End-Expiratory Occlusion Test Predicts Preload Responsiveness Independently of Positive End-Expiratory Pressure During Acute Respiratory Distress Syndrome

Objective: A 15-second end-expiratory occlusion increases cardiac preload and allows detection of preload dependence. We tested whether the reliability of this test depends upon positive end-expiratory pressure. Design: Prospective study. Setting: Medical ICU. Patients: Thirty-four patients presenting with acute circulatory failure and acute respiratory distress syndrome ventilated with a tidal volume of 6.7 mL/kg (interquartile range, 6.3-7.1). Measurements: At positive end-expiratory pressure = 5 cm H2O, we measured the changes in cardiac index induced by end-expiratory occlusion and a passive leg raising test. Preload dependence was defined by a passive leg raising-induced increase in cardiac index greater than or equal to 10%. Positive end-expiratory pressure was increased to a plateau pressure of 30 cm H2O, and end-expiratory occlusion and passive leg raising were performed again. Main Results: At positive end-expiratory pressure = 5 cm H2O, 29% of patients were passive leg raising responders. An end-expiratory occlusion-induced increase in cardiac index greater than or equal to 5% detected a passive leg raising-induced increase in cardiac index greater than or equal to 10% with a sensitivity of 90% (95% CI, 56-100) and a specificity of 88% (95% CI, 68-97). At higher positive end-expiratory pressure (15 cm H2O [interquartile range, 13-15]), the plateau pressure - positive end-expiratory pressure difference did not change (15 mm Hg [14-17] vs 15 mm Hg [13-18] before the positive end-expiratory pressure increase). Increasing positive end-expiratory pressure significantly reduced cardiac index in passive leg raising responders (-27% [interquartile range, -6 to -56]) but not in other patients. At high positive end-expiratory pressure, passive leg raising increased cardiac index to a larger extent than at positive end-expiratory pressure = 5 cm H2O (19% [interquartile range, 15-34] vs 16% [interquartile range, 13-23], respectively). The proportion of passive leg raising responders significantly increased (34 vs 29%, respectively), meaning preload dependence had increased. At higher positive end-expiratory pressure, an end-expiratory occlusion-induced increase in cardiac index greater than or equal to 6% detected a passive leg raising-induced increase in cardiac index greater than or equal to 10% with a sensitivity of 100% (95% CI, 75-100) and a specificity of 90% (95% CI, 70-99). Conclusions: The end-expiratory occlusion test is reliable for detecting preload dependence whatever the positive end-expiratory pressure during acute respiratory distress syndrome.