Title: Reduced Unloading in NAVA Improves distribution of Ventilation in ICU patients.

Objectives:

1. To investigate if NAVA targeted to moderate respiratory muscular unloading results in redistribution of ventilation to the dorsal regions of the lungs
2. To verify if the redistribution of ventilation translates to a better gas exchange and to a potentially lung protective ventilation strategy (lower airway pressures)
3. To verify the possibility to set NAVA at different levels of unloading, based on Neuro-Ventilatory Efficiency.

Study Design: Randomised Crossover of Pressure Support and NAVA at different levels of unloading.

Population: Adult Intubated patients at the Neurosurgical ICU, ventilated for more than 48h, in weaning phase from mechanical ventilation.

Study duration: 2,5h Number of subjects: 12

Critically ill patients on mechanical ventilation are at risk for developing respiratory muscle atrophy. Partial Assist modes such Pressure Support (PS) and Neurally Adjusted Ventilatory Assist (NAVA) are developed to maintain patients´own effort in breathing. However there are no recommendations on how to set the optimal ventilator support in NAVA to avoid over- or underassistance.

A previous Electrical Impedance Thomography (EIT) study has shown a redistribution of ventilation towards the dorsal regions of the lung in acute lung injury patients ventilated with NAVA, compared to PS.

In the present study, the assist is targeted to different respiratory muscle unloading, predefined and based on the Neuro-Ventilatory Efficiency (NVE). The NVE will be measured at 10min intervals and NAVA level adjusted if needed, to keep constant the level of unloading in each study step.

Protocol: Once enrolled, the patients are ventilated in PS (PScli1) as set by the clinician. They are then ventilated in NAVA at 3 different levels of muscle unloading in randomized order. At NAVAcli, the assist level matches to PScli1 in terms of muscle unloading. With NAVA40% and NAVA60%, the patients have 40% and 60% unloading, respectively. In the last study step the patients are back to PS (PScli2). Each patient is his/her own control and goes through the 5 ventilation periods, of 30min each. In the last 5 min of each study step, the CoV (obtained through the EIT data), blood gas samples (for oxygenation and ventilation) and ventilatory parameters are obtained and analyzed.

The investigators hypothesize that

1. It is possible to set NAVA at different levels of unloading, based on NVE.
2. Moderate muscle unloading (corresponding to NAVA40%) keeps the diaphragm active and thereby leads to more dorsal distribution of ventilation compared to PScli and to higher unloading in NAVA.
3. Secondarily and as a consequence of the redistribution of ventilation, we hypothesize that the gas exchange will remain unchanged or will improve and that the airway pressures will decrease for moderate unloading (NAVA40%).

Inclusion Criteria:

• Patients belonging to Neurosurgical ICU
• Intubated for ≥48h
• Weaning phase from Mechanical Ventilation

Exclusion Criteria:

• bleeding disorders (PK INR>1,5 or APTT>50s or platelet count <50000/µL) or
• unstable intracranial pressure (ICP>20 mmHg during the latest 8 hours) or
• unstable circulation (requiring high vasopressor dose, for example Noradrenalin >0,2µg/kg/min) or
• too severe lung disease (PFI ≤ 26,7 kPa or PEEP >10 cmH2O or FiO2>0,5 at study entry point) or
• fever> 38,5°C or
• tendency to hyperventilation (PaCO2 < 4,5 kPa at study entry point).