This observational study will analyze data already collected by the investigators as part of their routine clinical practice from patients with acute respiratory failure (ARF) treated with mechanical ventilation. The study itself does not require any specific intervention.

Mechanical ventilation can save the lives of patients with ARF. However, if used improperly, it can exacerbate lung disease and worsen outcomes (Slutsky et al.).

Despite decades of animal and clinical research, it remains unclear how to establish the positive end-expiratory pressure (PEEP) during mechanical ventilation to reduce the risk of lung damage. Several methods have been suggested, but none have consistently proven superior to the others (Sahetya et al.).

As part of their routine clinical practice, the investigators study the responses to different PEEP levels of patients with ARF undergoing mechanical ventilation by integrating information from various techniques, each examining different aspects of lung morphology and physiology. The methods the investigators use include lung computed tomography (CT) and electrical impedance tomography (EIT). Lung CT is the reference technique for measuring the morphological response to PEEP (Gattinoni et al.). It quantifies the volume of the hyperinflated and non-aerated lung, both of which are related to the risk of mechanical ventilation causing damage (Slutsky et al.). Lung EIT monitors the functional response to PEEP in terms of changes in regional compliance across different PEEP levels. Allegedly, an increase in compliance when PEEP is decreased reveals overdistention, the functional correlate of (worrisome) hyperinflation, at the higher PEEP. A decrease in compliance when PEEP is decreased signals new collapse, the functional correlate of (worrisome) loss of aeration (Franchineau et al.).

In the Unit where the investigators work, patients with ARF treated with mechanical ventilation are routinely studied as follows. First, a lung CT with a PEEP of 20 cmH2O and then of 5 cmH2O is obtained. Thereafter, a decremental PEEP test is performed with the EIT, where PEEP is decreased from 20 cmH2O down to 5 cmH2O in steps of 2 or 3 cmH2O. Finally, results are analyzed and compared offline.

At the lung CT, decreasing PEEP from 20 to 5 cmH2O is always associated with some decrease in the volume of the hyperinflated lung and some increase in the volume of the non-aerated lung. However, the magnitude of these two effects varies among individuals, and the net response may be defined as the difference between those two competing effects. If the decrease in the volume of the hyperinflated lung is greater than the increase in the volume of the non-aerated lung, the overall response (i.e., less hyperinflation) can be considered positive. PEEP should then be set closer to 5 than to 20 cmH2O. Diversely, if the decrease in the volume of the hyperinflated lung is smaller than the increase in the volume of the non-aerated lung, the overall response (i.e., more loss of aeration) can be considered negative. PEEP should then be set closer to 20 cmH2O (Protti et al.). Similarly, at the lung EIT, decreasing PEEP from 20 to 5 cmH2O is always associated with compliance improvement in some regions (i.e., less overdistension) and worsening in others (i.e., more collapse). Again, the magnitude of these two opposite effects varies among individuals. According to most experts on lung EIT, PEEP should be set at the level where both overdistension and collapse are minimized (the so-called "best" PEEP) (Jonkman et al.).

Lung CT requires transfer to the radiology unit, exposure of the patient to radiation, and complex analysis offline. By contrast, lung EIT is virtually risk-free, and analysis can be performed using an automatic algorithm. Nevertheless, lung EIT is less well validated than lung CT. For instance, the assumption that a decrease in compliance in response to a decrease in PEEP is due to new collapse has been questioned (Protti et al., Chiumello et al., Menga et al.). So far, lung CT remains the reference technique for studying individual responses to PEEP, while lung EIT requires further validation.

This study aims to verify whether the "best" PEEP identified using lung EIT is strongly associated with the net response assessed using lung CT, when PEEP is decreased from 20 to 5 cmH2O in patients with ARF treated with mechanical ventilation. If so, this would strengthen the rationale for using the lung EIT (which is safer and simpler than the lung CT) to set PEEP.

Inclusion Criteria:

• Adults (≥18 years of age) admitted to our Unit with ARF treated with mechanical ventilation
• The patient undergoes a lung CT and EIT to guide the setting of PEEP as part of our routine clinical practice

Exclusion Criteria:

• The patient cannot undergo a lung CT and/or EIT as judged by the attending physician (for instance, transport to the radiology unit may be considered too risky if the patient is extremely severe, or using EIT may be contraindicated because of the presence of a pacemaker)
• Pregnancy (as this condition alters the respiratory physiology)

Adults admitted to the Intensive Care Unit of the IRCCS Humanitas Research Hospital, in Rozzano (Milan, Italy), with acute respiratory failure treated with mechanical ventilation