Hylke Salverda

176 Chapter 10 implementation of automated oxygen control in our NICU as standard of care for preterm infants led to no statistically significant difference in neurodevelopmental outcome at two years of age. In chapter 8 we compared short-term clinical outcomes after using two different automated oxygen controllers (OxyGenie and CLiO2) in a propensity score matched retrospective observational study. Preterm infants (OxyGenie n=121, CLiO2 n=121) born between 24+0-29+6 weeks of gestation were included. Median [IQR] gestational age in the OxyGenie cohort was 28+3 [26+3.5–29+0] versus 27+5 [26+5–28+3] in the CLiO2 cohort, respectively 42% and 46% of infants were male and mean (SD) birth weight was 1034 (266) grams vs 1022 (242) grams. Again, inspired oxygen was titrated by OxyGenie (SLE6000) or CLiO2 (AVEA) during respiratory support. We compared mortality, retinopathy of prematurity, bronchopulmonary dysplasia, and necrotising enterocolitis and found that fewer infants in the OxyGenie group received laser coagulation for ROP (1 infant vs 10; risk ratio 0.1 (95%-CI 0.0 – 0.7); p=0.008), and infants had a shorter admittance in the NICU (28 [15-42] vs 40 [2561] days; median difference 13.5 days (95%-CI 8.5 – 19.5); p<0.001). Infants in the OxyGenie group had fewer days on continuous positive airway pressure (8.4 [4.819.8] days vs. 16.7 [6.3-31.1]; p<0.001) and a significantly lower number of days on invasive ventilation (0 [0-4.2] days vs. 2.1 [0-8.4]; p=0.012). There were no statistically significant differences between all other morbidities. In this propensity scorematched retrospective study, the OxyGenie epoch was associated with less morbidity when compared to the CLiO2 epoch. There were significantly fewer infants that received treatment for ROP, received less intensive respiratory support and, although there were more supplemental oxygen days, the duration of stay in the NICU was shorter. A larger study will have to replicate these findings. Finally, this thesis concludes with discussing the results of our studies. Automated oxygen control is up and coming and improves oxygen targeting by reducing hypoxia, hyperoxia and workload. This thesis provides the first evidence that the success of oxygen targeting is influenced by choice of AOC device. The OxyGenie controller was more effective than the CLiO2 controller, both in a controller crossover study as well as during the entire admission under routine clinical care. Ultimately our result demonstrated that OxyGenie is a better choice than CLiO2 for oxygenation targeting. The effect on clinical outcome of these devices is not yet clear. OxyGenie control was associated with better clinical outcome than CLiO2 control, but the retrospective nature of the studies performed precludes us from inferring causality. Nevertheless, all data in this thesis are pointing in the same direction.