Evaluation of pediatric near-infrared cerebral oximeter for cardiac disease.

ABSTRACT

BACKGROUND: Cerebral hypoxia-ischemia remains a complication in children with congenital heart disease. Near-infrared spectroscopy can be utilized at the bedside to detect cerebral hypoxia-ischemia. This study aimed to calibrate and validate an advanced technology near-infrared cerebral oximeter for use in children with congenital heart disease. METHODS: After institutional review board approval and parental consent, 100 children less than 12 years and less than 40 kg were enrolled. Phase I (calibration) measured arterial and jugular venous saturation (SaO(2), SjO(2)) by co-oximetry simultaneously with device signals to calibrate an algorithm to determine regional cerebral saturation against a weighted average cerebral saturation (0.7 SjO(2) + 0.3 SaO(2)). Phase II (validation) evaluated regional cerebral saturation from the algorithm against the weighted average cerebral saturation by correlation, bias, precision, and A(Root Mean Square) assessed by linear regression and Bland-Altman analysis. RESULTS: Of 100 patients, 86 were evaluable consisting of 7 neonates, 44 infants, and 35 children of whom 55% were female, 79% Caucasian, and 41% with cyanotic disease. The SaO(2) and regional cerebral saturation ranged from 34% to 100% and 34% to 91%, respectively. There were no significant differences in subject characteristics between phases. For the entire cohort, A(RMS), bias, precision, and correlation coefficient were 5.4%, 0.5%, 5.39%, and 0.88, respectively. Age, skin color, and hematocrit did not affect these values. CONCLUSIONS: This cerebral oximeter accurately measures the absolute value of cerebral saturation in children over a wide range of oxygenation and subject characteristics, offering advantages in assessment of cerebral hypoxia-ischemia in congenital heart disease.