Dear Editor,

We read with great interest the recent Intensive Care Medicine article by Cressoni et al. [1], who, using computed tomography scans, prospectively studied the effect of the traditional open lung strategy for ventilation in 33 patients with acute respiratory distress syndrome (ARDS). They showed that the open lung strategy with a plateau pressure of up to 30 cmH2O and positive end-expiratory pressure up to 15 cmH2O were not sufficient to open up the lung and keep it open. Since these are important findings in terms of the potentially harmful effects of the current open lung strategy, we appreciate this research for providing clinically useful information.

However, several factors potentially affecting their results should be discussed. First, the definition of lung inhomogeneity in this study seems to be controversial. They defined the lung inhomogeneity threshold as the percentage of lung volume presenting an inflation ratio of greater than 1.61 in accordance with a previous study by the same group [2]; however, the latter cohort did not have ARDS. The extent of lung inhomogeneities increased with the severity of ARDS, correlated with the physiological dead space, and was associated with overall mortality [3]. Therefore, the optimal threshold should have been calculated in an ARDS population similar to the current cohort.

Second, there was a significant difference in the background of patients. The pathogenesis of ARDS in the mild and moderate groups was predominantly related to sepsis and trauma, while the pathogenesis of ARDS in the severe group was predominantly pneumonia. Morisawa et al. demonstrated that pulmonary vascular permeability was more severely compromised in patients with pulmonary ARDS than those with extrapulmonary ARDS, independent of the Sequential Organ Failure Assessment score [4], although no significant differences were observed in 28-day mortality, mechanical ventilation days, and duration of hospitalization between the two groups. These results suggest that the differences in the pathogenesis of ARDS, and not just the severity of the disease as determined by oxygenation capacity, might have been associated with the different clinical manifestations as well as the observed effect of the open lung strategy.

Third, the interval between disease onset and study enrollment, as well as the treatment strategy, should be taken into account. The delay of 1–5 days in commencement of treatment, different ventilator settings, and fluid management [5] could have affected both the outcome in ARDS patients and the observed effect of the open lung strategy.

In conclusion, we suggest that the authors provide additional data to better understand the association between opening pressures and atelectrauma.