ContextIn recent years, protective ventilation with airway pressure limitation has constituted a major advance in acute respiratory distress syndrome treatment and has led to a substantial improvement in prognosis. With this therapeutic rationale, one may even question nowadays whether the severity of respiratory failureper sestill influences mortality.ObjectiveTo determine whether the severity of respiratory failure, scored according to the usual criteria, still influences mortality in acute respiratory distress syndrome patients when a low stretch ventilation was used and to assess the impact on mortality of other nonpulmonary organ dysfunction, particularly circulatory failure.Design and SettingA retrospective study conducted in the medical intensive care unit of a French university hospital from October 1993 to December 2001.PatientsA total of 150 acute respiratory distress syndrome patients who were administered uniform protective ventilation with a limited plateau pressure (<30 cm H2O), a low positive end-expiratory pressure (<10 cm H2O), and the same strategy concerning hemodynamic support and dialysis when required.Main Outcome and MeasuresMean age, general severity index (Simplified Acute Physiologic Score II), number of associated organ failures (Logistic Organ Dysfunction Score), respiratory severity indices (Pao2/Fio2, Lung Injury Severity Score), and severity of initial circulatory failure (circulatory failure present at admission or that developed during the first 48 hrs) were compared, according to recovery or death, and evaluated by a logistic regression model, which allows simultaneous control of multiple factors.ResultsThe average mortality rate for the whole group was 38%, with 93 patients recovering after an average duration of mechanical ventilation of 18 ± 13 days. The major factor significantly and independently associated with probability of dying was the severity of circulatory failure (p= .0001, odds ratio = 10.17). Patients free from initial circulatory failure (39 patients) had a 95% recovery rate.ConclusionWith our low stretch strategy, the severity of circulatory failure was the main determinant of acute respiratory distress syndrome prognosis. Patients with isolated respiratory failure during the first 48 hrs of respiratory support have an excellent chance of recovery when treated with protective ventilation associated with a low positive end-expiratory pressure.