Correlation and agreement of regional cerebral oxygen saturation measured from sensor sites at frontal and temporal areas in adult patients undergoing cardiovascular anesthesia

Abstract

Background The function and viability of the brain depend on adequate oxygen supply. A decrease in cerebral blood supply causing cerebral desaturation may lead to many neurological complications. Direct measurement of regional cerebral oxygen saturation (rScO2) assists in early detection and management. Near-infrared spectroscopy (NIRS) has been introduced for measuring rScO2. A pair of sensors are attached to the right and left forehead. However, there are some situations where the forehead of the patient is not accessible for sensor attachment (e.g., neurosurgery involving the frontal area; a bispectral index (BIS) sensor already attached, or a wound to the forehead); therefore, alternate sites for sensor attachment are required. The temporal area was proposed as an alternate site. The objective of this study was to assess the correlation and agreement of rScO2 measured at the forehead vs. the temporal area. Methods Adult patients undergoing cardiothoracic or vascular surgery were monitored for rScO2 using two pairs of ForeSight sensors. The first pair (A1 and A2) were attached to the right and left forehead, while the second pair (B1 and B2) were attached to the right and left temporal area. The rScO2 values measured from A1 vs. B1 and A2 vs. B2 were assessed for correlation and agreement using the Bland-Altman analysis. Results Data from 19 patients with 14,364 sets of data were analyzed. The data from A1 vs. B1 and A2 vs. B2 showed moderate positive correlation (r = 0.627; P < 0.0001 and r = 0.548; P < 0.0001). The biases of A1 vs. B1 and A2 vs. B2 were −2.3% (95% CI [−2.5 to −2.2]; P < 0.0001) and 0.7% (95% CI [0.6–0.8]; P < 0.0001). The lower and upper limits of agreement of A1 vs. B1 were −17.5% (95% CI [−17.7 to −17.3]) and 12.8% (95% CI [12.6–13.0]). The lower and upper limits of agreement of A2 vs. B2 were −14.6% (95% CI [−14.8 to −14.4]) and 16.0% (95% [CI 15.8–16.3]). Conclusions The rScO2 values measured from sensors at the frontal and temporal areas show a moderate correlation with sufficiently good agreement. The temporal area may be an alternative to the frontal area for cerebral oximetry monitoring.