Rethinking ARDS classification: oxygenation impairment fails to predict VILI risk

Abstract

Purpose: The selection and intensity of respiratory support for ARDS are guided by PaO₂/FiO₂. However, ventilator-induced lung injury (VILI) is linked to respiratory mechanics and ventilator settings. We explored whether the VILI risk is related to ARDS severity based on oxygenation.

Methods: We analysed data on 228 ARDS subjects with PaO₂/FiO₂ < 200 mmHg, categorized into three severity groups: one based on PaO₂/FiO₂ ratio, and the others based on tertiles of predictors of VILI: mechanical power ratio (MPR) and driving pressure (DP). In each group of oxygenation-based ARDS severity and MPR and DP tertiles, we measured CT anatomy, gas exchange, respiratory mechanics, VILI prerequisites (lung elastance and lung gas volume), and VILI determinants (tidal volume, PEEP, airway pressures).

Results: Predictors of VILI, such as MPR and DP, were similar across ARDS severity groups based on PaO₂/FiO₂ ratio, while oxygenation remained comparable across different levels of VILI risk defined by MPR and DP. Oxygenation impairment was associated with increased lung weight, recruitability, and reduced well-inflated tissue. In contrast, MPR and DP tertiles affected variables associated with the baby lung size, such as lung gas volume and well-inflated tissue. Mechanical ventilation intensity increased progressively across MPR and DP tertiles, but remained similar across PaO₂/FiO₂ severity groups.

Conclusions: ARDS severity based on oxygenation impairment does not reflect the prerequisites and determinants of VILI. This should prompt a reconsideration of recommending respiratory support based on oxygenation impairment, rather than VILI determinants.

Keywords: ARDS severity; Gas exchange; Mechanical power; Respiratory mechanics; Ventilator-induced lung injury.

Fig. 1

Mechanical power ratio and driving pressure as a function of ARDS severity_P/F classes. Statistical analysis was performed by one-way ANOVA or Kruskal–Wallis test as appropriate. Post hoc multiple comparisons were performed by pairwise Student’s t test or Wilcoxon test with Bonferroni correction. At linear regression mechanical power ratio vs P/F and driving pressure vs P/F had a p = 0.271, R² < 0.01 and p = 0.916, R² < 0.01, respectively. P/F (PaO₂/FiO₂): ratio of arterial partial pressure of oxygen to the inspired fraction of oxygen

Fig. 2

PaO₂/FiO₂ as a function of mechanical power ratio and driving pressure tertiles. Statistical analysis was performed by one-way ANOVA or Kruskal–Wallis test as appropriate. Post hoc multiple comparisons were performed by pairwise Student’s t test or Wilcoxon test with Bonferroni correction. At linear regression P/F vs mechanical power and P/F vs driving pressure had a p = 0.271, R² < 0.01 and p = 0.916, R² < 0.01, respectively. P/F (PaO₂/FiO₂): ratio of arterial partial pressure of oxygen to the inspired fraction of oxygen