Non-invasive assessment of lung elastance in patients with acute respiratory distress syndrome

BACKGROUND: Chest wall mechanics can be abnormal in patients with Acute Respiratory Disease Syndrome (ARDS). Therefore, partitioning respiratory system between lungs and chest wall at the bedside is useful to optimize ventilator settings. A non--invasive method for assessing lung elastance (EL), called lung barometry, was previously described on an animal model. METHODS: This prospective study was designed to compare EL assessed by lung barometry (ELLB) versus esophageal pressure (ELPeso). In sedated, paralyzed patients, PEEP was progressively increased from 5 to 40cmH2O then decreased from 40 to 5cmH2O by step of 5cmH2O every two minutes. ELLB was assessed for each step as the ratio between the change in PEEP and the induced end--expiratory lung volume change measured by direct spirometry. ELPeso was calculated from esophageal pressure measurement at each PEEP. EL and the ratio between EL and respiratory system elastance (ERS) calculated with the two methods were compared. RESULTS: Twenty six adult patients with early onset moderate or severe ARDS were included. There was a linear correlation between ELLB and ELPeso during the increase and decrease of PEEP (R?=0.26 and 0.42, respectively). Concordance using Bland and Altman method demonstrated bias and large limits of agreement during the increase (--0.5cmH2O/L; --25 to 24cmH2O/L) and during the decrease in PEEP (--0.3cmH2O/L; --21 to 20cmH2O/L). There were no linear correlation between ELLB/ERS and ELPeso/ERS during the increase and the decrease of PEEP (R?=0.00; R?=0.00, respectively). CONCLUSION: In ARDS patients, lung barometry method cannot be used instead of the esophageal pressure measurement to assess EL.