A Comprehensive Bench Assessment of Automatic Tube Compensation in ICU Ventilators for Better Clinical Management

BACKGROUND: Automatic tube compensation (ATC) unloads endotracheal tube (ETT) resistance. We conducted a bench assessment of ATC functionality in ICU ventilators to improve clinical management. METHODS: This study had 2 phases. First, we performed an international survey on the use of ATC in clinical practice, hypothesizing a rate of ATC use of 25%. Second, we tested 7 modern ICU ventilators in a lung model mimicking a normal subject (Normal), a subject with ARDS, and a subject with COPD. Inspiratory effort consisted of esophageal pressure over 30 consecutive breaths obtained in a real patient under weaning. A brand new 8-mm inner diameter ETT was attached to the lung model, and ATC was set at 100% compensation for the ETT. The 30 breaths were first run with ATC off and no ETT (ie, reference period), and then with ATC on and ETT (ie, active period). The primary end point was the difference in tidal volume (V(T)) between reference and active periods. We hypothesized that the V(T) difference should be equal to 0 in an ideally functioning ATC. V(T) difference was compared across ventilators and respiratory mechanics conditions using a linear mixed-effects model. RESULTS: The clinical use of ATC was 64% according to 644 individuals who responded to the international survey. The V(T) difference varied significantly across ventilators in all respiratory mechanics configurations. The divergence between V(T) difference and 0 was small but significant: the extreme median (interquartile range) values were -0.013 L (-0.019 to -0.002) in the COPD model and 0.056 L (0.051-0.06) in the Normal model. V(T) difference for all ventilators was 0.015 L (95% CI 0.013-0.018) in the ARDS model, which was significantly different from 0.021 L (95% CI 0.018-0.024) in the Normal model (P \textless .001) and 0.010 L (0.007-0.012) in the COPD model (P = .003). CONCLUSIONS: ATC is used more frequently in clinical practice than expected. In addition, V(T) delivery by ATC differed slightly though significantly between ventilators