Prediction of post-extubation work of breathing

Objective: To evaluate which mode of preextubation ventilatory support most closely approximates the work of breathing performed by spontaneously breathing patients after extubation. Design: Prospective observational design. Setting: Medical, surgical, and coronary intensive care units in a university hospital. Patients: A total of 22 intubated subjects were recruited when weaned and ready for extubation. Interventions. Subjects were ventilated with continuous positive airway pressure at 5 cm H2O, spontaneous ventilation through an endotracheal tube (T piece), and pressure support ventilation at 5 cm H2O in randomized order for 15 mins each. At the end of each interval, we measured pulmonary mechanics including work of breathing reported as work per liter of ventilation, respiratory rate, tidal volume, negative change in esophageal pressure, pressure time product, and the airway occlusion pressure 100 msec after the onset of inspiratory flow, by using a microprocessor-based monitor. Subsequently, subjects were extubated, and measurements of pulmonary mechanics were repeated 15 and 60 mins after extubation. Measurements and Main Results: There were no statistical differences between work per liter of ventilation measured during continuous positive airway pressure, T piece, or pressure support ventilation (1.17 +/- 0.67 joule/l, 1.11 +/- 0.57 joule/l, and 0.97 +/- 0.57 joule/l, respectively). However, work per liter of ventilation during all three preextubation modes was significantly lower than work measured 15 and 60 mins after extubation (p < .05). Tidal volume during pressure support ventilation and continuous positive airway pressure (0.46 +/- 0.11 L and 0.44 +/- 0.11 L, respectively) were significantly greater than tidal volume during both T-piece breathing and spontaneous breathing 15 mins after extubation (p < .05). Negative change in esophageal pressure, the airway occlusion pressure 100 msec after the onset of inspiratory flow, and pressure time product were significantly higher after extubation than during any of the three preextubation modes (p < .05). Conclusions: Work per liter of ventilation, negative change in esophageal pressure, the airway occlusion pressure 100 msec after the onset of inspiratory flow, and pressure time product all significantly increase postextubation. Tidal volume during continuous positive airway pressure or pressure support ventilation overestimates postextubation tidal volume.