Another important finding in the latter study was that the time dependency disappeared after the measurements were made at PEEP, which was able to reduce the differences in time constants of lung districts. External PEEP changed the value of PEEPi,dyn in both groups in comparison to ZEEP, while the PEEPi,stat remained substantially unchanged. The small difference between dynamic and static PEEP did not permit assessment of the relationship between duration of NPPV and PEEPi,dyn/PEEPi,stat.

In the first day of treatment of patients with severe exacerbation of COPD it may be necessary to apply NPPV continuously, and intolerance to the interface may be a complication. A subgroup of patients, after an initial rapid improvement in blood pH and gas exchange, may require tracheal intubation and sedation owing to intolerance to the face mask. Conti et al. [1] reported a percentage of success of NPPV by facial mask of 38 %, similar to those found by Squadrone et al. [2]. Among the patients who failed NPPV, 46 % failed due to intolerance [9].

In the study of Antonaglia et al. [6], respiratory mechanical parameters were assessed within 6 h of invasive mechanical ventilation under the same conditions, that is, they were sedated, paralyzed, and ventilated with a similar ventilatory setting and PEEP was used during NPPV. The mechanical parameters were measured again after a short period of stabilization at ZEEP and were within the range reported in the literature pertaining to COPD patients [7].

Peak tracheal pressure (Ptr,max) was followed by a rapid initial drop in Ptr (Pi) and by a slow decay of Ptr to an apparent plateau value. This plateau pressure, recorded 5 s after the occlusion, was taken to represent the static end-inspiratory elastic recoil pressure of the respiratory system (Pst). By dividing (Ptr,max – Pi) by the flow (V’) immediately preceding the occlusion, the Newtonian resistance of the respiratory system (Rint) was obtained. The static elastance of the respiratory system (Est) was computed by dividing the corresponding values of (Pst – PEEPi,stat – PEEP) by tidal volume (VT). ∆P was calculated as (Pi – Pst) and represents the pressure dissipated against viscoelastic properties and mechanical inhomogeneities in the respiratory system. We considered the change in end-expiratory lung volume (ΔEELV) relative to the relaxation volume of the respiratory system [6].

ΔP is an indicator of the degree of inhomogeneous time constant distribution and/or viscoelastic behavior within the respiratory system, and it also encompasses the pressure used to overcome the viscoelastic properties of the system. In this line, ΔP tended to be lower in patients intubated owing to intolerance to NPPV than in those who failed due to respiratory causes. In the study of Antonaglia et al. [6], PEEPi,dyn/PEEPi,stat was found to be inversely correlated with ΔP, thus supporting the aforementioned notion of the existence of more important mechanical inequalities in the lungs of these patients.