Pulse oximetry advantages in infants with bronchopulmonary dysplasia.

We studied 12 infants with a clinical and radiologic diagnosis of bronchopulmonary dysplasia who were oxygen dependent and older than 30 days. Simultaneous readings of hemoglobin oxygen saturation (SaO2) determined by two pulse oximeters (Nellcor 100, BTI Biox III) and transcutaneous (tc) PO2 (Sensor Medics, Transend) were correlated with SaO2 (Radiometer, OSM 2 Hemoximeter) and PaO2 (Corning 178) measured on blood from an indwelling arterial catheter. For each infant, the fractional inspiratory oxygen (FiO2) was adjusted to obtain three to five sets of data in the range of 70% to 95% SaO2. Fifty-three data points were generated and pooled for analysis. The slope of the regression line generated for the Nellcor 100 was .86; for the BTI Biox III, it was .91; and for the Sensor Medics Transend, it was .55, resulting in average errors of +2.5%, +1.0%, and -29%, respectively, when comparing corresponding transcutaneous and arterial values. When SaO2 was equal to or less than 95%, no infants were hyperoxic. These data confirm reports by others that tcPO2 values do not accurately represent PaO2 values in older infants with bronchopulmonary dysplasia. Pulse oximeters do not require user calibration, and their sensor is unheated so they will not cause skin burns. We conclude that pulse oximetry offers major advantages over tcPO2 measurements in the management of infants with bronchopulmonary dysplasia.