Hylke Salverda

155 9 General discussion and future perspectives Introduction Very preterm infants have an immature respiratory system leading to inadequate ventilation and gas exchange, eventually resulting in hypoxia. Supplemental oxygen is therefore often provided to prevent damage associated with hypoxia.1-3 In response to hypoxic episodes the concentration of supplemental oxygen is increased but, when not titrated down promptly after resolution of the event, can lead to iatrogenic hyperoxia. Both hypoxia and hyperoxia have been associated with organ injury.4-8 To guide titration of supplemental oxygen the infant’s oxygen saturation (SpO2) is used, in an attempt to minimise the occurrence of episodes outside the intended range. Despite best efforts with manual titration, the therapeutic range is narrow and preterm infants still spent up to 50% of the time outside this intended therapeutic oxygen saturation target range.9 Maintaining a stable oxygenation within the narrow therapeutic range is hampered by a multiplicity of reasons: (1) The neonatal oxygenation physiology is unstable, exemplified by respiratory pauses also called apnoea of prematurity. (2) The infant’s response to a therapeutic change in supplemental oxygen is non-linear, and there is a significant time delay between a change in the concentration of inhaled oxygen (FiO2) and a change in the infant’s SpO2. 10 (3) Workload of bedside staff may limit continuous titration to the infants need11 and there appears to be a tendency to accept a higher oxygen saturation by bedside staff, possibly because the infants are believed to be more stable and temporary hyperoxia may be considered less harmful.12, 13 The contribution of some of these issues to morbidity and mortality may be reduced by using an automated oxygen controller (AOC) for titration of supplemental oxygen, rather than manual titration by bedside staff. Although the concept was already thought of in the 1940s, automated oxygen titration has only become booming in the last decade. A recent survey in the UK by Kaltsogianni et al. showed that 10% of the surveyed units have embraced AOC, with the majority of those units using it in routine clinical care. In the last two decades, over 20 studies reported a comparison of manual titration with an automated oxygen controller in preterm infants.14-20 Six devices are commercially available at the time of writing, which are discussed in chapter 2. Automated oxygen titration by a device reduces time outside of the target range14, 21 by providing a more prompt response to deviations from the target range, and as such may help reduce associated morbidity and mortality. Several basic approaches are employed and combined by the available automated oxygen titration algorithms: