Scholarly article on topic 'Prone ventilation in acute respiratory distress syndrome'

Prone ventilation in acute respiratory distress syndrome Academic research paper on "Medical engineering"

Share paper
Academic journal
European Respiratory Review
OECD Field of science

Academic research paper on topic "Prone ventilation in acute respiratory distress syndrome"



Prone ventilation in acute respiratory distress syndrome

Claude Guérin

Number 1 in the Series ''Acute Respiratory Distress Syndrome'' Edited by S. John Wort and Stefano Nava

Affiliations: Réanimation Médicale, Hôpital de la Croix-Rousse, Hospices Civils de Lyon, Université de Lyon, Créatis INSERM 1044, Lyon, France.

Correspondence: Claude Guérin, Réanimation Médicale, Hôpital de la Croix-Rousse, Hospices Civils de Lyon, Université de Lyon, Créatis INSERM 1044, 103 Grande Rue de la Croix-Rousse, Lyon 69004, France. E-mail:

ABSTRACT Prone positioning has been used for many years in patients with acute lung injury (ALI)/ acute respiratory distress syndrome (ARDS), with no clear benefit for patient outcome. Meta-analyses have suggested better survival in patients with an arterial oxygen tension (PaO2)/inspiratory oxygen fraction (FIO2) ratio <100 mmHg. A recent randomised controlled trial was performed in ARDS patients after a 12-24 h stabilisation period and severity criteria (PaO2/FIO2 <150 mmHg at a positive end-expiratory pressure 05 cmH2O). This trial has demonstrated a significant reduction in mortality from 32.8% in the supine group to 16% in the prone group (p<0.001). The reasons for this dramatic effect are not clear but probably involves a reduction in ventilator-induced lung injury due to prone positioning, for which there is ample evidence in experimental and clinical studies.

The aims of this article are to discuss: the rationale of prone positioning in patients with ALI/ARDS; the evidence of its use based on trial analysis; and the limitations of its use as well as the current place of prone positioning in the management of patients with ALI/ARDS.

From the currently available data, prone positioning should be used as a first-line therapy in patients with severe ALI/ARDS.


Prone positioning and low tidal volume should be the first-line therapy in patients with continued ARDS and severity criteria


It took almost 40 years from the early recommendation by Bryan [1] to use prone positioning in patients with acute respiratory distress syndrome (ARDS) to the demonstration of its beneficial effect on patient survival (fig. 1) [2]. Meanwhile, research has explored the mechanisms by which prone positioning could improve oxygenation and reduce ventilator-induced lung injury (VILI). Furthermore, several randomised controlled trials (RCTs) have been performed to test the effect of prone positioning on patient outcome. Of interest over these 40 years was the continuous interaction between pathophysiological advances and the search for clinical evidence of efficiency, and the continuous refinement in trial design. The aims of this article are to summarise the rationale for prone positioning, the level of evidence supporting its use, its limitations and its place in the current management of ARDS.

For editorial comments see page 157.

Received: Feb 04 2014 | Accepted after revision: March 08 2014 Conflict of interest: None declared. Provenance: Submitted article, peer reviewed.

Copyright ©ERS 2014. ERR articles are open access and distributed under the terms of the Creative Commons Attribution Non-Commercial Licence 4.0.

FIGURE 1 A patient with acute respiratory distress syndrome receiving mechanical ventilation in the intensive care unit while in the prone position.



From an historical point of view, the first rationale for using prone positioning in ARDS patients is improvement in oxygenation. We now know that this goal is less important than other rationale that will be discussed later. A large number of reports have shown that the improvement in oxygenation in ARDS was frequent, observed in ~75% of the cases, and sometimes dramatic, i.e. during a prone session. The oxygenation response is commonly defined as an increase in the arterial oxygen tension (PaO2) by 20% or an increase in the PaO2/inspiratory oxygen fraction (FIO2) ratio by 020 mmHg from the supine position. Therefore, from the onset prone positioning has mainly been thought of and used as a rescue therapy to relieve life-threatening hypoxaemia. For instance, in 1997 Mure et al. [3] reported on 13 ARDS patients, mostly secondary lung injury (indirect lung injury stemming from nonrespiratory severe sepsis/septic shock), who received low positive end-expiratory pressure (PEEP) and high FIO2. 11 of the patients had a PaO2/FIO2 ratio of <100 mmHg in the supine position and in four patients this ratio was <50 mmHg. In six patients, the PaO2/FIO2 ratio increased by a factor of at least four in the prone position. From their results, the authors concluded that prone positioning should be used as a first-line therapy before the use of nitric oxide and extracorporeal membrane oxygenation (ECMO). Moving forward into the RCT era, the effects on oxygenation were then assessed for prone positioning using a group, not just as a single session. Therefore, the effects on oxygenation in the prone position group were compared to the supine position group. Metaanalyses showed a significantly greater risk, ~30-40%, for better oxygenation in the prone position group [4, 5].

