Acute lung injury: outcomes and new therapies.

Mortality rates in ARDS are improving, with several recent studies reporting mortality in the order of 20-40% rather than the early descriptions of this disease in which a mortality of 40-60% or higher was frequently cited. The ability to accurately predict outcomes plays an important role in the assessment of the impact of new therapies. Traditionally clinicians have relied on simple respiratory indices to assess mortality risk; however, the predictive ability of such indices, particularly early in the course of the disease, is somewhat limited. Adult data suggest that improved prediction not only of the outcome of established ARDS but also of the development of ARDS in at-risk patients may be obtained by measuring the concentrations of inflammatory mediators and/or surfactant-associated proteins in plasma or bronchoalveolar lavage samples. A bewildering array of therapies for ARDS is available; in many cases the benefits are uncertain. Treatments of proven value in adults include using PEEP beyond the lower inflection point of the pressure-volume curve and limiting tidal volumes to 6 ml/kg. Nitric oxide appears to offer no benefit to outcomes, although it does improve oxygenation in some patients. Surfactant is still undergoing assessment in randomised controlled trials; however, the use of aerosolised surfactant has been recently shown to be ineffective in adult patients with ARDS. Perfluorocarbon-assisted gas exchange (PAGE) or partial liquid ventilation is similarly still being assessed in randomised controlled trials in adults.