Pulse oximeter and transcutaneous arterial oxygen measurements in neonatal and paediatric intensive care.

Pulse oximeter (SaO2P) measurements were compared with direct arterial line oxygen saturation (SaO2) from co-oximeters in 92 instances in 43 patients, and with arterial line oxygen measurements (PaO2) in 169 instances in 81 patients. The mean (SD) absolute difference between SaO2P and SaO2 was 2.6% (2.4) after attempt to correct for the co-oximeter falsely measuring a proportion of fetal haemoglobin as carboxy haemoglobin. For 19 infants and children greater than or equal to 5 months old, who have very little fetal haemoglobin, the mean (SD) absolute difference of 27 comparisons was 1.8% (2.1). Comparison of SaO2P and PaO2 measurements in 46 instances when PaO2 was less than 6.67 kPa showed SaO2 to be less than 90% on 40 occasions. In 24 instances when PaO2 was greater than or equal to 13.3 kPa the SaO2P was greater than or equal to 98% on 22 occasions. In 23 infants undergoing neonatal intensive care, transcutaneous oxygen monitors were compared with arterial PO2 measurements in 60 instances. The mean (SD) absolute difference between PaO2 and transcutaneous oxygen measurements was 1.60 kPa (1.73). Ten of the 60 comparisons had differences greater than 2.67 kPa and three greater than 5.33 kPa (maximum 8.40 kPa). Pulse oximetry is a clinically useful technique for managing oxygenation but further studies are needed to confirm its safety in premature infants at risk of retinopathy of prematurity.