Automated oxygen titration with non-invasive ventilation in hypoxaemic adults with cardiorespiratory disease: a randomised cross-over trial

Abstract

BackgroundClosed-loop oxygen control systems automatically adjust the fraction of inspired oxygen (FiO2) to maintain oxygen saturation (SpO2) within a predetermined target range. Their performance with low and high-flow oxygen therapies, but not with non-invasive ventilation, has been established. We compared the effect of automated oxygen on achieving and maintaining a target SpO2range with nasal high flow (NHF), bilevel positive airway pressure (bilevel) and continuous positive airway pressure (CPAP), in stable hypoxaemic patients with chronic cardiorespiratory disease.MethodsIn this open-label, three-way cross-over trial, participants with resting hypoxaemia (n=12) received each of NHF, bilevel and CPAP treatments, in random order, with automated oxygen titrated for 10 min, followed by 36 min of standardised manual oxygen adjustments. The primary outcome was the time taken to reach target SpO2range (92%–96%). Secondary outcomes included time spent within target range and physiological responses to automated and manual oxygen adjustments.ResultsTwo participants were randomised to each of six possible treatment orders. During automated oxygen control (n=12), the mean (±SD) time to reach target range was 114.8 (±87.9), 56.6 (±47.7) and 67.3 (±61) seconds for NHF, bilevel and CPAP, respectively, mean difference 58.3 (95% CI 25.0 to 91.5; p=0.002) and 47.5 (95% CI 14.3 to 80.7; p=0.007) seconds for bilevel and CPAP versus NHF, respectively. Proportions of time spent within target range were 68.5% (±16.3), 65.6% (±28.7) and 74.7% (±22.6) for NHF, bilevel and CPAP, respectively.Manually increasing, then decreasing, the FiO2resulted in similar increases and then decreases in SpO2and transcutaneous carbon dioxide (PtCO2) with NHF, bilevel and CPAP.ConclusionThe target SpO2range was achieved more quickly when automated oxygen control was initiated with bilevel and CPAP compared with NHF while time spent within the range across the three therapies was similar. Manually changing the FiO2had similar effects on SpO2and PtCO2across each of the three therapies.Trial registration numberACTRN12622000433707.