Automated control of inspired oxygen in ventilated preterm infants: crossover physiological study

To evaluate the efficacy of automated control of the fraction of inspired oxygen (FiO2 ) in comparison with manual FiO2 control in maintaining target pulse oxygen saturation (SpO2 ) range.

Crossover physiological study involving preterm infants requiring mechanical ventilation and supplemental oxygen. Each infant was studied for two consecutive 12 hours in a random sequence. Outcome measures included the proportion of time with SpO2 within and outside the target range of 90-95%, extreme hypoxaemia (< 80%) and hyperoxaemia (≥ 98%).

Complete data set was available in 27 infants. The percentage of time (median, IQR) within the target range was higher during automated control 72.8 (58.8-82.6) compared to manual control 59.6 (49.3-73.3), p = 0.031. Corresponding reduction in per cent time below the target range was 18.1 (12.7-23.6) versus 25.9 (17.8-30.7), p = 0.028, and above the target range 4.8 (3-16) versus 10.1 (6.4-22.5), p = 0.026. Median (IQR) per cent time spent with severe hypoxaemia (SpO2 < 80%) and severe hyperoxaemia (SpO2 ≥ 98%) was 1.3 (0.1-2.9) versus 3.2 (1.4-6.1), p = 0.022 and 0.08 (0.05-0.36) versus 1.7 (0.7-6.8), p = 0.001, respectively. Median number of manual adjustments of FiO2 per 12 hour was 0 and 63, respectively.

Automated control of FiO2 significantly improved compliance of oxygen saturation targeting and significantly reduced exposure to hypoxaemia as well as hyperoxaemia.