Comparison of Minimally and More Invasive Methods of Determining Mixed Venous Oxygen Saturation

To investigate the accuracy of a minimally invasive, 2-step, lookup method for determining mixed venous oxygen saturation compared with conventional techniques.

Single-center, prospective, nonrandomized, pilot study.

Tertiary care hospital, university setting.

Thirteen elective cardiac and vascular surgery patients.

All participants received intra-arterial and pulmonary artery catheters. Minimally invasive oxygen consumption and cardiac output were measured using a metabolic module and lithium-calibrated arterial waveform analysis (LiDCO; LiDCO, London), respectively. For the minimally invasive method, Step 1 involved these minimally invasive measurements, and arterial oxygen content was entered into the Fick equation to calculate mixed venous oxygen content. Step 2 used an oxyhemoglobin curve spreadsheet to look up mixed venous oxygen saturation from the calculated mixed venous oxygen content. The conventional "invasive" technique used pulmonary artery intermittent thermodilution cardiac output, direct sampling of mixed venous and arterial blood, and the "reverse-Fick" method of calculating oxygen consumption.

LiDCO overestimated thermodilution cardiac output by 26%. Pulmonary artery catheter-derived oxygen consumption underestimated metabolic module measurements by 27%. Mixed venous oxygen saturation differed between techniques; the calculated values underestimated the direct measurements by between 12% to 26.3%, this difference being statistically significant.

The magnitude of the differences between the minimally invasive and invasive techniques was too great for the former to act as a surrogate of the latter and could adversely affect clinical decision making.