A method for measuring passive elastance during proportional assist ventilation

There are currently no reliable, noninvasive ways to monitor respiratory elastance (E) during assisted ventilation. We describe a method that is suited for proportional assist ventilation (PAV). In this mode, the end of the ventilator's inflation phase occurs during the declining phase of inspiratory effort (Pmus). If the opening of the exhalation valve is delayed, airway pressure (Paw) should initially rise as Pmus continues its decline. When Pmus declines to zero, a Paw plateau should appear. Paw at this point should reflect passive recoil at the prevailing volume. A cohort of 74 ventilator-dependent patients, ventilated in the PAV mode, were studied. Brief end-inspiratory occlusions were applied at random intervals. The magnitude of early change in Paw during the occlusion was inversely related to level of assist (r = 0.7, p < 0.00001). At high assist (> 75%), Paw was nearly flat or declined slightly, indicating minimal residual Pmus at the onset of occlusion. At lower assist levels, Paw increased exponentially in most patients with an average time constant of 0.21 +/- 0.06 s. Extraneous events that may corrupt the measurement (e.g., behavioral responses) were extremely rare (< 0.5%) in the first 0.25 s. From these findings, we concluded that Paw measured 0.25 s from occlusion onset (P-0.25) includes little inspiratory Pmus and is free of extraneous events. E, estimated from P-0.25 during PAV (E-PAV), agreed well (r = 0.92) with passive E measured during controlled ventilation (ECMV); the average difference (EPAV - ECMV) was (+/- SD) -0.3 +/- 4.9 cm H2O <bullet> L-1, corresponding to 0.9 +/- 16.4% of average E. We conclude that Paw measured at 0.25 s from the onset of end-inspiratory occlusion in the PAV mode provides a reliable estimate of passive elastic recoil.