Abstract
ObjectiveWe analysed the relationship between oscillatory volume (VOSC) and pressure amplitude (ΔP) in six neonatal high-frequency oscillatory (HFO) ventilators and related it to (1) the accuracy of VOSCand ΔP measurements and (2) the maximal delivered ΔP.DesignIn vitro study.SettingNeonatal intensive care unit.InterventionsVentilators tested were VN800 (Dräger), Servo-n (Maquet Getinge), SensorMedics 3100A (Vyaire Medical), Fabian HFOi (Vyaire Medical), SLE6000 (SLE UK) and Humming Vue (Metran). We changed various settings and mechanical characteristics of the test lung to mimic preterm and term conditions.Main outcome measuresFor each condition, we measured VOSCand ΔP. We assessed the accuracy of the VOSCand ΔP measurements versus a reference measurement system using linear regression and Bland-Altman analysis. We evaluated the maximum delivered ΔP at different oscillatory frequencies.ResultsWe observed large variability between machines in the ΔP displayed at any target VOSC. Most ventilators over-read ΔP with errors up to 30 cmH2O or 60%. The error in the measurement of VOSCwas up to ±2 mL or ±30%. We observed high variability in the accuracy of ΔP and VOSCmeasurements; the SLE6000 committed the lowest errors in ΔP measurements and the Fabian HFOi in VOSC. The maximum delivered ΔP varied depending on the ventilator, being maximal for the Humming Vue, followed by the SLE6000 and SensorMedics 3100A.ConclusionsThe variability in the relationship between VOSCand ΔP among HFO ventilators is largely explained by the variable accuracy in ΔP and VOSCmeasurement. Different ventilators also exhibit important differences in the maximal generated ΔP.