ACCURACY OF 4 OXIMETRY DEVICES FOR CONTINUOUS MONITORING OR MIXED-VENOUS OXYGEN-SATURATION DURING ABDOMINAL AORTIC-SURGERY

Several systems for mixed-venous oximetry are now available. There are one three-wave-length system (Abbott) and three two-wave-length systems with (Spectramed) and without automatic correction for hemoglobin or hematocrit (Edwards). The purpose of this prospective randomized study was to compare the different systems and to examine the accuracy of continuous mixed-venous oximetry during abdominal aortic surgery. Eighty patients had a radial artery cannula and one of the following fiberoptic pulmonary artery catheters inserted before induction of anesthesia: Swan-GanZ(TM) oximetry TD catheter (Edwards), Swan-GanZ(TM) flow-directed oximetry thermodilution paceport catheter (Baxter, Edwards Division), SpectraCath STP(TM) (Spectramed), and Opticath(TM) (Abbott), Mixed-venous O2 saturation was monitored by oximetry computers: SAT-1 (Edwards), SAT-2 (Baxter, Edwards Division), Hemopro2 (Spectramed), and Oximetrix 3 (Abbott). As a method of reference, mixed-venous blood samples were drawn and immediately analyzed by an OSM3-Hemoximeter. Data sets were obtained at eight predetermined times. Hemoglobin was kept constant at +/- 1 g.dl-1. Continuous oximetry in comparison to in-vitro measurements yielded a correlation coefficient of r = 0.873 (P less-than-or-equal-to 0.0001) and a value of bias and precision (b +/- p) of -0.9 +/- 2.6% for the SAT-1, r = 0.815 (P less-than-or-equal-to 0.0001) and b +/- p = -2.2 +/- 2.5% for the SAT-2, r = 0.901 (P less-than-or-equal-to 0.0001) and b +/- p = 0.35 +/- 2.5% for the Hemopro2, and r = 0.920 (P less-than-or-equal-to 0. 0001) and b +/- p = 0.1 +/- 1.8% for the Oximetrix 3, respectively. Significant differences between the catheters were found concerning r (P less-than-or-equal-to 0.01) and bias (P less-than-or-equal-to 0.05): the accuracy of the Oximetrix 3 was superior to both SATs and that of the Hemopro2 was superior to the SAT-2. Precision was best in the Oximetrix 3, but was not significantly different between the groups. Our findings demonstrate that mixed-venous oximetry is a reliable on-line method of measuring SvBARO2. Our data suggest that the Oximetrix 3 and Hemopro2 are more accurate than the SAT-2, possibly because the Oximetrix 3 is a three-wave-length system and the Hemopro2 is a two-wave-length but three-optical-fiber system in comparison to the SAT, which is a two-wave-length, two-optical-fiber system. Since bias and not precision was significantly different between the groups, it might also be that another algorithm in both SATs, especially the SAT-2, could correct the differences in the presence of stable hemoglobin levels.