Physiological response to pressure support ventilation delivered before and after extubation in patients not capable of totally spontaneous autonomous breathing.

We designed a prospective, physiological study in 12 patients affected by chronic respiratory disorders. The study was aimed at assessing the diaphragm energy expenditure (PTPdi), lung resistance (RL) and elastance (EL), arterial blood gases (ABG), breathing pattern, and dyspnea measured by a visual analog scale during invasive pressure support ventilation (i-PSV) and noninvasive PSV (n-PSV). The ventilator settings were kept the same. Both i-PSV and n-PSV significantly reduced the PTPdi per minute, compared with that during a T-piece trial (204.4 +/- 93.8 cm H(2)O x s/min [i-PSV]; 197.5 +/- 119.8 [n-PSV]; 393.8 +/- 129.0 [T-piece]). Expired tidal volume (VTe) was significantly higher (p < 0.05) during n-PSV (615 +/- 166 ml) than during i-PSV (519 +/- 140 ml). The respiratory pump (PTPdi/VTe) was more effective (p < 0.05) with noninvasive ventilation (22.3 +/- 2.3 cm H(2)O x s/L for i-PSV versus 17.2 +/- 3.3 for n-PSV). RL and EL were similar with the two modes of ventilation. Overall dyspnea was significantly (p < 0.05) better during n-PSV than i-PSV, whereas ABG were similar. We have shown, in patients affected by stable chronic respiratory disorders not ready to sustain totally spontaneous breathing, that i-PSV and n-PSV are equally effective in reducing the PTPdi and in improving ABG, but that n-PSV seems to be better tolerated.