Abstract
Mixed venous oxygen saturation (SvO(2)) is an important variable in anesthesia and intensive care but currently requires pulmonary artery catheterization. Recently, non-invasive determination of SvO(2) (Capno-SvO(2)) using capnodynamics has shown good agreement against CO-oximetry in an animal model of modest hemodynamic changes. The purpose of the current study was to validate Capno-SvO(2) against CO-oximetry during major alterations in oxygen delivery. Furthermore, evaluating fiberoptic SvO(2) for its response to the same challenges. Eleven mechanically ventilated pigs were exposed to oxygen delivery changes: increased inhaled oxygen concentration, hemorrhage, crystalloid and blood transfusion, preload reduction and dobutamine infusion. Capno-SvO(2) and fiberoptic SvO(2) recordings were made in parallel with CO-oximetry. Respiratory quotient, needed for capnodynamic SvO(2), was measured by analysis of mixed expired gases. Agreement of absolute values between CO-oximetry and Capno-SvO(2) and fiberoptic SvO(2) respectively, was assessed using Bland-Altman plots. Ability of Capno- SvO(2) and fiberoptic SvO(2) to detect change compared to CO-oximetry was assessed using concordance analysis. The interventions caused significant hemodynamic variations. Bias between Capno-SvO(2) and CO-oximetry was + 3% points (95% limits of agreements - 7 to + 13). Bias between fiberoptic SvO(2) and CO-oximetry was + 1% point, (95% limits of agreements - 7 to + 9). Concordance rate for Capno-SvO(2) and fiberoptic SvO(2) vs. CO-oximetry was 98% and 93%, respectively. Capno-SvO(2) generates absolute values close to CO-oximetry. The performance of Capno-SvO(2) vs. CO-oximetry was comparable to the performance of fiberoptic SvO(2) vs. CO-oximetry. Capno-SvO(2) appears to be a promising tool for non-invasive SvO(2) monitoring.