ObjectiveTo determine if oxygen consumption (VO2) calculated using the Fick relationship (calculated VO2) determines total body VO2accurately and precisely enough to employ this method during clinical assessment of oxygen transport.DesignMethods comparison, using repeated measures during four physiologic states: normal heart/normal lungs, heart failure/normal lungs, normal heart/acute lung injury, heart failure/acute lung injury.SettingsUniversity research laboratory.SubjectsThirteen adult Yucatan pigs.InterventionsOleic acid-induced acute lung injury; heart failure was induced with a continuous infusion of esmolol.Measurements and Main ResultsCalculated VO2was determined by multiplying thermodilution cardiac output by the arterialvenous oxygen content difference in anesthetized, spontaneously breathing animals. Conditions were tightly controlled so that calculated VO2would be as accurate as possible. "True" VO2was measured simultaneously with a water-sealed spirometer (spirometry VO2). Calculated VO2and spirometry VO2were determined and analyzed during the four physiologic states listed above. Pooled data also were evaluated. Mean spirometry VO2and calculated VO2differed significantly during all four physiologic states and when data were pooled (spirometry VO2273 plus minus 70, calculated VO2178 plus minus 58 mL/min; p less than .01). Calculated VO2consistently underestimated spirometry VO2, as demonstrated by the large, positive bias in pooled data (95 plus minus 59 mL of oxygen/min) and in the four physiologic states. Linear regression of data from all four states yielded slopes that were indistinguishable from 1, but y intercepts that varied from minus 152 to plus 182. For pooled data, the following equation was used: calculated VO2equals 0.5 times (spirometry VO2plus 46); r2equals .35. Precision in pooled data was 22% of the mean spirometry VO2. Data analysis for the four physiologic states demonstrated results similar to those results obtained when data were pooled.ConclusionsEven in a tightly controlled, clinical simulation in the laboratory, calculated VO2from the Fick relationship systematically underestimated VO2measured with a water-sealed spirometer. If true VO2changes, the magnitude and direction of change will be reflected by calculated VO2but with approximate 20% error in the absolute value. Heart failure, acute lung injury, and their combination did not affect the accuracy of calculated VO2. Therefore, calculating VO2using the Fick relationship is too inaccurate to be used for research purposes. Because assessment of the directional change of VO2may be clinically useful, calculated VO2can be employed with discretion during clinical oxygen transport evaluation, bearing in mind the calculation's inherent imprecision.(Crit Care Med 1996; 24:86-90)