Patient-ventilator interaction during pressure support ventilation and neurally adjusted ventilatory assist

Objective: To compare the effect of pressure support ventilation and neurally adjusted ventilatory assist on breathing pattern, patient-ventilator synchrony, diaphragm unloading, and gas exchange. Increasing the level of pressure support ventilation can increase tidal volume, reduce respiratory rate, and lead to delayed ventilator triggering and cycling. Neurally adjusted ventilatory assist uses diaphragm electrical activity to control the timing and pressure of assist delivery and is expected to enhance patient-ventilator synchrony. Design: Prospective, comparative, crossover study. Setting: Adult critical care unit in a tertiary university hospital. Patients: Fourteen nonsedated mechanically ventilated patients (n = 12 with chronic obstructive pulmonary disease). Interventions: Patients were ventilated for 10-min periods, using two pressure support ventilation levels (lowest tolerable and +7 cm H2O higher) and two neurally adjusted ventilatory assist levels (same peak pressures and external positive end-expiratory pressure as with pressure support ventilation), delivered in a randomized order. Measurements and Main Results: Diaphragm electrical activity, respiratory pressures, air flow, volume, neural and ventilator respiratory rates, and arterial blood gases were measured. Peak pressures were 17 +/- 6 cm H2O and 24 +/- 6 cm H2O and 19 +/- 5 cm H2O and 24 +/- 6 cm H2O with high and low pressure support ventilation and neurally adjusted ventilatory assist, respectively. The breathing pattern was comparable with pressure support ventilation and neurally adjusted ventilatory assist during low assist; during higher assist, larger tidal volumes (p = .003) and lower breathing frequencies (p = .008) were observed with pressure support ventilation. Increasing the assist increased cycling delays only with pressure support ventilation (p = .003). Compared with pressure support ventilation, neurally adjusted ventilatory assist reduced delays of ventilator triggering (p < .001 for low and high assist) and cycling (high assist: p = .004; low assist: p = .04), and abolished wasted inspiratory efforts observed with pressure support ventilation in six subjects. The diaphragm electrical activity and pressure-time product for ventilator triggering were lower with neurally adjusted ventilatory assist (p = .005 and p = .02, respectively; analysis of variance). Arterial blood gases were similar with both modes. Conclusions: Neurally adjusted ventilatory assist can improve patient-ventilator synchrony by reducing the triggering and cycling delays, especially at higher levels of assist, at the same time preserving breathing and maintaining blood gases. (Crit Care Med 2010; 38:518-526)