A comparison between the Fick method and indirect calorimetry for determining oxygen consumption in patients with fulminant hepatic failure

OBJECTIVETo compare the Fick method of determining oxygen consumption (VO (2)) with a gas exchange method in a group of patients in whom the cardiac output and mixed venous oxygen saturation values were consistently high. DESIGNA prospective, observational study. SETTINGA ten-bed intensive therapy unit at a university teaching hospital. PATIENTSSeventeen patients suffering from fulminant hepatic failure who required ventilatory support and invasive hemodynamic monitoring. All patients were sedated and paralyzed throughout the study period. INTERVENTIONSNone. MEASUREMENTS AND MAIN RESULTSVO2 was determined simultaneously by indirect calorimetry and by the Fick method five or six times in each patient over a 5-hr period after resuscitation with fluids and, if clinically indicated, norepinephrine infusion. The agreement between the methods was poor (limits of agreement +19 to -101 mL/min/m) and the Fick method consistently underestimated gas exchange measurements (mean bias 41 mL/min/m). The bias varied widely, both between and within individual patients. The reproducibility of the Fick-derived VO2 was worse than the indirect calorimetry measurements, indicating that the dispersion of data attributable to measurement error was greater with the Fick method. CONCLUSIONSUnder clinical conditions, the agreement between Fick calculations and indirect calorimetry measurements of VO2 in hyperdynamic patients with fulminant hepatic failure was extremely poor. The reproducibility of Fick calculations was less than the reproducibility derived by gas exchange measurements because of the large measurement errors that may occur with the Fick method when the cardiac output is large and the arterialvenous oxygen content difference is small. Fick calculations systematically underestimate gas exchange measurements. The Fick method is inaccurate and unreliable when an estimation of VO2 is required in patients with this hemodynamic pattern. (Crit Care Med 1998; 26:1200-1207)