Neurally adjusted ventilatory assist : from animal studies to clinical practice

Abstract

Patients in the Intensive Care Unit (ICU) undergoing ventilator treatment may experience asynchrony with the ventilator, which has been associated with increased need of sedation, sleep disruption, prolonged mechanical ventilation and unsuccessful weaning from the ventilator. The search for new strategies to improve patient-ventilator interaction is ongoing. Neurally Adjusted Ventilatory Assist (NAVA) is a recently developed ventilator support that uses the Electrical Activity of the diaphragm (EAdi) as an input signal to control the ventilator. Each breath is delivered in proportion to the EAdi amplitude and follows the timing of the EAdi start and ending. NAVA may potentially improve patient-ventilator synchrony, which could be beneficial, especially in paediatric ICU patients. Further, NAVA use could be of interest during surgery and general anaesthesia, potentially reducing risk for intraoperative atelectasis formation and postoperative complications. However, the feasibility of NAVA during general anaesthesia has not been investigated. There is some uncertainty among clinicians on how to set NAVA bedside. A pragmatic strategy, targeting NAVA to different levels of muscle unloading has not been studied, but could be a means of standardizing the approach to unload. Finally, it is not known if lower unloading could improve the distribution of ventilation in favour of the dorsal regions of the lungs.

Patient-ventilator synchrony and diaphragm unloading with NAVA compared to Pressure Support (PS) were evaluated in Acute Lung Injured (ALI) rabbits with increasing level of assist and in rats breathing with an added dead space in the respiratory circuit. The feasibility and efficacy of NAVA was investigated in small species, close in weight to the smallest viable human being. The pattern of breathing in NAVA and PS was studied in small species, when dead space was added. The feasibility of NAVA in sedation and anaesthesia with sevoflurane and propofol and the impact of these drugs on the pattern of breathing and muscle contractility were investigated in a big animal model. The possibility to target NAVA to different levels of respiratory muscle unloading was evaluated in Neurosurgical Intensive Care (NICU) patients and the effect of moderate unloading on the distribution of ventilation was investigated with Electrical Impedance Tomography.

Synchrony with NAVA was maintained and diaphragm unloading increased when raising the assist level, while for high PS the synchrony and unloading worsened, due to wasted inspiratory efforts causing a larger work of breathing. With NAVA, oxygenation and ventilation remained in the physiologic range in small species and when dead space was added, similar PaCO2 was achieved with a lower increase in respiratory rate and minute ventilation, compared to PS. Sedation and anaesthesia with sevoflurane and propofol, during NAVA in pigs, preserved the EAdi signal and spontaneous breathing, keeping the gas exchange in the physiologic range. The tidal volume variability in NAVA was preserved both with sevoflurane and propofol, being higher in propofol, due to more frequent sighs, followed by post-sigh apnoea. With sevoflurane Neuro-mechanical and Neuro-Ventilatory Efficiency (NVE) were higher, suggesting that sevoflurane could better preserve muscle contractility, compared to propofol. In NAVA it was feasible to set the assist at different levels of respiratory muscle unloading in NICU patients, by means of the NVE. Lower muscle unloading was shown to redistribute ventilation towards the dorsal regions of the lungs.

In conclusion this thesis demonstrates that NAVA improves patient-ventilator synchrony for increased assist levels and unloads the respiratory muscles with lower pressures and volumes compared to PS. NAVA is feasible and efficacious in small species, close in weight to the lowest viable human being and appears to be more efficient than PS in eliminating the CO2. NAVA is feasible during sedation and anaesthesia with sevoflurane and propofol in pigs and preserves the natural variability in pattern of breathing. Propofol in combination with NAVA is associated with more sighs and post-sigh apnoea than sevoflurane. Sevoflurane appears to preserve muscle contractility in NAVA better than propofol. NAVA can be targeted to different levels of muscle unloading by means of the NVE and lower unloading redistributes ventilation towards the dorsal regions of the lungs, providing the premise for a better matching of ventilation and perfusion.