Regional haemoglobin oxygen saturation during surgical haemorrhage

Experimental research and early clinical studies have shown that regional haemoglobin desaturation, measured by near infrared spectroscopy (NIRS), follows blood loss. To further assess NIRS as a blood loss monitor, we evaluated it on patients undergoing major surgery.

Regional haemoglobin oxygen saturation from the cerebral cortex (CsO(2)) and the left gastrocnemius muscle (PsO(2)) was continuously monitored with an INVOS 4100 NIRS oximeter in 10 patients during surgery. Haemoglobin, mean arterial pressure, arterial and central venous oxygen saturation and tension, temperature, pH and lactate concentration were also measured.

The median (IQR) blood loss was 650 (400-1 800) ml, equivalent to 16% (11-35) of the patients' blood volume. During surgery, CsO(2) fell by a mean (95% CI) of 8.4% (2-14.8) (p=0.016) and PsO(2) fell by a mean (95% CI) of 3.3% (-1.5-8.1) (p=0.16). CsO(2), but not PsO(2), correlated with the volume of blood lost (R=0.44; p=0.004), central venous oxygen saturation (R=0.52; p=0.001), and venous oxygen tension (R=0.58; p<0.001). These correlations remained significant after calculations were adjusted for changes in arterial pressure, temperature and pH.

Cerebral, but not peripheral, haemoglobin oxygen saturation decreases proportionally to blood loss and correlates with measurements of systemic oxygen extraction. With further research, NIRS measurements of CsO(2) may be developed into a useful tool to monitor blood loss.