Proportional assist versus pressure support ventilation in patients with acute respiratory failure: cardiorespiratory responses to artificially increased ventilatory demand.

OBJECTIVE: To test the hypothesis that in response to increased ventilatory demand, dynamic inspiratory pressure assistance better compensates for increased workload compared with static pressure support ventilation (PSV).

DESIGN: Randomized clinical crossover study.

SETTING: General intensive care u nits of a university hospital.

PATIENTS: Twelve patients with acute respiratory failure.

INTERVENTIONS: Patients received PSV, proportional assist ventilation (PAV), and PAV+ automatic tube compensation (ATC) in random order while maintaining mean inspiratory airway pressure constant. During each setting, ventilatory demand was increased by adding deadspace without ventilator readjustment.

MEASUREMENTS AND MAIN RESULTS: Cardiorespiratory, ventilatory, and work of breathing variables were assessed by routine monitoring plus pneumotachography; airway, esophageal, and abdominal pressure measurements; and nitrogen washout. After deadspace addition, tidal volume and end-expiratory lung volume increased similarly in all ventilatory modalities. Ventilator work, peak inspiratory flow, and maximum airway pressure increased significantly during PAV+ATC when compared with PSV after deadspace addition. However, increase in ventilator work did not result in a smaller increase in patients' work of breathing with elevated ventilatory demand during PAV+ATC (PSV 807 +/- 204 mJ/L, PAV 802 +/- 193 mJ/L, and PAV+ATC 715 +/- 202 mJ/L, p = .11). Increase in patients' work of breathing was mainly caused by a significantly higher resistive workload during PAV and PAV+ATC.

CONCLUSION: In patients with acute respiratory failure, dynamic inspiratory pressure assistance modalities are not superior to PSV with respect to cardiorespiratory function and inspiratory muscles unloading after increasing ventilatory demand. The latter might be explained by higher peak flows resulting in nonlinearly increased resistive workload that was incompletely compensated by PAV+ATC.