In 1988, Langer et al. [6] reported on the effects of prone positioning on oxygenation in 13 ARDS patients, mostly primary lung injury (direct lung injury mainly stemming from pneumonia or aspiration), who received higher PEEP and lower FIO2. Furthermore, the authors performed lung computed tomography (CT) scans in two patients, not only in the supine position but also in the prone position. In patient 4, PaO2 increased from 76 mmHg in the supine position to 141 mmHg in the prone position, with a similar PEEP of 10 cmH2O and Fio2 of 0.60. The lung densities observed on the lung CT scan redistributed from the dorsal to the ventral regions with the corresponding body position change. In contrast, in patient 9, for an almost similar redistribution of the dorsal lung densities towards the ventral regions when in prone position, PaO2 decreased from 101 mmHg to 64 mmHg from the supine to the prone position, with the same PEEP of 10 cmH2O and FIO2 of 0.60. These unexpected findings have not been previously observed. Thus, the authors performed a systematic investigation of lung CT scans in the supine and prone position in normal subjects and in patients with acute respiratory failure [7]. The patients received, on average, PEEP of 11 cmH2O, FIO2 of 0.58 and tidal volume (VT) of 10 mL?kg-1 measured body weight. The gas/tissue ratio assessed on the CT scan decreased progressively across the ventral to dorsal gradient in the supine position, indicating a progressive loss of aeration. With the prone position, the gas/tissue ratio increased towards the dorsal regions, reflecting more aeration in these regions, but was decreased in the ventral regions, which were dependent on the gravitational force. Accordingly, the improvement in oxygenation in the prone position is due to a reduction in intrapulmonary shunt [8] and results from the concomitant effect of the increase in aeration in the dorsal lung regions, which was greater than the loss of aeration in the ventral regions, and the prevalence of lung perfusion into those dorsal regions. The fact that blood continues to flow towards the dorsal regions in the prone position has been demonstrated in many studies. Therefore, the reduction in intrapulmonary shunt is due to more ventilation in well-perfused lung areas.

Ventilator-induced lung injury

The second rationale to use prone positioning is in the prevention of VILI [9]. Preventing VILI has been established as the primary goal of mechanical ventilation after the ARDS network demonstrated that lower VT improved survival compared to higher VT in ARDS patients [10]. This trial was the ultimate demonstration, after decades of research [11, 12], that lung overdistension was detrimental not only to lung healing but also to patient outcome. Lung overdistension is the first component of VILI and, thus, preventing lung overdistension is the main concern of intensivists when setting mechanical ventilation. As it has become clear that lung overdistension pertains to volutrauma [13], not barotrauma (fig. 2), the relevant measurement should not be airway pressure but transpulmonary pressure determination. Transpulmonary pressure is the difference between alveolar pressure (airway pressure at zero airflow when the alveoli and conducting airways are communicating) and pleural pressure. Transpulmonary pressure can even be high if airway pressure is low. The typical example is pressure support ventilation when mechanical breaths are synchronised with a patient's inspiratory effort, a combination that may promote very high VT, causing a risk for volutrauma [14]. Active research is currently dedicated to spontaneous breathing in ARDS patients.

The second component of VILI is called atelectrauma (fig. 2), which results from the repeated opening and closing of the small airways [15]. When VT is reduced, atelectrauma may be less important than overdistension as three large RCTs failed to demonstrate a beneficial effect on patient outcome of higher PEEP compared to lower PEEP [16-18]. However, individual meta-analysis found a statistically significant lower mortality rate in the higher PEEP group in the subgroup of patients with a PaO2/FIO2 ratio f 200 mmHg [19].

