Reliability of in vivo mixed venous oximetry during experimental hypertriglyceridemia.

Abstract

The spectral absorbance of iv lipid emulsion produces interference in the in vitro spectrophotometric measurement of hemoglobin saturation. Therefore, we investigated in vivo mixed venous oximetry during lipid emulsion infusions. Boluses of lipid emulsion, increasing by 0.1-g/kg increments, were infused every 30 mins into nine anesthetized dogs. After each lipid bolus, laboratory and hemodynamic data measurements were repeated, and arterial and mixed venous blood gases were analyzed on a laboratory cooximeter that was equipped for canine blood. The in vivo mixed venous oxygen saturation was continuously monitored via an indwelling, optically equipped pulmonary artery catheter. As expected, increasing concentrations of lipid emulsion increased (r2 = .92) serum triglyceride concentration. As a result, the in vitro measurements of percentage methemoglobin increased artifactually, reducing in vitro measurements of arterial (r2 = .74) and mixed venous (r2 = .44) oxygen saturations. The in vivo mixed venous oxygen saturation, however, remained constant during hypertriglyceridemia (r2 less than .01). In vivo mixed venous oximetry remained constant during hypertriglyceridemia, while in vitro measurements of arterial and mixed venous saturations were artifactually reduced.