A comparison of formulas used to estimate mixed venous saturations

True mixing of venous blood in the absence of shunt occurs in the pulmonary artery. In the presence of left to right shunt at a level proximal to the pulmonary artery, mixed venous blood for oxygen saturation (MVO2) is estimated by using an average of blood samples taken from the chamber proximal to the shunt. In atrial septal defect, the determination of MVO2 is calculated by using blood samples from the superior vena cava (SVC) and the inferior vena cava (IVC). Several formulas have been proposed, utilizing varying combinations of blood samples taken from the SVC and IVC. In the present investigation, 100 patients without evidence of shunt were studied during routine cardiac catheterization. Duplicate blood samples were taken from the pulmonary artery (PA), the SVC, and the IVC, and were analyzed for oxygen-saturation. If one assumes that the PA blood sample represents true venous blood mixing (TMVO2), the following formulas were used for comparison: 1)PA = SVC; 2) PA = IVC; 3) PA = (SVC + IVC)/2; 4) PA = (2SVC + IVC)/3; 5) PA = (3SVC + IVC)/4; and 6) PA = (2IVC + SVC)/3. When one uses the standard two variable regression equations, this study shows that the 90% confidence limits are wide. The correlation, however, is somewhat better if one uses the formulas 3)-6). Therefore, the error that may be introduced in calculating the TMVO2 may be substantial and can critically alter the estimation of the shunted blood volume.