A clear and direct demonstration that prone positioning can prevent VILI was reported by Broccard et al. [20] who ventilated normal dogs with 77 mL?kg-1 measured body weight VT to reach a transpulmonary plateau pressure of 35 cmH2O. During 6 h of such mechanical ventilation in the supine position the lungs were macroscopically and microscopically severely injured. When the dogs received the same ventilator settings for 6 h in the prone position the lung injury was markedly reduced. Furthermore, the histological lung injury due to high VT was more homogeneously distributed throughout the lungs in the prone position.




Atelectrauma (shear stress]

FIGURE 2 a) The distribution of normally aerated (white circles), poorly aerated (grey circles), non-aerated (black rectangle) and consolidated (green rectangle) lung areas during acute respiratory distress syndrome while in the supine position during end-inspiration and end-expiration. b) Barotrauma (alveolar rupture with air leaks) and volutrauma (overdistension in the normally aerated lung areas). c) Atelectrauma, i.e. shear stress in the poorly aerated lung areas close to the consolidated non-recruitable lung areas. Biotrauma (biochemical and biological response) results from volutrauma and/or atelectrauma with activation of pro-inflammatory mediators within the lungs and distant end organs. The red circle represents the heart.





K^y HD

FIGURE 3 Homogenisation of the distribution of lung aeration as a result of moving from a) the supine to b) the prone position during acute respiratory distress syndrome. The red circle represents the heart. White circles: normally aerated lung areas; grey circles: poorly aerated lung areas; black rectangle: non-aerated lung areas; green rectangle: consolidated lung areas.

Further findings pertaining to VILI prevention using prone positioning have increased over time in the literature. Prone positioning makes the following more homogeneously distributed in the anterior-to-posterior direction throughout the lungs: lung densities (fig. 3) [7, 21], as previously discussed; intrapulmonary shunt [22]; lung ventilation [23]; and transpulmonary pressure [24]. By favouring such a homogenisation, prone positioning prepares the lung to receive the strain imposed by mechanical ventilation [25], and hence makes the distribution of the resulting stress more homogeneous across the lung. Stress/strain homogenisation is associated with less risk for VILI. Furthermore, the global lung stress and strain is reduced in the prone position. Mentzelopoulos et al. [26] found that transpulmonary pressure, i.e. lung stress, and VT to end-expiratory lung volume ratio, i.e. lung strain, were lower in the prone position than in the supine position. Galiatsou et al. [27] performed a lung CT scan in ARDS patients in the supine position and then in the prone position, same patient restricted to different interventions. The authors found that the prone position was associated with significant alveolar recruitment and less hyperinflation compared to the supine position, and that these effects were more important in lobar than diffuse ARDS anatomical pattern. These findings were confirmed and even expanded on by Cornejo et al. [28], who assessed the lung recruitability in the supine position. The prone position promoted lung recruitment and reduced overdistension in patients in both categories of low and high recruitability at either low or high PEEP in the prone position. However, tidal recruitment and derecruitment, i.e. cyclic opening and closing of the small airways (atelectrauma), and tidal hyperinflation were significantly reduced in the prone position in only the subgroup of ARDS patients who had high recruitability in the supine position and who were receiving higher PEEP in the prone position.

VILI is subtended by biochemical and biological events implicated in the regulation of lung inflammation, a process termed biotrauma (fig. 2). Papazian et al. [29] found that the prone position was associated with lower lung concentration of interleukin (IL)-8, IL-6 and IL-1p compared to the supine position in ARDS patients ventilated with 6 mL?kg-1 predicted body weight VT and higher PEEP. In rodents subjected to injurious mechanical ventilation (VT 18 mL?kg-1 and PEEP 0 cmH2O), prone ventilation has been shown to maintain the expression of mitogen-activated protein kinase (MAPK)-phosphatase 1, a pivotal regulator in VILI, while the supine position was associated with a significant downregulation [30]. Mice deficient in MAPK-phosphatase 1 were more susceptible to VILI [30].

TABLE 1 Characteristics of the five largest randomised controlled trials testing the role of prone positioning in patient survival

First author [ref.]

