Noninvasive detection of positive end-expiratory pressure in COPD patients recovering from acute respiratory failure

The detection of intrinsic Positive End-Expiratory Pressure (PEEPi) is one of the main clinical challenge in patients with COPD, since it requires the insertion of gastro-esophageal balloons, to record the changes of transdiaphragmatic pressure (PdiPEEPi) and flow. In this study we aimed to record the PEEPi, using a noninvasive method, comparing it with PdiPEEPi. We recorded a total of 84 measures in 7 COPD patients recovering from an episode of acute hypercapnic respiratory failure requiring mechanical ventilation. In the first part of the study we tested the hypothesys that subtle change from linear to granular pattern of diaphragm-liver interface, during M-mode echography was due to the start of diaphragm contraction. When compared to the start of the Pdi rise (i.e.indication of beginning of diaphragm contraction) the difference between the two we recording was 2.5+1.3 ms. Subsequently we measured with the echographyc method the presence of PEEPi (echoPEEPI) using the following equation: echo PEEPi= P0.1 x T lat /100 where P0.1=mouth occlusion pressure at 100ms,T lat = echography time latency between diaphragm contraction and onset of inspiratory flow. P0.1 was used for measuring the slope of the pressure signal to estimate the effort the patients must overcome to generate inspiratory flow. The Bland and Altman plot showed a bias of only -0,3 cmH20, between PdiPEEPi and ecoPEEPi, with estimated limits of agreement +1.1 and -1.6 cmH20 and a high interoperator concordance (interclass correlation coefficient=0.98, CI95% 0.97-0.99). We described for the first time the good agreement between an invasive recording of PEEPi and a noninvasive one.