Cerebral hypoxia-ischemia remains a not infrequent compli- cation in children with congenital heart disease (CHD) before, during and after cardiothoracic interventions. Near-infrared spectroscopy (NIRS), a non-invasive optical technology, can be utilized at the bedside to monitor brain oxygenation and detect hypoxia-ischemia. The purpose of this multi-center study was to calibrate and validate a novel NIRS device for CHD.
The study sensor, EQUANOXTM AdvanceTM 8004CB sensor (adhesive and non-adhesive), utilizes a dual-emitter/dual-detec- tor sensor architecture which has been shown to more effec- tively eliminate signal contamination due to surface and shal- low tissue variability to better isolate the targeted tissue bed.1 Spacing between the emitter/detector components is 12.5 mm, thus the distance from each emitter to the nearest detector is 12.5 mm and the distance from each emitter to the farther detec- tor is 25.0 mm. (FIGURE 1). Rather than require multiple sen- sors or including age into the algorithm, EQUANOX Advance 8004CB incorporates Nonin’s proprietary Dynamic Compensa- tion algorithm to account for the variability in light absorption characteristics of the developing brain tissue of the child. This methodology allows for adjustment based on each individual subject’s development rather than relying on an age-based algo- rithm representing an excessively broad-ranged group.
FiGure 1: Equinox Advance 8004cB sensor
After parental consent and IRB approval, 100 children aged <12 years and <40 kg with CHD undergoing cardiac catheter- ization were enrolled into a two phase study. In Phase I (cali- bration phase), arterial and jugular bulb venous blood were obtained for co-oximetry simultaneously with NIRS signals to calibrate an algorithm for NIRS regional cerebral saturation
(rScO2) to a weighted average cerebral saturation (SavO2= 0.7 SjO2 + 0.3 SaO2). In Phase II, (validation phase), arterial and jugular bulb venous samples to calculate SavO2 were compared with NIRS rScO2 from the device algorithm. Correlation, bias, precision, and Arms between SavO2 and rScO2 were determined by linear regression and Bland-Altman analysis.
Of 100 subjects, 86 met criteria for inclusion in the analysis. There were 7 neonates, 44 infants, and 35 children; of these, 54.7 % were female, 79.1% Caucasian, and 40.7% with cya- notic disease. SaO2 and rScO2 ranged from 34.4 to 100% and 34 to 92%, respectively. There were no differences in these data between calibration and validation phases. The Arms, bias, preci- sion, and correlation coefficient were 5.4%, 0.5±5.38%, 5.39%, and 0.88 respectively. (FIGURE 2) Age, skin color, and hema- tocrit did not affect these values.
Figure 2: rso2 versus savo2
The Nonin EQUANOX Regional Oximeter System with EQUANOX Advance 8004CB sensor accurately measures the absolute value of cerebral saturation in children over a wide range of oxygenation levels, and may offer advantages in the detection of tissue hypoxia-ischemia in CHD over other devices.
1 Davie S, Grocott H. The impact of extra cranial contamination on regional cerebral oxygen saturation: a comparison of three different cerebral oximeters. NIRS measurements of Society of Cardiovascular Anesthesiologists 33rd Annual Meeting. Savannah, GA. May 3, 2011. Poster session.