The control of breathing during weaning from mechanical ventilation

Using the recruitment threshold technique, we measured the CO2 responsiveness of the unloaded respiratory pump in 14 mechanically ventilated patients prior to weaning. The CO2 recruitment threshold (CO2RT) was compared with the arterial CO2 tension during unassisted breathing (CO2SB) and with the PaCO2 during mechanical ventilation (CO2MV) at machine settings determined by the primary physician. Based on these comparisons, we tested the hypotheses that (1) patients without weaning-induced respiratory distress (group 1) maintain CO2SB near CO2RT, (2) patients with weaning-induced respiratory distress (group 2) retain CO2SB above CO2RT, thereby manifesting incomplete load compensation, and (3) CO2MV is ventilator setting dependent and provides insufficient information about the ventilatory requirement during weaning. Respiratory distress was prospectively defined as sustained tachypnea (rate greater than or equal to 30) or intense dyspnea (Borg scale rating) and limited weaning in nine of 14 patients. The average CO2RT was 40 mm Hg in both groups. All patients in group 1 maintained CO2SB near CO2RT (p greater than 0.1). Seven of nine patients in group 2 retained CO2 by greater than or equal to 3 mm Hg above CO2RT (p less than 0.01). There was no significant difference between CO2MV and CO2SB in either group. We conclude that CO2RT provides a better reference of the adequacy of ventilatory load compensation during weather than CO2MV.