Abstract 371: Physiological Predictors of Cerebral Oxygenation During Hemorrhage in Swine

Introduction: Measurements of cerebral oxygenation (rSO ₂ ) using near-infrared spectroscopy (NIRS) have been used to predict survival in retrospective clinical studies. However, the physiology underlying changes in rSO ₂ during critical states such as hemorrhage is not completely understood. This analysis examined the effect of blood pressures, intracranial pressure, and heart rate on rSO ₂ .

Methods: Eight swine were instrumented with invasive pressure catheters in the aorta, right atrium, and cerebrum and underwent a pressure-targeted arterial hemorrhage protocol. Animals were bled at a rate of 20mL/min to a target diastolic pressure of 35 mmHg. Brief periods of increased baseline airway pressure (15 cm H ₂ O) were introduced to cause transient drops in arterial blood pressure. The aortic pressure waveform was used to identify individual heartbeats. The first 100 heartbeats represent baseline data, while heart beats 4900-5000 (time from start of bleed; 40±7 min) capture data late into the bleed for all animals. Data from 53,315 heart beats were entered into a regression model: rSO ₂ = mean diastolic aortic pressure + mean diastolic right atrial pressure + intracranial pressure + heart rate + animal.

Results: The hemorrhage reduced diastolic aortic pressure from 61.7±9.6 (heart beats 1-100) to 35.0±9.2 mmHg (heart beats 4900-5000), increased right atrial pressure from 4.1±3.6 to 4.9±6.9 mmHg, reduced intracranial pressure from 33.6±23.2 mmHg to 16.3±2.5 mmHg, and increased heart rate from 101±14 to 144±40 beats/min. Output from the regression model (Table 1) shows that mean diastolic aortic pressure, but not right atrial pressure, intracranial pressure or heart rate, is a significant predictor of rSO ₂ .

Conclusions: rSO ₂ values are largely driven by arterial blood pressure and may serve as a useful non-invasive tool as a surrogate for arterial blood pressure during hemorrhage. Further work should investigate the use of this technology to provide real-time feedback.