Context: Continuous positive airway pressure (CPAP) is widely used in the belief that it may reduce the need for intubation and mechanical ventilation in patients with acute hypoxemic respiratory insufficiency. Objective: To compare the physiologic effects and the clinical efficacy of CPAP vs standard oxygen therapy in patients with acute hypoxemic, nonhypercapnic respiratory insufficiency. Design, Setting, and Patients: Randomized, concealed, and unblinded trial of 123 consecutive adult patients who were admitted to 6 intensive care units between September 1997 and January 1999 with a PaO2/FIO2 ratio of 300 mm Hg or less due to bilateral pulmonary edema (n=102 with acute lung injury and n=21 with cardiac disease). Interventions: Patients were randomly assigned to receive oxygen therapy alone (n=61) or oxygen therapy plus CPAP (n=62). Main Outcome Measures: Improvement in PaO2/FIO2 ratio, rate of endotracheal intubation at any time during the study, adverse events, length of hospital stay, mortality, and duration of ventilatory assistance, compared between the CPAP and standard treatment groups. Results: Among the CPAP vs standard therapy groups, respectively, causes of respiratory failure (pneumonia, 54% and 55%), presence of cardiac disease (33% and 35%), severity at admission, and hypoxemia (median [5th-95th percentile] PaO2/FIO2 ratio, 140 [59-288] mm Hg vs 148 [62-283] mm Hg; P=.43) were similarly distributed. After 1 hour of treatment, subjective responses to treatment (P<.001) and median (5th-95th percentile) PaO2/FIO2 ratios were greater with CPAP (203 [45-431] mm Hg vs 151 [73-482] mm Hg; P=.02). No further difference in respiratory indices was observed between the groups. Treatment with CPAP failed to reduce the endotracheal intubation rate (21 [34%] vs 24 [39%] in the standard therapy group; P=.53), hospital mortality (19 [31%] vs 18 [30%]; P=.89), or median (5th-95th percentile) intensive care unit length of stay (6.5 [1-57] days vs 6.0 [1-36] days; P=.43). A higher number of adverse events occurred with CPAP treatment (18 vs 6; P=.01). Conclusion: In this study, despite early physiologic improvement, CPAP neither reduced the need for intubation nor improved outcomes in patients with acute hypoxemic, nonhypercapnic respiratory insufficiency primarily due to acute lung injury.
