Abstract Sun606: Assessment of Racial Bias in Cerebral Oximetry During Cardiac Arrest: An Ancillary Analysis of the AWARE II Study

Abstract

Background: Cerebral oximetry is a tool to assess the quality of resuscitation and to predict the return of ROSC and survival with neurological outcomes. The use of near-infrared spectroscopy (NIRS) allows for continuous monitoring of cerebral oxygen saturation. Recent studies highlight potential racial disparities due to the effect of skin pigmentation in pulse oximetry, leading to possible misrepresentation of oxygen saturation in darker individuals, raising concerns about accuracy and nihilism. NIRS targets deeper cerebral tissues but may still be impacted by skin melanin. It is unknown if similar biases exist in cerebral oximetry. This study aimed to determine whether cerebral oximetry is impacted by skin pigmentation during cardiac arrest. Methods: We conducted an ancillary analysis of data from the AWAreness during REsuscitation (AWARE II) study, a multi-center prospective investigation of consciousness and brain oxygenation during cardiac arrest. Patients aged ≥18 years experiencing in-hospital cardiac arrest were included. Cerebral oximetry data were recorded using the Nonin EQUANOX™ Model 7600. 189 patients (136 Caucasians, 41 African Americans) met inclusion criteria. Continuous cerebral oximetry data were captured during resuscitation and a rank-sum test was used to compare average regional cerebral oxygen saturation (rSO2) between groups, including subanalyses by ROSC outcome. Results: We examined the average rSO2 among the patients. The average rSO2 was 47.25% in Caucasian patients and 50.27% in African American patients (p = 0.2) (Table 1). Among those achieving sustained ROSC (n = 68), average rSO2 was 53.88% for Caucasian and 56.55% for African American patients (p = 0.4) (Table 2). In patients without sustained ROSC (n = 109), the average rSO2 was 43.14% and 46.26% (p = 0.2) (Table 3). No statistically significant differences in rSO2 values were observed based on race. Discussion: This is the largest study to evaluate the impact of skin pigmentation on cerebral oximetry during cardiac arrest. There were no statistically significant differences in measured brain oxygenation between Caucasian and African American patients. These findings suggest cerebral oximetry may provide a reliable measure of brain oxygenation in diverse patient populations. Limitations include use of a single NIRS device and a specific clinical context (cardiac arrest); further studies across varied settings and technologies are warranted to confirm generalizability.