Introduction: The stress index (SI), a parameter derived from the shape of the pressure-time curve, can identify injurious mechanical ventilation. We tested the hypothesis that adjusting tidal volume (VT) to a non-injurious SI in an open lung condition avoids hypoventilation while preventing overdistension in an experimental model of combined lung injury and low chest-wall compliance (Ccw).
Methods: Lung injury was induced by repeated lung lavages using warm saline solution, and Ccw was reduced by controlled intra-abdominal air-insufflation in 22 anesthetized, paralyzed and mechanically ventilated pigs. After injury animals were recruited and submitted to a positive end-expiratory pressure (PEEP) titration trial to find the PEEP level resulting in maximum compliance. During a subsequent four hours of mechanical ventilation, VT was adjusted to keep a plateau pressure (Pplat) of 30 cmH2O (Pplat-group, n = 11) or to a SI between 0.95 and 1.05 (SI-group, n = 11). Respiratory rate was adjusted to maintain a 'normal' PaCO2 (35 to 65 mmHg). SI, lung mechanics, arterial-blood gases haemodynamics pro-inflammatory cytokines and histopathology were analyzed. In addition Computed Tomography (CT) data were acquired at end expiration and end inspiration in six animals.
Results: PaCO2 was significantly higher in the Pplat-group (82 versus 53 mmHg, P = 0.01), with a resulting lower pH (7.19 versus 7.34, P = 0.01). We observed significant differences in VT (7.3 versus 5.4 mlKg(-1), P = 0.002) and Pplat values (30 versus 35 cmH2O, P = 0.001) between the Pplat-group and SI-group respectively. SI (1.03 versus 0.99, P = 0.42) and end-inspiratory transpulmonary pressure (PTP) (17 versus 18 cmH2O, P = 0.42) were similar in the Pplat- and SI-groups respectively, without differences in overinflated lung areas at end- inspiration in both groups. Cytokines and histopathology showed no differences.
Conclusions: Setting tidal volume to a non-injurious stress index in an open lung condition improves alveolar ventilation and prevents overdistension without increasing lung injury. This is in comparison with limited Pplat protective ventilation in a model of lung injury with low chest-wall compliance.