Gattinoni [31] Guerin [32] Mancebo [33] Taccone [34] Guérin [2]

Patients n

Supine position 152 378 60 174 229

Prone position 152 413 76 168 237

Patients with ARDS %

Supine position 93.3 28 100 100 100

Prone position 94.7 33.9 100 100 100

Pa02/Fl02 at inclusion mmHg 127 150 147 113 100

Tidal volume at inclusion mLkg-1 10.3 MBW 8 MBW 8.4 PBW 8 PBW 6.1 PBW

PEEP at inclusion cmH20 10 8 12 10 10

Prone position session duration* h 7 8 17 18 17

Mortality %

Supine position 25 31.5 58 32.8 32.8

Prone position 21.1 32.4 43 31 16

ARDS: acute respiratory distress syndrome; PaO2: arterial oxygen tension; Fio2: inspiratory oxygen fraction; PEEP: positive end-expiratory pressure; MBW: measured body weight; PBW: predicted body weight. #: average hours per session.

So, there is large body of evidence supporting the fact that prone position ventilation has relevant beneficial physiological effects in ARDS. These benefits were progressively discovered together with major advances in mechanical ventilation. Whether these physiological benefits can translate into improvement in patient outcome will be discussed in the next section.


To date, five large RCTs (table 1), as well as some smaller ones, have tested the role of prone positioning in patient survival. Over time there has been a continuous refinement of the design of the RCTs taking into account the advances made in mechanical ventilation (table 2). For example, protective lung ventilation was only applied in the two most recent trials [2, 34], the duration of prone sessions was increased from the third trial onwards [2, 33, 34], and the targeted population focused on the most severe ARDS patients in terms of hypoxaemia, and crossover was not allowed for patients allocated to the supine position group except as a rescue therapy in the two most recent trials [2, 34]. The four largest RCTs performed metaanalyses on both study and patient levels [31-34]. Both analyses consistently found that prone positioning was associated with a significant reduction in mortality in the subgroup of patients with a PaO2/FIO2 ratio <100 mmHg.

We designed, performed and published the fifth large RCT on prone positioning in a specific ARDS subgroup, termed severe ARDS [2]. This was defined according to local criteria before the Berlin definition was released [35]. We found a significant reduction in 28-day mortality (the primary end-point) with 32.8% mortality in the supine group versus 16% mortality in the prone group (p<0.001). At 90 days, mortality in the supine and prone groups was 41% and 23.6%, respectively (p<0.001).

Our RCT included 10 specific features (table 3), which, in addition to the physiological benefits of prone positioning discussed previously, may explain this result. The stabilisation period of 12-24 h we mandated had previously been used by VILLAR et al. [36] to select the most severe ARDS patients. It could be that this stabilisation period in our RCT included selected patients who may have benefited from prone positioning. However, it remains unclear whether this had occurred and what could be the mechanisms underlying this hypothesis. It should be stressed that the PROSEVA (Proning Severe ARDS Patients) trial [2] included highly selected patients by study design: severity criteria, stabilisation period, and several noninclusion criteria. Therefore, one issue for a wider use of prone positioning may be related to the capability to select the right patients. Furthermore, the skills of the caregivers are essential for the safety of the procedure. This important point will be discussed below.

The PROSEVA trial [2] and the second Italian RCT (PSII: Prone Supine II) [34] were very close in their design to the five largest RCTs (table 4), as they implemented most of the advances achieved in mechanical ventilation, prone positioning and adjunct therapies. The reasons why PROSEVA succeeded and PSII did not are unclear but may be related to the items presented in table 4, in particular strict lung protective ventilation, concentrated prone period, less complications and larger sample size in the former study.

TABLE 2 Limitations in randomised controlled trials

First author [ref.]

Gattinoni [31]

Güérin [32]

Mancebo [33]

Taccone [34]

Güérin [2]

Year of trial

Protective lung ventilation Long prone sessions Target population

Crossover allowed Rate of crossover

2001 No No ALI/ARDS

Yes Not reported

2004 No No

Hypoxaemic acute respiratory failure Yes

81 (21.4%) out of 378

2006 No Yes ARDS

5 (8.1%) out of 62

2009 Yes Yes ARDS

No (rescue) 20 (11.5%) out of 174

2013 Yes Yes

ARDS with severity criteria No (rescue) 17 (7.4%) out of 229

ALI: acute lung injury; ARDS: acute respiratory distress syndrome.

