Cardiac output during exercise: A comparison of four methods

Several techniques assessing cardiac output (Q) during exercise are available. The extent to which the measurements obtained from each respective technique compares to one another, however, is unclear. We quantified Q simultaneously using four methods: the Fick method with blood obtained from the right atrium (Q(Fick-M)), Innocor (inert gas rebreathing; Q(Inn)), Physioflow (impedance cardiography; Q(Phys)), and Nexfin (pulse contour analysis; Q(Pulse)) in 12 male subjects during incremental cycling exercise to exhaustion in normoxia and hypoxia (FiO2=12%). While all four methods reported a progressive increase in Q with exercise intensity, the slopes of the Q/oxygen uptake (VO2) relationship differed by up to 50% between methods in both normoxia [4.9 +/- 0.3, 3.9 +/- 0.2, 6.0 +/- 0.4, 4.8 +/- 0.2L/min per L/min (mean +/- SE) for Q(Fick-M), Q(Inn), Q(Phys) and Q(Pulse), respectively; P=0.001] and hypoxia (7.2 +/- 0.7, 4.9 +/- 0.5, 6.4 +/- 0.8 and 5.1 +/- 0.4L/min per L/min; P=0.04). In hypoxia, the increase in the Q/VO2 slope was not detected by Nexfin. In normoxia, Q increases by 5-6L/min per L/min increase in VO2, which is within the 95% confidence interval of the Q/VO2 slopes determined by the modified Fick method, Physioflow, and Nexfin apparatus while Innocor provided a lower value, potentially reflecting recirculation of the test gas into the pulmonary circulation. Thus, determination of Q during exercise depends significantly on the applied method.