Change in cardiac output during Trendelenburg maneuver is a reliable predictor of fluid responsiveness in patients with acute respiratory distress syndrome in the prone position under protective ventilation

Background Predicting fluid responsiveness may help to avoid unnecessary fluid administration during acute respiratory distress syndrome (ARDS). The aim of this study was to evaluate the diagnostic performance of the following methods to predict fluid responsiveness in ARDS patients under protective ventilation in the prone position: cardiac index variation during a Trendelenburg maneuver, cardiac index variation during an end-expiratory occlusion test, and both pulse pressure variation and change in pulse pressure variation from baseline during a tidal volume challenge by increasing tidal volume (VT) to 8 ml.kg-1. Methods This study is a prospective single-center study, performed in a medical intensive care unit, on ARDS patients with acute circulatory failure in the prone position. Patients were studied at baseline, during a 1-min shift to the Trendelenburg position, during a 15-s end-expiratory occlusion, during a 1-min increase in VT to 8 ml.kg-1, and after fluid administration. Fluid responsiveness was deemed present if cardiac index assessed by transpulmonary thermodilution increased by at least 15% after fluid administration. Results There were 33 patients included, among whom 14 (42%) exhibited cardiac arrhythmia at baseline and 15 (45%) were deemed fluid-responsive. The area under the receiver operating characteristic (ROC) curve of the pulse contour-derived cardiac index change during the Trendelenburg maneuver and the end-expiratory occlusion test were 0.90 (95% CI, 0.80–1.00) and 0.65 (95% CI, 0.46–0.84), respectively. An increase in cardiac index ≥ 8% during the Trendelenburg maneuver enabled diagnosis of fluid responsiveness with sensitivity of 87% (95% CI, 67–100), and specificity of 89% (95% CI, 72–100). The area under the ROC curve of pulse pressure variation and change in pulse pressure variation during the tidal volume challenge were 0.52 (95% CI, 0.24–0.80) and 0.59 (95% CI, 0.31–0.88), respectively. Conclusions Change in cardiac index during a Trendelenburg maneuver is a reliable test to predict fluid responsiveness in ARDS patients in the prone position, while neither change in cardiac index during end-expiratory occlusion, nor pulse pressure variation during a VT challenge reached acceptable predictive performance to predict fluid responsiveness in this setting. Trial registration ClinicalTrials.gov, NCT01965574. Registered on 16 October 2013. The trial was registered 6 days after inclusion of the first patient. Electronic supplementary material The online version of this article (doi:10.1186/s13054-017-1881-0) contains supplementary material, which is available to authorized users.