It should also be highlighted that, for a greater severity of patients enrolled in the PROSEVA trial, the outcome in the control group was similar to that in the PSII trial (table 1). This finding may reflect the continuous improvement in the patient's care and the skills in performing prone positioning in the participating intensive care units in the PROSEVA trial. This latter argument is supported by the fact that the same rate of complications between the two groups was observed in the PROSEVA trial contrary to the previous RCTs, including PSII.


Our trial has some limitations as the two groups were not completely balanced (by chance) at the time of randomisation. In particular, the Sequential Organ Function Assessment score was lower in the prone group. This was observed even though in order to be included the patients had to have a mean arterial blood pressure 0 65 mmHg. Even though this criterion was met by every patient in both groups, the rate of patients receiving vasopressors was greater in the supine group. However, in the multivariate analysis the effect of the prone position was still highly significant.

From the onset it was pointed out that the procedure of prone positioning exposes serious complications, in particular those related to airways, such as endotracheal tube displacement (main stem intubation or non-scheduled extubation), endotracheal tube obstruction or kinking, and vascular access kinking or removal. In our trial [32], as in others [31, 33], the rate of complications was significantly greater in the prone group compared to the supine group [4]. However, the mortality was not higher in the prone group. The same was true in the PSII RCT, in which the use of rotoprone was implicated in the observed higher rate of complications in the prone group. Therefore, the caregivers were very reluctant to expand the use of prone positioning in their intensive care unit given the lack of a clear significant benefit. The benefit/risk ratio was judged to not be in favour of extensive use of prone positioning. In the PROSEVA trial, for the first time, the rate of serious complications was similar between the two groups. This finding is probably the result of the expertise and skills of the centres involved in the trial that performed the procedure safely. Following the PROSEVA trial the benefit/risk ratio has greatly improved due to the significant increase in survival.

TABLE 3 Key features in the PROSEVA trial on prone positioning in severe acute respiratory distress syndrome (ARDS) patients

1. ARDS criteria confirmed after 12-24 h

2. ARDS with severity criteria:

PaO2/Fio2 <150 mmHg with Fio2 00.6 + PEEP 05 cmH2O + VT 6 ml-kg"1 predicted body weight

3. Several noninclusion criteria

4. Strict lung protective mechanical ventilation (in both groups)

5. First prone position session started within 1 h after randomisation

6. Prone sessions of at least 16 h consecutively

7. Predetermined stopping criteria of prone position

8. Crossover not allowed except as a lifesaving procedure in the supine group

9. Neuromuscular blockade in both groups

10. Centres with expertise in prone positioning for many years

PROSEVA: Proning Severe ARDS Patients; PaO2: arterial oxygen tension; Fio2: inspiratory oxygen fraction; PEEP: positive end-expiratory pressure; VT: tidal volume.

TABLE 4 Characteristics of the PSII and PROSEVA randomised controlled trials on prone positioning in acute respiratory disease syndrome patients

PSII [34]


Centres n Support Patients n Supine position Prone position Inclusion criteria Stratification of randomisation Stabilisation period Stopping prone criteria

Target tidal volume mL-kg-1 PEEP management Target oxygenation Target plasma pH SAPS II SOFA score

Pa02/Fl02 at inclusion mmHg PEEP at inclusion cmH20 Tidal volume at inclusion mL-kg'1 Pplat at inclusion cmH20 Average prone session duration h Time in prone# % Average prone position sessions per patient in the prone position group

Rotoprone in 20 centres

168 174

Pa02/Fl02 f 200 mmHg + PEEP 5-10 cmH2O Yes No

Resolution of ARF

8 PBW Local PEEP/Fl02 table Pa02 70-90 mmHg 7.30-7.45 41 + 15 6.8 + 3.9 113 + 39 10 + 3 8.0 + 1.7 PBW Not available 18+4 50 8+ 6

Standard bed

229 237

Pa02/Fl02 <150 mmHg + PEEP 05 cmH2O + Fio2 00.60 No Yes

Improvement in oxygenation (Pa02/Fl02 >150 mmHg + PEEP <10 cmH20 + Fio2 <0.60) 6 PBW

