Non-invasive measurement of cardiac output using an iterative, respiration-based method

Current non-invasive respiratory-based methods of measuring cardiac output (Q˙) make doubtful assumptions and encounter significant technical difficulties. We present a new method using an iterative approach (Q˙IT), which overcomes limitations of previous methods. Sequential gas delivery (SGD) is used to control alveolar ventilation (V˙A) and CO2 elimination (V˙CO2) during a continuous series of iterative tests. Each test consists of four breaths where inspired CO2(PICO2) is controlled; raising end-tidal Pco2(PE′CO2) by about 1.33 kPa (10 mm Hg) for the first breath, and then maintaining PE′CO2 constant for the next three breaths. The PICO2 required to maintain PE′CO2 constant is calculated using the differential Fick equation (DFE), where Q˙ is the only unknown and is arbitrarily assumed for the first iteration. Each subsequent iteration generates measures used for calculating Q˙ by the DFE, refining the assumption of Q˙ for the next test and converging it to the true Q˙ when PE′CO2 remains constant during the four test breaths. We compared Q˙IT with Q˙ measured by bolus pulmonary artery thermodilution (Q˙TD) in seven pigs undergoing liver transplantation. Q˙IT implementation and analysis was fully automated, and Q˙TD varied from 0.6 to 5.4 litre min−1 through the experiments. The bias (between Q˙IT and Q˙TD) was 0.2 litre min−1 with 95% limit of agreement from −1.1 to 0.7 litre min−1 and percentage of error of 32%. During acute changes of Q˙, convergence of Q˙IT to actual Q˙ required only three subsequent iterations. Q˙IT measurement is capable of providing an automated semi-continuous non-invasive measure of Q˙.