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

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.