For personal use. Only reproduce with permission from The Lancet Publishing Group. REVIEW 332 THE LANCET • Vol 361 • January 25, 2003 • www.thelancet.com Protective ventilation in patients with acute respiratory distress syndrome (ARDS) Interest in the effects of mechanical stimuli on lung tissue has been generated by clinical studies that have clearly shown the importance of such effects, especially in patients with acute respiratory failure. Results of an international survey 1 showed that most patients with ARDS were ventilated with tidal volumes of at least 10 mL/kg. These large tidal volumes (compared with about 7 mL/kg when spontaneously breathing at rest) tend to maintain a normal partial pressure of carbon dioxide (PaCO 2 ) and prevent atelectasis in patients with lung injury. Conversely, limitation of pressure and volume leads to retention of carbon dioxide, with potentially harmful effects. 2 Although the advantages and disadvantages of hypercapnia are still debated, the permissive hypercapnia technique was associated with a low death rate in a non-randomised study 3 in patients with ARDS in 1990. In 1993, a consensus conference 4 while acknowledging the lack of convincing data for human beings, recommended that this approach should be adopted, by limiting tidal volume to 5–7 mL/kg and plateau pressure to 35 cm H 2 O. Subsequently, clinical studies 5–9 have been published in which protective and conventional ventilatory strategies were compared (table). Whether the conventional groups were ventilated in accordance with best practice in these trials is debatable since prescribed limits were exceeded. Apart from the ARDS Network study, 8 the other studies 5–7,9 were underpowered to show a difference in death rates between the two groups. Similarly, death rates of patients in the conventional treatment group of the Amato study 9 were much higher than would be predicted for this patient group, therefore we have not considered this study any further. Results from only one of the remaining four studies showed a benefit from protective ventilation, however the data perhaps more convincingly show the negative effect of high tidal volumes and pressures on outcome. The ARDS Network study 8 provides the most compelling data for benefits of low tidal volume ventilation in patients with lung injury, not only because the number of patients (n=861) was sufficient to show a survival advantage, but also because the protective ventilation group achieved the lowest tidal volume (6·2 mL/kg ideal bodyweight) and plateau pressure (<30 cm H 2 O), and because results of this trial showed the greatest difference in tidal volume between the protective and conventional ventilation groups compared with the other studies. 8 Although features of the ARDS Network trial protocol provide less biologically plausible explanations for the survival difference (eg, management of respiratory acidosis in the protective ventilation group needed administration of bicarbonate), the need to limit tidal volume and inflation pressures has been established. Aspects of the ARDS Network trial protocol raise issues that need further consideration. First, the required response to respiratory acidosis was initially to increase the respiratory rate and then to give sodium bicarbonate if necessary. The effects of hypercapnia and mild respiratory acidosis on the outcome of patients with ARDS are not known. The protocol also prescribed increased positive end-expiratory pressure as the concentration of oxygen was increased to meet targets for arterial oxygenation. The ARDS Network has recently reported on the Alveoli trial, 10 which confirmed the benefit of a strategy that lowers tidal volume and limits the plateau pressure in patients with ARDS. The death rate of patients ventilated with high and conventional positive end-expiratory pressure strategies was less than 30% and did not differ significantly between the two groups. Hence, the optimum positive end- expiratory pressure in patients with ARDS could not be Lancet 2003; 361: 332–40 Unit of Critical Care and National Heart and Lung Institute (L Pinhu MSc) and Department of Intensive Care Medicine (Prof T W Evans DSc), Imperial College School of Medicine, London, UK (M J D Griffiths MRCP); Department of Respiratory Medicine (T Whitehead MRCP) and Adult Intensive Care Unit (Prof T W Evans, M J D Griffiths), Royal Brompton Hospital, London, UK Correspondence to: Dr Mark J D Griffiths, Adult Intensive Care Unit, Royal Brompton Hospital, London SW3 6NP (e-mail: m.griffiths@ic.ac.uk) Mechanical ventilation is indispensable in support of patients with respiratory failure who are critically ill. However, use of this technique has adverse effects, including increased risk of pneumonia, impaired cardiac performance, and difficulties associated with sedation and paralysis. Moreover, application of pressure to the lung, whether positive or negative, can cause damage known as ventilator-associated lung injury (VALI). Despite difficulties in distinguishing the effects of mechanical ventilation from those of the underlying disorder, VALI greatly assists patients with the most severe form of lung injury, acute respiratory distress syndrome (ARDS). Moreover, modification of mechanical ventilation so that VALI is kept to a minimum improves survival of patients with ARDS. Here, we outline the effects of mechanical ventilation on injured lungs and explore the underlying mechanisms. Ventilator-associated lung injury Liao Pinhu, Thomas Whitehead, Timothy Evans, Mark Griffiths Search strategy and selection criteria We selected references by searching English language articles published in the past 20 years in Pubmed under each subheading of the review. We have done a non-systematic review of the results of our searches and articles collected by the authors. Priority was given to articles published in journals with high impact factors. Review