ABSTRACT
Continuous mixed venous oxygen saturation (SVO2) measurements from fiberoptic pulmonary artery catheters, arterial oxygen saturation (SaO2) measurements from pulse oximetry, and minute-by-minute oxygen consumption (VO2) measurements from indirect calorimetry can be used for near-continuous estimation of cardiac output (Qt) and oxygen delivery (DO2) by application of the Fick Principle. Assumptions required for calculation of blood oxygen contents include constant hemoglobin concentration (Hgb) and constant or negligible physically dissolved oxygen. First, the influence of these assumptions on continuous Qt and DO2 determinations was tested. Unmeasured changes in Hgb resulted in substantial error in calculated Qt, whereas calculated DO2 was minimally affected. Both Qt and DO2 were little altered by errors in PaO2 or PVO2. Second, the effects of SVO2, SaO2, and VO2 measurement errors on Qt and DO2 calculations under normal and extreme conditions were quantified. Relative errors in SVO2 were increased by a factor of 4.2 in Qt estimations and by a factor of 3.2 in DO2 measurements under normal conditions. These factors increased with increasing SVO2, and thus, Qt and DO2 became increasingly unreliable as oxygen extraction fell. Third, we compared continuous measurements of Qt and DO2 with intermittent measurements made by thermodilution cardiac output and blood sampling, and found correlation coefficients of 0.85 for Qt and 0.89 for DO2. Fourth, common measurement errors in VO2 and DO2 calculated in this way were found to bias regressions between VO2 and DO2, and this bias could be minimized only if the DO2 range were high and SVO2 were low.(ABSTRACT TRUNCATED AT 250 WORDS)
