Continuous measurement of cerebral oxygen saturation (rSO₂) for assessment of cardiovascular status during hemorrhagic shock in a swine model.

Abstract

Early trauma care is dependent on subjective assessments and sporadic vital sign assessments. We hypothesized that near-infrared spectroscopy-measured cerebral oxygenation (regional oxygen saturation [rSO₂]) would provide a tool to detect cardiovascular compromise during active hemorrhage. We compared rSO₂ with invasively measured mixed venous oxygen saturation (SvO₂), mean arterial pressure (MAP), cardiac output, heart rate, and calculated pulse pressure. Six propofol-anesthetized instrumented swine were subjected to a fixed-rate hemorrhage until cardiovascular collapse. rSO₂ was monitored with noninvasively measured cerebral oximetry; SvO₂ was measured with a fiber optic pulmonary arterial catheter. As an assessment of the time responsiveness of each variable, we recorded minutes from start of the hemorrhage for each variable achieving a 5%, 10%, 15%, and 20% change compared with baseline. Mean time to cardiovascular collapse was 35 minutes ± 11 minutes (54 ± 17% total blood volume). Cerebral rSO₂ began a steady decline at an average MAP of 78 mm Hg ± 17 mm Hg, well above the expected autoregulatory threshold of cerebral blood flow. The 5%, 10%, and 15% decreases in rSO₂ during hemorrhage occurred at a similar times to SvO₂, but rSO₂ lagged 6 minutes behind the equivalent percentage decreases in MAP. There was a higher correlation between rSO₂ versus MAP (R² =0.72) than SvO₂ versus MAP (R² =0.55). Near-infrared spectroscopy-measured rSO₂ provided reproducible decreases during hemorrhage that were similar in time course to invasively measured cardiac output and SvO₂ but delayed 5 to 9 minutes compared with MAP and pulse pressure. rSO₂ may provide an earlier warning of worsening hemorrhagic shock for prompt interventions in patients with trauma when continuous arterial BP measurements are unavailable.