Noninvasive proportional assist ventilation compared with noninvasive pressure support ventilation in hypercapnic acute respiratory failure

Abstract

OBJECTIVES: To compare short-term administration of noninvasive proportional assist ventilation (NIV-PAV) and pressure support ventilation (NIV-PSV). DESIGN: Prospective, crossover, randomized study. SETTING: Medicosurgical intensive care unit in a nonteaching hospital. PATIENTS: Twelve chronic obstructive pulmonary disease patients admitted for hypercapnic acute respiratory failure. INTERVENTION: NIV-PSV and NIV-PAV given in a randomized order after baseline evaluation in continuous positive airway pressure. Using a flow-triggering ventilator, NIV-PAV was adjusted using the runaway method and compared with NIV-PSV at similar peak inspiratory airway pressure. MEASUREMENTS AND MAIN RESULTS: Flow, airway pressure, and changes in esophageal pressure were measured and the tidal volume, the patient's inspiratory work of breathing, and the esophageal pressure--time product were calculated. Arterial pH and PaCO(2) were measured and breathing comfort was assessed using a visual analogic scale. Peak inspiratory airway pressure (17 +/- 3 cm H(2)O) and tidal volume were similarly increased with the two modalities with no change in respiratory rate. The change in esophageal pressure was similarly decreased (from 20 +/- 8 cm H(2)O in continuous positive airway pressure to 12 +/- 7 in NIV-PSV and 10 +/- 5 cm H(2)O in NIV-PAV) as well as inspiratory muscle effort indexes. Arterial pH and PaCO(2) were similarly improved. Breathing comfort was significantly improved in NIV-PAV (+38 +/- 38%) but not in NIV-PSV (+11 +/- 23%). The tidal volume was more variable in NIV-PAV (89 +/- 18%) than in NIV-PSV (15 +/- 8%) and changes in tidal volume variability were significantly correlated (p =.02) with changes in breathing comfort. CONCLUSIONS: In chronic obstructive pulmonary disease patients with hypercapnic acute respiratory failure, NIV-PAV was able to unload inspiratory muscles similarly to NIV-PSV but may be more comfortable than NIV-PSV.