| Abstract | The vast majority (95%) of very preterm infants receive oxygen-therapy monitored by oxygen pulse saturation (SpO2 ). However, they spend a significant percentage of time out of the SpO2 target with a high risk of severe complications such as bronchopulmonary dysplasia (BPD) and retinopathy of prematurity (ROP). Recently, systems allowing the automated control of inspired oxygen (FiO2 ) for patient delivery to maintain target SpO2 has been become commercially available. We reviewed literature and individuated sixteen studies on the effectiveness of automated control of FiO2 in preterm infants. These studies demonstrate that automated devices are significantly more effective than manual control in maintaining target SpO2 and in preventing hyperoxia, while they seem to be less effective in preventing hypoxia. The studies were very heterogeneous for design, population size, duration, and device used, and this precludes firm conclusions regarding effectiveness and best setting of these systems. Moreover, none of the studies investigated if automated control of FiO2 can actually improve outcome in preterm infants. We conclude that further large-scale studies are warranted to assess the actual clinical relevance of these devices and to decide if they should become the standard of care. |