Neurally adjusted ventilatory assist improves patient-ventilator interaction during postextubation prophylactic noninvasive ventilation.

Objectives: To compare the respective impact of pressure support ventilation and naturally adjusted ventilatory assist, with and without a noninvasive mechanical ventilation algorithm, on patient-ventilator interaction.
Design: Prospective 2-month study.
Setting: Adult critical care unit in a tertiary university hospital.
Patients: Seventeen patients receiving a prophylactic postextubation noninvasive mechanical ventilation.
Interventions: Patients were randomly mechanically ventilated for 10 mins with: pressure support ventilation without a noninvasive mechanical ventilation algorithm (PSV-NIV-), pressure support ventilation with a noninvasive mechanical ventilation algorithm (PSV-NIV+), neurally adjusted ventilatory assist without a noninvasive mechanical ventilation algorithm (NAVA-NIV-), and neurally adjusted ventilatory assist with a noninvasive mechanical ventilation algorithm (NAVA-NIV+).
Measurements and Main Results: Breathing pattern descriptors, diaphragm electrical activity, leak volume, inspiratory trigger delay, inspiratory time in excess, and the five main asynchronies were quantified. Asynchrony index and asynchrony index influenced by leaks were computed. Peak inspiratory pressure and diaphragm electrical activity were similar for each of the four experimental conditions. For both pressure support ventilation and neurally adjusted ventilatory assist, the noninvasive mechanical ventilation algorithm significantly reduced the level of leakage (p < .01). Inspiratory trigger delay was not affected by the noninvasive mechanical ventilation algorithm but was shorter in neurally adjusted ventilatory assist than in pressure support ventilation (p < .01). Inspiratory time in excess was shorter in neurally adjusted ventilatory assist and PSV-NIV+ than in PSV-NIV- (p < .05). Asynchrony index was not affected by the noninvasive mechanical ventilation algorithm but was significantly lower in neurally adjusted ventilatory assist than in pressure support ventilation (p < .05). Asynchrony index influenced by leaks was insignificant with neurally adjusted ventilatory assist and significantly lower than in pressure support ventilation (p < .05). There was more double triggering with neurally adjusted ventilatory assist.
Conclusions: Both neurally adjusted ventilatory assist and a noninvasive mechanical ventilation algorithm improve patient-ventilator synchrony in different manners. NAVA-NIV+ offers the best compromise between a good patient-ventilator synchrony and a low level of leaks. Clinical studies are required to assess the potential clinical benefit of neurally adjusted ventilatory assist in patients receiving noninvasive mechanical ventilation.
Trial Registration: Clinicaltrials.gov Identifier NCT01280760.