Respiratory airway resistance monitoring in mechanically ventilated patients

Abstract

Physiological models of respiratory mechanics can be used to optimise mechanical ventilator settings to improve critically ill patient outcomes. Models are generally generated via either physical measurements or analogous behaviours that can model experimental outcomes. However, models derived solely from physical measurements are infrequently applied to clinical data. This investigation assesses the efficacy of a physically derived airway branching model (ABM) to capture clinical data. The ABM is derived via classical pressure-flow equations and branching based on known anatomy. It is compared to two well accepted lumped parameter models of the respiratory system: the linear lung model (LLM) and the Dynostatic Model (DSM). The ABM significantly underestimates the total pressure drop from the trachea to the alveoli. While the LLM and DSM both recorded peak pressure drops of 17.8 cmH2O and 10.2 cmH2O, respectively, the maximum ABM modelled pressure drop was 0.66 cmH2O. This result indicates that the anatomically accurate ABM model does not incorporate all of the airway resistances that are clinically observed in critically ill patients. In particular, it is hypothesised that the primary discrepancy is in the endotracheal tube. In contrast to the lumped parameter models, the ABM was capable of defining the pressure drop in the deep bronchial paths and thus may allow further investigation of alveoli recruitment and gas exchange at that level given realistic initial pressures at the upper airways.