PEEP/Fl 02 table used in the ARMA trial Pa02 55-80 mmHg 7.20-7.45 46 + 16 10.0 + 3.3 104 + 25 10 + 4 6.1+0.6 PBW 24 + 5 17 + 4 73 4 + 4

Data are presented as mean + SD, unless otherwise stated. PS I I: Prone Supine I I : PROSEVA: Prone Severe ARDS Patients; PaO2: arterial oxygen

tension; Fio2: inspiratory oxygen fraction; PEEP: positive end-expiratory pressure; ARF: acute respiratory failures; SAPS: Simplified Acute

Physiology Score; SOFA: Sequential Organ Function Assessment; Pplat: plateau pressure; PBW: predicted body weight. position between the start of the first session and the end of the last session.

: time spent in the prone

However, there is another side-effect of prone positioning that may cause some concern to the patients, their family and the caregivers, namely pressure ulcers. Previous trials and meta-analyses, also reported a greater incidence of pressure ulcers with the use of prone positioning. Furthermore, the facial location of pressure ulcers may have a psychological impact on the patients and their family. In the PROSEVA trial, as an ancillary study, the location and stage of pressure ulcers at time of inclusion, 7 days after inclusion and at intensive care unit discharge was prospectively reported in both groups. A greater incidence of new pressure ulcers was found in the prone position group during the first week and at the time of intensive care unit discharge [37]. However, in the logistic regression analysis the position group was no longer significantly associated with new pressure ulcers. An unresolved issue was the role of prone position per se and greater survival in the prone group with the subsequent greater risk for pressure ulcers. At any rate, preventive means should be tested in the prone position.

Place of prone positioning in the management of ARDS patients

Three interventions have proven beneficial in ARDS: lower VT [10], neuromuscular blocking agents (ACURASYS (ARDS et Curarisation Systematique) trial) [38], and prone positioning [2]. Lower VT is the common strategy that must be applied to any ARDS patient regardless of the level of hypoxaemia [39]. Two recent meta-analyses have been published, which included PROSEVA in addition to the previous RCTs. Both found that prone positioning improves survival irrespective of the level of hypoxaemia provided a lower VT is set [40, 41].

Neuromuscular blocking agents and prone positioning were investigated in ARDS patients with a PaO2/FIO2 ratio <150 mmHg at PEEP 05 cmH2O. An Fio2 of at least 0.60 and a 12-24-h stabilisation period were added to the inclusion criteria. Clearly, these two interventions are tightly linked and should be used together as first-line therapy in patients exhibiting the criteria mentioned above. The recent Berlin proposal split ARDS patients into mild, moderate and severe categories at 300, 200 and 100 mmHg PaO2/FlO2 ratio

thresholds, respectively. It is unclear whether mortality does regularly increase from the mild to the severe ARDS category in the Berlin definition [42].

Based on the positive results of the ACURASYS and PROSEVA trials, it would be more appropriate to split ARDS patients into only two categories, a PaO2/FlO2 ratio of <150 mmHg and >150 mmHg assessed at PEEP 05 cmH2O. In ARDS patients with an PaO2/FlO2 ratio <150 mmHg, the mechanical ventilation should start with the following settings: lower VT, neuromuscular blockade for the first 48 h and prone positioning for long sessions until PaO2/FlO2 is >150 mmHg.

Large multicentre trials are ongoing to investigate the effects of the following interventions on patient outcome: PEEP set according to end-expiratory transpulmonary pressure, ECMO and spontaneous ventilation. It is worth noting that in the control group, i.e traditional lung protective ventilation, the use of prone positioning is not mandated. These trials were, however, designed before the release of the results of the PROSEVA trial.

In conclusion, there is now a large body of evidence supporting the fact that prone positioning improves mortality in patients with severe ARDS. Accordingly, prone positioning should be used as a first-line therapy in this setting.


1 Bryan AC. Conference on the scientific basis of respiratory therapy. Pulmonary physiotherapy in the pediatric age group. Comments of a devil's advocate. Am Rev Respir Dis 1974; 110: 143-144.

