• Sabri Soussi, Fabrice Vallée, Florian Roquet, Vincent Bevilacqua, Mourad Benyamina, Axelle Ferry, Alexandru Cupaciu, Maïté Chaussard, Christian De Tymowski, David Boccara, Maurice Mimoun, Marc Chaouat, James Anstey, Alexandre Mebazaa, Matthieu Legrand, and PRONOBURN study group.
• *Department of Anesthesiology and Critical Care and Burn Unit, AP-HP, Saint-Louis Hospital, Paris, France †Plastic Surgery and Burn Unit, AP-HP, Saint-Louis Hospital, Paris, France ‡Intensive Care Unit, Royal Melbourne Hospital, Parkville, Melbourne, Australia §Lariboisière, Hospital, INSERM UMR 942, Paris, France ||Paris Diderot University, Paris, France.
• Shock. 2017 Nov 1; 48 (5): 532-538.

ObjectivesTo evaluate the interchangeability of oxygen consumption variations measured with the Fick equation (ΔVO2Fick) and indirect calorimetry (ΔVO2Haldane) in critically ill burns patients.MethodsProspective observational single-center study conducted in a university hospital. Twenty-two consecutive burns patients with circulatory insufficiency and hyperlactatemia (>2 mmol/L) who required a fluid challenge (FC) were included. All patients had cardiac output monitoring (transpulmonary thermodilution technique) and were ventilated and sedated. Simultaneous measurements of VO2Fick and VO2Haldane were performed before and immediately after the FC, at rest, and in hemodynamic conditions stabilized for at least 1 h. VO2Fick and VO2Haldane were measured, respectively, with the standard formulae (using arterial and central venous saturation measured with a blood gas analyzer) and with a metabolic monitor.ResultsForty-four paired measurements of VO2 were obtained. At each timepoint, the median (interquartile range, 25-75) VO2Haldane values were significantly higher than the median VO2Fick values (126 (103-192) vs. 90 (66-149) mL O2/min/m (P = 0.004) before FC and 129 (105-189) vs. 80 (54-119) mL O2/min/m (P = 0.001) after FC). Correlation between the ΔVO2Fick and the ΔVO2Haldane (%) measurements was poor, with an r = 0.06, (P = 0.77). The mean bias was 8.6% [limits of agreement (LOA): -75.7%, 92.9%].ConclusionsAnalysis of agreement showed poor concordance for the ΔVO2Haldane and the ΔVO2Fick (%) with a low mean bias but large and clinically unacceptable LOA. ΔVO2Haldane and ΔVO2Fick (%) are not interchangeable in these conditions.

18 keywords...

hide…