Evaluation of oxygen uptake and delivery in critically ill patients: A statistical reappraisal

Objective The evaluation of oxygen consumption (VO₂) and oxygen delivery (DO₂) has gained increasing importance in the monitoring of critically ill patients. They can be obtained from either direct measurements or by indirect calculations based on the Fick principle. However the choice between these two approaches remains controversial. The aim of the study was to investigate whether these 2 methods provide similar results, and if not, to define the best one in terms of reproducibility. Design Oxygen delivery and oxygen consumption were prospectively analyzed in 171 consecutive critically ill patients. Metabolic data were obtained simultaneously. Setting The study was completed in the intensive care unit as part of the management of the patients studied. Patients A first “group” of 279 evaluations was carried out in 73 consecutive critically ill patients. The results were subsequently validated by 423 observations performed in the 98 following patients. Interventions Before and during each evaluation, the patients were kept in stable hemodynamic and metabolic conditions. All were mechanically ventilated. Measurements and results VO₂ was evaluated by calculation (Fick principle) and direct measurement using indirect calorimetry. Cardiac output was both measured by the thermodilution technique and calculated (Fick principle) and the data were used for the evaluation of the directly measured and indirectly calculated DO₂. For both VO₂ and DO₂ the agreement between direct and indirect evaluations was not satisfactory. Differences as great as 55 ml/min·m² and 267 ml/min·m² between simultaneously measured and calculated VO₂ and DO₂ respectively may be expected. Finally, the indirect calculated methods were less reproducible than the measured ones. These observations resulted mainly from the cumulative effects of the random errors in the metabolic data entering into the calculation of VO₂ and DO₂. Conclusions Our data suggested that the indirect calculation (Fick equation) and the direct measurement (indirect calorimetry, thermodilution) of both VO₂ and DO₂ did not provide similar results. Direct measurements are more reproducible methods and must be preferred.