Comparison between neurally-assisted, controlled, and physiologically variable ventilation in healthy rabbits.

Abstract

Various ventilation strategies have been proposed to reduce ventilation-induced lung injury that occurs even in individuals with healthy lungs. We compared new modalities based on an individualised physiological variable ventilation model to a conventional pressure-controlled mode. Rabbits were anaesthetised and ventilated for up to 7 h using pressure-controlled ventilation with (Group PCS, n=10), and without (Group PC, n=10) regular sighs. Variable ventilation in the other two groups was achieved via a pre-recorded spontaneous breathing pattern [Group physiologically variable ventilation (PVV), n=10] or triggered by the electrical activity of the diaphragm [Group neurally adjusted ventilation assist (NAVA), n=9]. Respiratory elastance, haemodynamic profile, and gas exchange were assessed throughout the ventilation period. Cellular profile, cytokine content of bronchoalveolar lavage fluid, and wet-to-dry lung weight ratio (W/D) were determined after protocol completion. Lung injury scores were obtained from histological analysis. Marked deteriorations in elastance were observed (median and 95% confidence interval) in Group PC [48.6 (22)% increase from baseline], while no changes were detected in Groups PCS [3.6 (8.1)%], PVV [18.7 (13.2)%], and NAVA [-1.4 (12.2)%]. In comparison with Group PC, Group PVV had a lower lung injury score [0.29 (0.02) compared with 0.36 (0.05), P<0.05] and W/D ratio [5.6 (0.1) compared with 6.2 (0.3), P<0.05]. There was no difference in blood gas, haemodynamic, or inflammatory parameters between the groups. Individualised PVV based on a pre-recorded spontaneous breathing pattern provides adequate gas exchange and promotes a level of lung protection. This ventilation modality could be of benefit during prolonged anaesthesia, in which assisted ventilation is not possible because of the absence of a respiratory drive.