The Flo, was adjusted with an oxygen blender (Bird 960, Palm Springs, CA) to maintain an arterial oxygen saturation between 90 percent and 93 percent. The ventilator assist rate was then increased slowly until spontaneous respiratory activity ceased, as assessed from inspection of the Pao and flow tracings. Absence of phasic respiratory muscle activity was assumed when mechanical inflation with a constant flow caused a step-like increase in airway pressure followed by a linear rise to a peak value and a quasiexponential decline of expiratory flow. After 15 minutes of mechanical ventilation at these settings, the inspired CO, concentration (FIco,) was raised by 1 percent to 3 percent in stepwise increments of 3 to 5 min until spontaneous inspiratory muscle activity reappeared. Respiratory muscle recruitment was defined by the absence of relaxation criteria on ten or more consecutive breaths (Fig 1). An arterial blood gas sample was obtained within 30 s of recruitment, the CO, tension of which defined CO,RT. In seven patients, the CO,RT measurement was repeated immediately following the weaning trial.
COtMV: The arterial CO, tension during mechanical ventilation on the day of the study but prior to experimental measurements was recorded. The mode and ventilator settings were determined entirely by the patients primary physician, who had no knowledge of the specific aims of the study.
Figure 1. Tracings of the fractional CO, concentration (FcoJ and of the pressure (Pao) measured at the airway opening along with flow (V) through the endotracheal tube are shown for a patient (patient 4) receiving mechanical ventilation. Pressure and flow waveforms on the left-hand side of the figure (panel A) indicate absence of phasic respiratory motor activity. CO, supplementation to the inspired gas (panel B) induced phasic respiratory motor activity, as evident from the change in pressure and flow tracings.