The use of pulse oximetry to detect congenital heart disease

Objective To evaluate whether pulse oximetry can be used as a screening test to detect congenital heart disease (CHD) in otherwise well newborns before nursery discharge. Methods Births at Lakeland Regional Medical Center (LRMC) who met inclusion criteria, had a single pulse oximeter reading before discharge. A normal reading was considered to be >94%. Abnormal readings were repeated on a different extremity and if still abnormal, an echocardiography was performed. The number of children with CHD and the number of echocardiograms were compared with the previous year at LRMC. Results There were 2114 infants enrolled, 88 echocardiograms (3.8%) were performed, and 43 were abnormal (yield = 48.9%), of which 12 required management (13.6%). The control group consisted of 2851 births, producing 108 echocardiograms (3.8%), 42 of which were abnormal (38.9%), and 13 required management (12.0%). No comparisons were significant. During both years, 3 children with cyanotic CHD were born. Routine pulse oximetry failed to detect a child with total anomalous pulmonary venous return. Not a single normal echocardiogram was generated by the study. Conclusion Routine pulse oximetry was nearly 100% specific for detecting cyanotic CHD; as a result there was no increase in the number of echocardiograms. Routine pulse oximetry did detect a child who might not have been detected otherwise but failed to detect one that should have been detected. (J Pediatr 2003;142:268-72)