Objective: We studied the ability of a new pulse oximeter to detect hyperoxemia in neonates.
Methods: One hundred arterial blood samples were obtained from umbilical artery catheters of 41 acutely ill neonates (birth weight 560-4100 grams) while SpO2 was simultaneously measured at a postductal site. The relationship between PaO2 and SpO2 was examined and expressed on a matrix. Hyperoxemia was defined as PaO2 greater than 90 mm Hg and upper alarm limit for pulse oximetry was set at 95%. The incidence of true hyperoxemia (PaO2 greater than 90 mm Hg, SpO2 greater than 95%) was determined and sensitivity and specificity for detection of hyperoxemia by the SpaceLabs device was calculated.
Results: Hyperoxemia was detected in eighteen samples with sixteen having SpO2 greater than 95% (89% sensitivity). In 82 samples having a PaO2 value less than or equal to 90 mm Hg, 53 were associated with SpO2 less than or equal to 95% (65% specificity). Among 55 samples with SpO2 less than or equal to 95% hyperoxemia occurred in only 2 (3.6%). Sensitivity of 95% would have been achieved with upper alarm limit set at 94%.
Conclusion: The new pulse oximeter evaluated detected 89% of episodes of hyperoxemia with upper alarm limits set at 95% and 95% of episodes if upper limit was set at 94%. As previously reported, both individual device algorithms and upper alarm limits chosen determine ability of pulse oximeters to detect hyperoxemia.