Inaccuracies of oximetry in identifying the location of intracardiac left-to-right shunts in adults.

Although oximetric analysis of blood from the right heart chambers is the most commonly used method for assessing the presence of intracardiac left-to-right shunting, it sometimes provides misleading information because a patient with a left-to-right shunt in only 1 location sometimes manifests a significant oxygen step-up (1) in >1 chamber or (2) in a chamber other than that in which the shunt is actually located. This study was performed (1) to assess the frequency with which oximetric data provide such misleading results and (2) to determine which variables (if any) may contribute to the occurrence of such erroneous oximetric results. Accordingly, we analyzed oximetric data from 168 patients (61 men, 107 women, 14 to 76 years of age) with a proved left-to-right shunt at only 1 site and oximetric evidence of significant oxygen step-up. Using the criteria of Dexter et al (J Clin Invest 1947;26:554-560), Antman et al (Am J Cardiol 1980;46:265-271), or Pirwitz et al (Am Heart J 1997;133:413-417) for a significant oxygen step-up, 34%, 42%, and 35% of subjects, respectively, manifested a significant step-up in >1 chamber or a step-up only in an incorrect chamber. Compared with those with a step-up in the correct chamber only, those with a step-up in >1 chamber had larger Qp/Qs ratios (mean +/- SD, 2.7 +/- 1.2 vs 3.5 +/- 1.8, respectively, p <0.001), whereas subjects with a step-up only in an incorrect chamber had smaller Qp/Qs ratios than those with a step-up in the correct chamber (2.2 +/- 0.9 vs 2.8 +/- 1.3, respectively, p <0.001). In conclusion, in adult subjects with left-to-right intracardiac shunting in only 1 location, carefully obtained oximetric data often provide erroneous information.