Objective: To assess the reliability of estimation of cardiac index based on the mixed venous oxygen saturation and methods of improving the estimation of cardiac index.
Setting: PICU in an university hospital.
Design: In the post-operative period following complete repair of congenital heart disease we carried out 55 measurements of blood gases in 25 infants and children (mean age 16.1 months, mean body surface 0.43 m2) from a systemic artery (arterial) and the pulmonary artery (mixed venous). We also determined the pulmonary oxygen uptake and calculated the cardiac index (CI) using Fick's principle. In the analysis we compared the CI with the mixed venous oxygen saturation (SvO2) and with the quotient of the arterial oxygen content (CaO2) and the oxygen extraction (Ca-vDO2). This quotient is equal to arterial oxygen delivery (DO2) divided by the oxygen consumption (VO2).
Results: Pearson's correlation coefficient was 0.77 when SvO2 was compared to CI in a linear regression model. Assuming an inverse relationship between SvO2 and CI the correlation was much better (r = 0.90). However, the best estimation of CI provides the quotient CaO2/Ca-vDO2 (r = 0.93).
Conclusions: CaO2/Ca-vDO2 correlates much better with CI than the SvO2, therefore CI could be better estimated based on CaO2/Ca-vDO2. Furthermore CaO2/Ca-vDO2 provides good information about the oxygen supply situation of the body.