Oximetric detection of intracardiac left-to-right shunts

Abstract

To assess prevailing guidelines for the oximetric detection of intracardiac left-to-right shunts, we examined the variability of right heart oxygen saturation step-ups in 1121 catheterisations in children with aortic and pulmonary stenosis, who were studied as part of the joint study of the natural history of congenital heart defects. It is estimated that in the absence of intracardiac shunting, a step-up of more than 5.9 per cent between the superior vena cava and the pulmonary artery, 3.5 per cent between the right atrium and the pulmonary artery, and 3.6 per cent between the right ventricle and pulmonary artery, occurs 5 per cent of the time. By extrapolation from a 'normal' distribution, a step-up of 8.7 per cent between the superior vena cava and pulmonary artery, 5.6 per cent between the right atrium and pulmonary artery, and 5.2 per cent between the right ventricle and the pulmonary artery would occur 1 per cent of the time in the absence of intracardiac shunting. Assuming a normal haemoglobin concentration of 15 g/dl, the 1 per cent false positive rate corresponds to a step-up of 1.7 volumes per cent from the superior vena cava to the pulmonary artery, 1.1 volumes per cent from the right atrium to the pulmonary artery, and 1.0 volumes per cent from the right ventricle to the pulmonary artery. These data provide guidelines for the detection of abnormal shunts, but other independent methods such as indicator dilution or angiography should be employed in borderline situations.