Background: Cerebral oximetry is a noninvasive optical technology that measures frontal cortex blood hemoglobin-oxygen saturation. Commercially available cerebral oximeters have not been evaluated independently. Unlike pulse oximeters, there are currently no Food and Drug Administration standards for performance or accuracy. We tested the hypothesis that cerebral oximeters accurately measure a fixed ratio of the oxygen saturation in cerebral mixed venous and arterial blood.
Methods: We evaluated the performance of 5 commercially available cerebral oximeters: the EQUANOX® 7600 in 3- and 4-wavelength versions (Nonin Medical, Plymouth, MN), FORE-SIGHT® (Casmed, Branford, CT), INVOS® 5100C (Covidien, Boulder, CO), and the NIRO-200NX® (Hamamatsu Photonics, Hamamatsu City, Japan) during stable isocapnic hypoxia in volunteers. Twenty-three healthy adults (14 men, 9 women) had sensors placed on each side of the forehead. The subject's inspired oxygen (FIO2) was then changed to produce 6 steady-state arterial oxygen saturation (SaO2) levels between 100% and 70%, while end-tidal CO2 was maintained constant. At each plateau, simultaneous blood samples from the jugular bulb and radial artery were analyzed with a hemoximeter (OSM-3, Radiometer Medical A/S, Copenhagen, Denmark). Each cerebral oximeter's bias was calculated as the difference between the instrument's reading (cerebral saturation, ScO2) with the weighted saturation of venous and arterial blood (Sa/vO2), as specified by each manufacturer (INVOS: 25% arterial/75% venous; FORE-SIGHT, EQUANOX, and NIRO: 30% arterial/70% venous).
Results: Five hundred forty-two comparisons between paired blood samples and oximeter readings were analyzed. The pooled root mean square error was 8.06%, a value higher than for pulse oximeters, which is ±3% by Food and Drug Administration standards. The mean % bias ± SD (precision) and root mean square errors were: FORE-SIGHT 1.76 ± 3.92 and 4.28; INVOS 0.05 ± 9.72 and 9.69; NIRO-200NX -1.13 ± 9.64 and 9.68; EQUANOX-3 λ 2.48 ± 8.12 and 8.47; EQUANOX-4 λ 2.84 ± 6.27 and 6.86. The FORE-SIGHT, NIRO-200NX, and EQUANOX-3 λ had significantly more positive bias at lower SaO2. The amount of bias during hypoxia was reduced when the bias was calculated on the basis of difference between oximeter reading and the arterial and mixed venous saturation difference rather than the weighted average of blood saturation, indicating that differences in the ratio between arterial and venous blood volumes account for some of the positive bias at low saturation. Dark skin pigment tended to produce more negative bias in all instruments but bias was significantly larger than zero only for the FORE-SIGHT oximeter. Bias was significantly more negative in women for INVOS and EQUANOX devices but not for the FORE-SIGHT device.
Conclusions: While responsive to desaturation, cerebral oximeters exhibited large variation in reading errors between subjects, with mean bias possibly related to variations in the ratio of arterial and venous blood in the sampling area of the brain. This ratio is probably not fixed, as assumed by the manufacturers, but dynamically changes with hypoxia. Better understanding these factors could improve the performance of cerebral oximeters and help establish saturation or blood flow thresholds for brain well-being.