Prone positioning in ARDS

Prone positioning, first proposed in 1974 and first applied in ARDS patients in 1976, results in improved arterial oxygenation in most patients. After the introduction of CT scanning, showing lung consolidation located in the dependent lung regions and the aerated baby lung in the nondependent lung regions, we integrated prone positioning as standard practice in clinical treatment of ARDS patients to improve systemic oxygenation. The initial hypothesis was that better perfusion of the baby lung, located in the dependent lung regions after prone positioning, would provide advantages in gas exchange. The picture observed was quite different, however. We did observe an improvement in arterial oxygenation, but the mechanism was likely different because CT scans taken in the prone position showed a density redistribution toward the dependent lung areas. This observation led to our introduction of the \u201csponge model\u201d as our pathophysiologic understanding of ARDS. Whatever the position of the patient, the increased weight of the nondependent lung tissue squeezes the gas out of the dependent regions of the lung. The mechanisms of improved gas exchange were different from that first hypothesized. It is not the aim of this chapter to discuss the possible physiologic mechanisms of prone positioning, which may be found elsewhere. Taken together, all of the studies, including small and large series of patients, consistently showed that in 70% of the patients systemic oxygenation improves in prone compared with supine positioning, without any change in the applied airway pressure. There is no doubt that in life-threatening severe hypoxemia a trial in the prone position is indicated. A different issue is the effectiveness of the prone position in improving ARDS outcome. Is mechanical ventilation in ARDS less harmful in the prone compared with the supine position? Does mechanical ventilation induce less alveolar stress and strain in the prone position? There is a consistent physiologic rationale to believe that this is the case. In experimental settings and in normal subjects and patients affected by ARDS, CT scan shows a more homogeneous distribution of gas throughout the lung parenchyma in the prone compared with the supine position. This observation strongly suggests that the distribution of alveolar stress and strain is more homogeneous in the prone position. In experimental models of ARDS, there is evidence that prone positioning prevents or significantly delays the development of VILI. Two large randomized studies on prone positioning were unable to show a significant benefit on outcome; however, prone positioning was applied for only about a quarter of the day, and mechanical ventilation was not controlled. In a more recent trial, in which prone positioning was applied for 20 hours per day and mechanical ventilation was strictly controlled, a positive benefit was found for the patients treated with prone positioning. On these basis the Prone-Supine II study was organized to detect potential survival benefit of prone positioning avoiding the limitations of previous trials. Although, the study was not able to show a significant survival benefit in the general population, a favorable trend was detected in the subgroup of patients with severe ARDS. In a meta-analysis including 10 clinical trials on adults and children Sud et al. found that prone ventilation reduced mortality in severely hypoxemic patients (PaO2/FiO2 64100 mmHg, p = 0.01) but not in patients with PaO2/FiO2 >100 mmHg (p = 0.36). The authors suggestion was that prone position may provide benefits in severely hypoxemic but it should not be routinely used in all patients affected by acute hypoxemic respiratory failure. In a pooled analysis of the four largest databases of trials on prone position, the absolute mortality reduction in severe ARDS treated in prone position was approximately 10% (log-rank = 0.03). On the contrary in patients with moderate ARDS prolonged prone position may be useless or possibly harmful