2 Guerin C, Reignier J, Richard JC, et al. Prone positioning in severe acute respiratory distress syndrome. N Engl J Med 2013; 368: 2159-2168.

3 Mure M, Martling CR, Lindahl SG. Dramatic effect on oxygenation in patients with severe acute lung insufficiency treated in the prone position. Crit Care Med 1997; 25: 1539-1544.

4 Sud S, Sud M, Friedrich JO, et al. Effect of mechanical ventilation in the prone position on clinical outcomes in patients with acute hypoxemic respiratory failure: a systematic review and meta-analysis. CMAJ2008; 178: 1153-1161.

5 Abroug F, Ouanes-Besbes L, Elatrous S, et al. The effect of prone positioning in acute respiratory distress syndrome or acute lung injury: a meta-analysis. Areas of uncertainty and recommendations for research. Intensive Care Med 2008; 34: 1002-1011.

6 Langer M, Mascheroni D, Marcolin R, et al. The prone position in ARDS patients. A clinical study. Chest 1988; 94: 103-107.

7 Gattinoni L, Pelosi P, Pesenti A, et al. CT scan in ARDS: clinical and physiopathological insights. Acta Anaesthesiol Scand Suppl 1991; 95: 87-94.

8 Albert RK, Leasa D, Sanderson M, et al. The prone position improves arterial oxygenation and reduces shunt in oleic-acid-induced acute lung injury. Am Rev Respir Dis 1987; 135: 628-633.

9 Slutsky AS, Ranieri VM. Ventilator-induced lung injury. N Engl J Med 2013; 369: 2126-2136.

10 Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. The Acute Respiratory Distress Syndrome Network. N Engl J Med 2000; 342: 1301-1308.

11 Tobin MJ. Culmination of an era in research on the acute respiratory distress syndrome. N Engl J Med 2000; 342: 1360-1361.

12 Webb HH, Tierney DF. Experimental pulmonary edema due to intermittent positive pressure ventilation with high inflation pressures. Protection by positive end-expiratory pressure. Am Rev Respir Dis 1974; 110: 556-565.

13 Dreyfuss D, Saumon G. Ventilator-induced lung injury: lessons from experimental studies. Am J Respir Crit Care Med 1998; 157: 294-323.

14 Leray V, Bourdin G, Flandreau G, et al. A case of pneumomediastinum in a patient with acute respiratory distress syndrome on pressure support ventilation. Respir Care 2010; 55: 770-773.

15 Muscedere JG, Mullen JB, Gan K, et al. Tidal ventilation at low airway pressures can augment lung injury. Am J Respir Crit Care Med 1994; 149: 1327-1334.

16 Brower RG, Lanken PN, MacIntyre N, et al. Higher versus lower positive end-expiratory pressures in patients with the acute respiratory distress syndrome. N Engl J Med 2004; 351: 327-336.

17 Meade MO, Cook DJ, Guyatt GH, et al. Ventilation strategy using low tidal volumes, recruitment maneuvers, and high positive end-expiratory pressure for acute lung injury and acute respiratory distress syndrome: a randomized controlled trial. JAMA 2008; 299: 637-645.

18 Mercat A, Richard JC, Vielle B, et al. Positive end-expiratory pressure setting in adults with acute lung injury and acute respiratory distress syndrome: a randomized controlled trial. JAMA 2008; 299: 646-655.

19 Briel M, Meade M, Mercat A, et al. Higher vs lower positive end-expiratory pressure in patients with acute lung injury and acute respiratory distress syndrome: systematic review and meta-analysis. JAMA 2010; 303: 865-873.

20 Broccard A, Shapiro RS, Schmitz LL, et al. Prone positioning attenuates and redistributes ventilator-induced lung injury in dogs. Crit Care Med 2000; 28: 295-303.

21 Richard JC, Janier M, Lavenne F, et al. Effect of position, nitric oxide, and almitrine on lung perfusion in a porcine model of acute lung injury. J Applied Physiol 2002; 93: 2181-2191.

22 Richter T, Bellani G, Scott Harris R, et al. Effect of prone position on regional shunt, aeration, and perfusion in experimental acute lung injury. Am J Respir Crit Care Med 2005; 172: 480-487.

