In Vivo Comparison of Two Mixed Venous Saturation Catheters

The accuracy and stability of mixed venous saturation pulmonary arterial catheters under adverse physiologic conditions has not been assessed. Either a Shaw Opticath catheter (three-wavelength) or a Swan-Ganz oximetry TD catheter (two-wavelength) was calibrated in vitro and positioned in the pulmonary artery in each of ten mongrel dogs. The in vivo saturations were compared to measured saturations from anaerobically collected mixed venous blood analyzed with a reference cooximeter at each step in the protocol. Oxygen delivery was varied to obtain a broad range of mixed venous saturations (Sv-O2) by altering inspired oxygen concentration, isovolemic hemodilution, reducing cardiac output, and increasing afterload. Calculated oxygen consumption varied from 128 to 311 ml/min. Pre-insertion calibration for both catheter types compared favorably with the cooximeter prior to physiologic manipulations, although the three-wavelength catheter more closely approximated the cooximeter. The three-wavelength catheter tracked measured Sv-O2 accurately under adverse conditions for up to 10 h (R = .994; SEE = 2.2%). The two-wavelength catheter tended to drift under the same conditions (R = .808; SEE = 10.6%). At the conclusion of the experiment, the two-wavelength system was uniformly higher then the cooximeter by 5-31% with a mean of 21% (P less than or equal to .003 as compared with the initial difference by paired Student's t test). Pending further analysis of the tendency of the two-wavelength system to drift it would seem prudent to limit its clinical application.