The increase in C_CW in the prone position [17], which has also been documented by other authors [8, 18], partly results from an increase in abdominal pressure [17]. However, whether or not abdomen is supported in the prone position does not change the effect on either oxygenation or end-expiratory lung volume [19]. If we assume a decrease in C_CW in the prone position from the supine position, the absence of change in the compliance of the respiratory system (C_RS) should indicate a concomitant increase in lung compliance (C_L). By contrast a decrease in C_RS should herald an increase in C_L and, hence, a net lung recruitment. Interestingly, across the trials the impact of the prone positioning on C_RS was not similar. In the trial by Mancebo et al. [20], C_RS has been found decreasing with proning, while no significant change has been observed in the PROSEVA trial [21]. Two CT scan studies demonstrated alveolar recruitment in the prone position as compared to the supine position in patients taken as own control [22, 23]. This result was more likely focal as compared to diffuse loss of aeration in one study [22]. In the other CT scan study, alveolar recruitment was observed regardless the patient had a low or a high potential of recruitability in the supine position [23]. This finding of prone position-induced lung recruitment would indicate that higher PEEP should be set in prone position than in the supine position. A recent study showed that actually the clinicians did not use as much PEEP as they should have in the prone position [24]. The level of PEEP in the prone position is still an open question for two reasons. First, an experimental study in pigs concluded that prone position had same effect as PEEP of 7 cmH₂O would have on oxygenation by investigating various combinations of PEEP and position [25]. Second, using lower PEEP may have hemodynamic benefits, as discussed below. Furthermore, it has been suggested that the strongest predictor of mortality in ARDS patients is the driving pressure of the respiratory system, well above C_RS, tidal volume, or plateau pressure [26]. As prone position may affect C_RS due to its effect on C_CW, it is not clear whether same results may be relevant for the use of proning. By computing the trans-pulmonary driving pressure from the data of a recent trial, the following results deserve mention [27]. Between patients in whom PEEP was set from esophageal pressure as compared to those receiving PEEP set from a PEEP/FiO₂ table, the mean value of trans-pulmonary driving pressure was 10.6 and 10.5 cmH₂O, respectively. At day 3 after randomization, these values averaged 7.3 and 8.7 cmH₂O, respectively. The reduction of trans-pulmonary driving pressure was significantly higher in the esophageal pressure group than in the control group and significantly greater in survivors than in non-survivors (personal communication with D Talmor).

Protecting the lung from VILI is the main goal of mechanical ventilation. There are several lines of evidence to support that prone position can achieve this objective. The abovementioned CT scan studies showed that not only prone position could promote lung recruitment but also reduce hyperinflation [22, 23]. However, only higher PEEP used in the prone position was able to minimize cycling opening and closure during tidal breath, the so-called atelectrauma [23]. The lung concentration of pro-inflammatory cytokines was found reduced in prone as compared to supine position in ARDS patients [28]. The overall stress and strain is reduced in prone position in ARDS patients [18]. Experimental studies found that prone position reduced VILI due to high tidal volume and made it more homogeneously distributed throughout the lung in dogs [29], increased the time required to double elastance of the respiratory system as compared to supine position in rats [30], modulated the expression of a kinase strongly involved in VILI in rats [31], and attenuated VILI due to injurious ventilation in mice deficient for this kinase [31]. Therefore, there is a strong background for VILI prevention by using prone position. The likely mechanism for this to occur is by making the lung distribution of tidal volume, and hence the strain more homogenous, and by minimizing the compression of the lung by its own weight and also that of the heart. It should be made clear from the onset that VILI was not assessed in the trials that investigated the effect of prone position on patient outcome as we shall discuss below. That means that the mechanics by which prone position improves survival in ARDS patients should stem from these beneficial physiological effects.

In daily practice the rate of use of prone position is still limited (Table 6.2). The most recent data, not published as yet, from the large observational international Large Observational Study to Understand the Global Impact of Severe Acute Respiratory Failure (LUNG SAFE) study found that prone position was used in 7% of ARDS patients and in 14% of those with severe ARDS. It is likely that the risk of complications, like vascular line kinking or withdrawal and endotracheal tube obstruction or removal, are barriers that restraint the caregivers from performing prone positioning more frequently. It should be noted that in the PROSEVA trial, the rate of these complications was not significantly different between the two groups contrary to previous trials. This may reflect the fact that centers that participated in this trial have used prone positioning for many years and were able to provide with a safe procedure. Another concern is the pressure sores that occurred more frequently in the prone than in the supine position [57], which should deserve innovative preventive means.