23 Richard JC, Bregeon F, Costes N, et al. Effects of prone position and positive end-expiratory pressure on lung perfusion and ventilation. Crit Care Med 2008; 36: 2373-2380.

24 Mutoh T, Guest RJ, Lamm WJ, et al. Prone position alters the effect of volume overload on regional pleural pressures and improves hypoxemia in pigs in vivo. Am Rev Respir Dis 1992; 146: 300-306.

25 Chiumello D, Carlesso E, Cadringher P, et al. Lung stress and strain during mechanical ventilation for acute respiratory distress syndrome. Am J Respir Crit Care Med 2008; 178: 346-355.

Mentzelopoulos SD, Roussos C, Zakynthinos SG. Prone position reduces lung stress and strain in severe acute respiratory distress syndrome. Eur Respir J 2005; 25: 534-544.

Galiatsou E, Kostanti E, Svarna E, et al. Prone position augments recruitment and prevents alveolar overinflation in acute lung injury. Am J Respir Crit Care Med 2006; 174: 187-197.

Cornejo RA, Diaz JC, Tobar EA, et al. Effects of prone positioning on lung protection in patients with acute respiratory distress syndrome. Am J Respir Crit Care Med 2013; 188: 440-448.

Papazian L, Gainnier M, Marin V, et al. Comparison of prone positioning and high-frequency oscillatory

ventilation in patients with acute respiratory distress syndrome. Crit Care Med 2005; 33: 2162-2171.

Park MS, He Q, Edwards MG, et al. Mitogen-activated protein kinase phosphatase-1 modulates regional effects of

injurious mechanical ventilation in rodent lungs. Am J Respir Crit Care Med 2012; 186: 72-81.

Gattinoni L, Tognoni G, Pesenti A, et al. Effect of prone positioning on the survival of patients with acute

respiratory failure. N Engl J Med 2001; 345: 568-573.

Guerin C, Gaillard S, Lemasson S, et al. Effects of systematic prone positioning in hypoxemic acute respiratory failure: a randomized controlled trial. JAMA 2004; 292: 2379-2387.

Mancebo J, Fernandez R, Blanch L, et al. A multicenter trial of prolonged prone ventilation in severe acute respiratory distress syndrome. Am J Respir Crit Care Med 2006; 173: 1233-1239.

Taccone P, Pesenti A, Latini R, et al. Prone positioning in patients with moderate and severe acute respiratory distress syndrome: a randomized controlled trial. JAMA 2009; 302: 1977-1984.

Ranieri VM, Rubenfeld GD, Thompson BT, et al. Acute respiratory distress syndrome: the Berlin definition. JAMA 2012; 307: 2526-2533.

Villar J, Perez-Mendez L, Lopez J, et al. An early PEEP/F1o2 trial identifies different degrees of lung injury in patients with acute respiratory distress syndrome. Am J Respir Crit Care Med 2007; 176: 795-804. Girard R, Baboi L, Ayzac L, et al. The impact of patient positioning on pressure ulcers in patients with severe ARDS. Results from a multicentre randomised controlled trial on prone positioning. Intensive Care Med 2014; 40: 397-403. Papazian L, Forel JM, Gacouin A, et al. Neuromuscular blockers in early acute respiratory distress syndrome. N Engl J Med 2010; 363: 1107-1116.

Ferguson ND, Fan E, Camporota L, et al. The Berlin definition of ARDS: an expanded rationale, justification, and supplementary material. Intensive Care Med 2012; 38: 1573-1582.

Lee JM, Bae W, Lee YL, et al. The efficacy and safety of prone positional ventilation in acute respiratory distress syndrome: updated study-level meta-analysis of 11 randomized controlled trials. Crit Care Med 2013 [In press DOI: 10.1097/CCM.0000000000000122].

Beitler JR, Shaefi S, Montesi SB, et al. Prone positioning reduces mortality from acute respiratory distress syndrome in the low tidal volume era: a meta-analysis. Intensive Care Med 2014; 40: 332-341.

Hernu R, Wallet F, Thiolliere F, et al. An attempt to validate the modification of the American-European consensus definition of acute lung injury/acute respiratory distress syndrome by the Berlin definition in a university hospital. Intensive Care Med 2013; 39: 2161-2170.