Impact of PAV Versus NAVA on Patient-Ventilator Synchrony and Respiratory Muscle Unloading

New, "intelligent" ventilator modes with more complex closed loops have been developed, some with a demonstrated clinical benefit. The modes of proportional assist ventilation (PAV) and neurally-adjusted ventilatory assist (NAVA) are capable of delivering ventilation proportional to patient effort and may be associated with improved patient-ventilator synchrony when compared to pressure support ventilation (PSV), a classic simple closed-loop assisted ventilation mode. This study will study both modes of ventilation (PAV and NAVA) in each patient who is recruited, and measures of patient-ventilator synchrony and work of breathing will be taken during each mode. Baseline data will be collected on a standardized volume-cycled control mode ventilator setting for 5 minutes and on a standardized pressure support ventilator mode for 25 minutes. Subjects will then be randomized to either PAV or NAVA ventilation and will be ventilated on that mode for 30 minutes. Equivalence of support levels between PAV and NAVA trials will be ensured by targeting the same peak (Ppeak) airway pressures. Data collected will include (but are not limited to) vital signs, sedation score, dyspnea assessment using visual analog scale, respiratory rate, tidal volume (Vt), peak airway pressure (Ppeak), inspiratory time and neural inspiratory time (Ti and Tni), total physiologic and neural respiratory cycle time (Ttot and Tntot), end-tidal CO2, esophageal pressure waveforms, waveforms of all ventilatory patterns, Edi waveforms, and peak Edi. Subjects will then be switched to the other mode of ventilation and undergo an identical 30-minute evaluation period with identical data collection. Arterial blood gas measurements will be done after any changes in ventilator settings only in subjects who have had arterial lines inserted for clinically-indicated reasons. This will be a replicate crossover study, meaning that all subjects that are initially randomized to one sequence (PAV-NAVA or NAVA-PAV) will afterwards be "crossed-over" to the other one. These two additional periods of measurements will allow us to account for potential carry over effects of the different interventions.