Relationship between mixed venous oxygen saturation and regional cerebral oxygenation in awake, spontaneously breathing cardiac surgery patients

In search of a non-invasive method for estimation of the oxygen balance this prospective study evaluates the relationship between regional cerebral oxygenation (rScO2) and mixed venous oxygen saturation in awake, spontaneously breathing patients after cardiac surgery.

After approval by the local ethical committee and written informed consent, 26 consecutive patients after cardiac surgery with cardiopulmonary bypass were enrolled. On intensive care unit (ICU), several hours after extubation, patients were connected to the INVOS 5100 cerebral monitor. Blood samples for determination of mixed venous oxygen saturation (SvO2) and central venous oxygen saturation (ScvO2) were drawn successively and rScO2 was documented. Patients were studied twice, breathing room air for the first measurement cycle, and breathing 4L/min supplemental oxygen by face mask, achieving a transcutaneous oxygen saturation above 98%, for a second measurement.

Hemodynamic variables, hemoglobin and arterial carbon dioxide tension (PaCO2) remained stable between the two measurement cycles. Without oxygen, SvO2 and rScO2 showed a bias of -2.0%, limits of agreement (LOA) of -15.0 to 10.9% and a percentage error (PE) of 20.3%. SvO2 and ScvO2 showed a bias of -3.9%, LOA of -13.9 to 6.2% and PE 15.7%. With oxygen, the bias between SvO2 and rScO2 was -2.5%, LOA -14.2 to 9.2%, PE 17.2%. Between SvO2 and ScvO2 the bias was -4.1%, LOA -10.2 to 2.1%, PE 9.0%.

The rScO2 measured by near infrared spectroscopy was sufficiently representing mixed venous oxygen saturation in awake, hemodynamically stable, spontaneously breathing patients after cardiac surgery. The agreement was comparable to the agreement between SvO2 and ScvO2 with smaller differences in the lower ranges of SvO2.