17 1 General introduction and outline of this thesis cohort studies24-26 and randomised controlled trials27-29 point to the difficulty of SpO 2 targeting, with most studies reporting SpO2 values to be within the target range less than half of the time. Although bedside staff frequently adjusts FiO2 to maintain SpO2 within the target range prescribed by the clinician, their workload limits time availability and makes continuous tailoring of FiO2 to the infant’s needs difficult. This is further complicated by the neonatal oxygenation physiology being unpredictable and non-linear, with a long time delay between an adjustment in FiO2 and a stable SpO2. 30 Even in the presence of a dedicated respiratory therapist to titrate FiO 2, time within target range was only 66%.31 In premature infants with frequent fluctuations in oxygenation, clinical personnel usually respond to the occurrence of alarms in the pulse oximeter triggered by episodes of hypoxaemia with a manual increase in FiO2. When these episodes resolve and SpO2 returns to the desired range, FiO2 should be reset to the basal level. However, under routine clinical conditions, staff limitations can result in inconsistencies in response and timing. As a consequence, premature infants are often exposed to periods of insufficient oxygenation, unnecessary oxygen exposure and hyperoxaemia.32 Also, in these infants, the FiO 2 set by the bedside staff often exceeds the level required to maintain an acceptable range of SpO2. This is done in an attempt to reduce the frequency of the hypoxemic episodes. However, this is not always effective and can increase the exposure to supplemental oxygen and hyperoxaemia.33 Considering the effect of target range deviations and the difficulty of targeting SpO2, feedback-controlled adjustment of FiO2 by a machine –an automated oxygen controller (AOC)– is a logical improvement on current practice. In essence, SpO2 readings are continuously fed into a device holding a set of computational instructions (an algorithm), which then gives an output, an updated value for FiO2. The effectiveness of automated control of inspired oxygen and its effects on the fluctuation of oxygenation during the care of premature infants may result in improved neurodevelopmental outcomes.34 Randomised trials comparing automated FiO2 systems with manual titration for short periods demonstrated an increase in the proportion of time spent with SpO2 within target range varying between 8% and 24%.35-43 Automated FiO 2 control also decreased the required nursing time in preterm infants with frequent severe desaturations.36, 37, 44 Several automated oxygen control devices are commercially available and used in NICUs, but it is unknown whether these devices lead to different clinical or long-term outcome. Furthermore, it is unknown which of these controllers is most effective, as no comparisons have been made between the performance of different AOCs.