Selecting ventilator settings according to variables derived from the quasi-static pressure/volume relationship in patients with acute lung injury.

Knowledge of the pressure/volume (P/V) relationship of the lung may allow selection of tidal volume and positive end-expiratory pressure (PEEP) to optimize gas exchange without adversely affecting lung function or hemodynamics. Ten patients with acute lung injury were stabilized on controlled mechanical ventilation, based on conventional practice, using criteria from arterial blood gas data. The P/V relationship was determined under quasi-static conditions (end-expiratory and end-inspiratory, no flow periods > 0.8 s) during mechanical ventilation with an automated procedure that changed PEEP in a stepwise fashion. Differences in expiratory tidal volumes before and after a change in PEEP equaled the change in functional residual capacity (delta FRC). PEEP was set above the lowest point of the steepest section of the P/V curve (inflection pressure) to prevent end-expiratory lung collapse. Inspiratory tidal volumes (VTI) were adjusted to avoid an end-inspiratory lung volume reaching the flat part of the P/V curve. Averaged delta FRC versus PEEP curves were shifted to the left and the slope increased 1, 6, and 12 h after changing ventilator settings compared to baseline (P < 0.01). Averaged baseline delta FRC versus PEEP curves showed a marked inflection pressure that decreased after adjusting ventilator settings (P < 0.01). PEEP was increased from 7.4 +/- 1.8 cm H2O (baseline) to 11.9 +/- 1.6 cm H2O (1 h) (P < 0.001) according to measured baseline inflection pressures. Simultaneously, VTI had to be reduced from 759 +/- 161 mL (baseline) to 664 +/- 101 mL (1 h) (P < 0.01) to avoid end-inspiratory overinflation. To maintain minute volume constant ventilator frequency was increased from 14 +/- 1.2 (baseline) to 16 +/- 1.2 breaths/min (1 h) (P < 0.01). Maximum quasi-static compliance of 38 +/- 7 mL/cm H2O (baseline) increased to 46 +/- 9 mL/cm H2O (1 h) (P < 0.01). Maintaining FIO2 constant, PaO2 increased from a baseline of 90 +/- 16 mm Hg to 122 +/- 24 mm Hg (1 h) (P < 0.001), to 130 +/- 20 mm Hg (6 h) (P < 0.01), and to 138 +/- 19 mm Hg (12 h) (P < 0.01). Intrapulmonary shunt decreased from 0.28 +/- 0.08 (baseline) to 0.14 +/- 0.05 (12 h) (P < 0.001). Hemodynamic variables did not change. Our data suggest that using variables derived from a quasi-static P/V loop during mechanical ventilation under muscle paralysis is clinically superior compared to blood gas criteria for titration of ventilator settings.