To assess cardiac output, the velocity of blood flowing in the aorta can be measured via ultrasound. A small transducer is inserted through a patients nose or mouth into the esoph- agus, where it is directed toward and along the descending aorta. Once there, blood velocity in the vessel can be measured via the Doppler shift. Knowing the area of the aorta, the vol- umetric flow can be calculated from the distribution of velocities, and thus the cardiac out- put determined. (Stroke volume can itself be determined by integrating the volumetric flow over time from beat to beat.) Studies in mechanically ventilated patients have demonstrated the reliability of this technique (68,69). Like stroke volume and pulse pressure variation, the variations in aortic blood flow resulting from respiratory induced pressure changes have been shown to be strongly predictive of fluid responsiveness [50, 67].
Although this technique has proven useful, it has not been broadly implemented. This is likely due to many confounding factors serving as obstacles to its adoption including the steep learning curve needed to obtain and interpret the signal of interest, the need for sedated mechanically ventilated patients, the inability to obtain continuous measurements reliably, and the actual obstacle of having to deal with a probe in the patients esophagus for the duration of the procedure. Until this conspiracy of mitigating factors is overcome, such a technique, though useful, will not enjoy extensive use.