RESUMO
BACKGROUND: Noisy pressure support ventilation has been reported to improve respiratory function compared to conventional assisted mechanical ventilation. We aimed at determining the optimal level of pressure support variability during noisy pressure support ventilation. METHODS: Twelve pigs were anesthetized and mechanically ventilated. Acute lung injury was induced by surfactant depletion. At four levels of pressure support variability (coefficients of variation of pressure support equal to 7.5, 15, 30, and 45%, 30 min each, crossover design, special Latin squares sequence), we measured respiratory variables, gas exchange, hemodynamics, inspiratory effort, and comfort of breathing. The mean level of tidal volume was constant among variability levels. RESULTS: Compared to conventional pressure support ventilation, different levels of variability in pressure support improved the elastance of the respiratory system, peak airway pressure, oxygenation, and intrapulmonary shunt. Oxygenation and venous admixture benefited more from intermediate (30%) levels of variability, whereas elastance and peak airway pressure improved linearly with increasing variability. Heart rate as well as mean arterial and pulmonary arterial pressures decreased slightly at intermediate to high (30-45%) levels of variability in pressure support. Inspiratory effort and comfort of breathing were not importantly influenced by increased variability in pressure support. CONCLUSION: In a surfactant depletion model of acute lung injury, variability of pressure support improves lung function. The variability level of 30% seems to represent a reasonable compromise to improve lung functional variables during noisy pressure support ventilation.