelines. Recent reviews and meta-analyses have shown that SpO2 outside this range is associated with adverse outcome. Hyperoxia is associated with BPD (bronchopulmonary dysplasia) and ROP (retinopathy of prematurity), whereas hypoxia is associated with an increased risk of necrotizing enterocolitis, death and brain damage.

Keeping SpO2 within TR is difficult for the clinical staff and may require frequent manual adjustments of FiO2 (fraction of inspired oxygen) thus representing a huge workload. Recent studies have shown that babies often spend less than 50% of time with SpO2 within target range.

Due to limitations of manual control of FiO2, several commercially available algorithms have been developed facilitating automatic oxygen control (AOC) or closed loop automated oxygen control (CLAC) using feedback from the pulse oximetry as input. Observational trials, randomized clinical trials as well as meta-analyses of short-term outcomes have with a reasonable consistency shown automatic oxygen control to be associated with a significant increase in SpO2 spent in TR from approximately 50% to above 70-80% of time, a reduction in number and duration of episodes with hypoxemia and hyperoxemia , and a substantial reduction in the need for manual adjustments of FiO2 from more than 20 times to one during automated oxygen control.. It still remains to be determined if automatic oxygen control improves relevant long term outcomes, but a large international multi-center randomized clinical trial (FiO2-C Trial) is expected to report on long-term outcomes within 12-24 months.

The current commercially available systems for automatic oxygen control are integrated into brand- specific ventilator-systems and may only work with devices and consumables from that particular company. O2matic® is a danish automatic oxygen control system registered for use in adult patients and has been shown to be significantly better than manual control to maintain SpO2 within target O2matic® has never been tested in a neonatal population but can potentially be used as an add-on to several respiratory support systems used in neonatal care, possibly making automatic oxygen control more universally available at a lower cost. This protocol describes a proposed pilot-trial of automatic oxygen control using O2matic® in spontaneously breathing newborn infants needing respiratory support.

Research question and aim The primary objective is to evaluate if automated adjustment of FiO2 by O2Matic during neonatal nasal highflow therapy or CPAP improves the duration of SpO2 being within target range when compared to nurse-controlled (manual) adjustment of FiO2.

This will be done in a randomized controll cross over trial.

O2Matic - investigational device The O2Matic oxygen controller is a closed-loop system that, based on continuous monitoring of pulse and SpO2 by a wired pulse oximeter, adjusts oxygen flow to the patient. The algorithm in O2matic samples the last 15 seconds of input from the pulse oximeter and calculates increments or decrements in oxygen flow every second based on the last 15 seconds' average. Increments and decrements are proportionally increased relative to the difference between actual SpO2 and target SpO2. Maximal oxygen flow can be specified to fit the actual condition treated and the device used for delivering oxygen to the patient. O2matic allows for flow up to 15 L/minute in automatic mode, but will in this project be limited to 0-8 L/minute which are the flowrates used for nHFT (see further down). During the study alarms from SpO2 and heart-rate-out-of-range will be silenced and blinded (see further down) as study participants will be monitored using the usual central monitoring system used in the NICU The only alarm allowed will be for loss of signal from the oximetry sensor which can be individualized from 0 to 5 minutes, to avoid repeated alarms due to signal loss of shorter duration. The O2Matic is CE-marked and designed for adult care but not registered for use in patients less than 18 years of age. Consequently, this study also aims to test the use of O2Matic in a new patient group.

The O2Matic is only compatible with the Nonin Purelight pulse oximetry sensors - including Nonin Neonatal Oximetry sensors.

During the intervention the O2Matic will be operating in automatic mode adjusting the oxygen flow in response to changes in SpO2.

During the control period O2Matic will be deactivated.

Nasal High Flow Therapy via AIRVO-2 Nasal high flow therapy (nHFT) is a standard treatment modality in neonatal respiratory care which can be provided using the OptiFlowTM interface from Fischer & Paykell and a gas source providing a mixture of humidified air and oxygen.

The AIRVO-2TM is a humidifier with an integrated flow generator able to deliver warmed and humidified respiratory gases to spontaneously breathing patients through a variety of specific patient interfaces in varying sizes (OptiFlowTM). In clinical practice AIRVO-2 is used with OptiFlow prongs from size "preterm" and up, although the recommended size is from size "infant".

The AIRVO-2 is designed to be able to deliver flow at a constant set flow rate which will be kept constant regardless of manual changes in the fraction of inspired oxygen (FiO2) - the resulting FiO2 will be shown on the display. When used in this study the AIRVO-2 will be used in junior-mode which is necessary when used with a neonatal interface. In junior-mode humidification-temperature is limited to 34 degrees and airflow to a maximum of 25 L/min but in the current study only up to 8 L/min will be used.

Closed loop adjustment of oxygen via O2Matic and AIRVO-2 Oxygen supply to AIRVO-2 can be delivered by connecting the oxygen output from O2Matic to the oxygen inlet on AIRVO-2. In this way nHFT from AIRVO-2 can be delivered with automatically adjusted oxygen supply via the closed loop system from O2Matic at a constant set-flowrate controlled by AIRVO- 2, with the resulting FiO2 measured and displayed by AIRVO-2. O2Matic will adjust FiO2 based on the reading from the Nonin Purelight pulse oximeter connected to O2Matic. It is important to note, that the reading from this pulse oximeter will be blinded to the clinical staff, and all manual adjustments to FiO2 will be done based on readings from the pulse oximeter connected to the central monitoring system in the NICU (Avant 9600 Nonin Pulsoximeter connected via Philips Intellivue).

Study outline and trial design

The study is a pilot trial and designed as an order-randomized, cross-over trial. Infants will be randomized to treatment-order by block randomization to ensure even distribution between the sequence of treatment. All infants will at study start receive nHFT via OptiFlow interface, and then be allocated to:

• Automated adjustment of FiO2 by O2Matic followed by manual FiO2 adjustment or
• Manual adjustment of FiO2 followed by Automated adjustment of FiO2 by O2Matic

Each patient will serve as their own control, and each treatment-period will last 24 hours, and each patient will be exposed to 1 period of each treatment assignment.

Inclusion Criteria:

• Newborn infants receiving nHFT via Optiflow 4-8 l/min and FiO2>0.25, requiring adjustments of FiO2 beyond the extra need during care procedures
• Likely to be maintained on nHFT during study period of 48 hrs

Exclusion Criteria:

• major congenital abnormalities
• hemodynamic instability including large PDA with right to left shunting
• clinician's concern regarding infant stability.
• Unstable infants with frequent apneas, predominantly having need for oxygen during apneic events