Objective: To compare the effects of pressure support ventilation (PSV) delivered at the same level by three different ventilators on patients' work of breathing (WOB), breathing pattern and gas exchange.
Design: Prospective, self-controlled clinical study.
Setting: Intensive care unit of a tertiary university hospital.
Patients: Nine intubated adult patients during weaning from mechanical ventilation.
Interventions: Patients were randomly connected to one of three ventilators: the Siemens Servo 900 C (SC), the Ohmeda CPU 1 (CPU), and the Engström Erica (EE) during both zero cmH2O PSV and 15 cmH2O PSV.
Measurements and results: During zero PSV, there was no significant difference in terms of WOB, VT, VE, or auto-PEEP among the three ventilators, although there was a trend towards higher levels of WOB with EE. During 15 cmH2O PSV, WOB was significantly less with SC than with EE or CPU (0.47 +/- 0.48 J/l for SC, 1.0 +/- 0.48 for EE and 0.78 +/- 0.51 for CPU1, p = 0.003). WOB was 64% less than at zero PSV with SC but only 38% less with EE. This was associated with a different pressurization shape, as assessed by the interior surface of Paw-VT loops (1.23 +/- 0.09 J/l for SC, 0.9 +/- 0.02 for EE, and 0.79 +/- 0.18 for CPU; p < 0.001). At 15 cmH2O PSV, auto-PEEP was significantly lower with SC than with EE (1.7 +/- 2.1 cmH2O for SC, 4.7 +/- 3.6 for EE, and 2.8 +/- 0.3 for CPU; p = 0.04). External expiratory resistances, in cmH2O/l/s, were significantly higher with EE than with CPU or SC (12.9 +/- 3.2 EE, 7.5 +/- 2.4 CPU, 5.9 +/- 0.5 SC; p < 0.001).
Conclusion: During PSV, the different working principles of different mechanical ventilators profoundly affect patient's WOB. Among the various factors, velocity of pressurization of PSV may play a role in its efficacy in unloading the respiratory muscles.