A new method for the non-invasive measurement of cardiac output during spontaneous ventilation

Introduction. All respiratory-based methods of measuring cardiac output depend on imposing a step-change in alveolar ventilation (V˙ A) for a particular gas. For CO2, this can be accomplished by interposing respiratory deadspace in the breathing circuit, but therefore requires a uniform breathing pattern. I developed a method of reducing V˙ A that is independent of breathing pattern using a new breathing circuit. I evaluated this method's accuracy by comparing its measurements (QNI ) with those of thermodilution (QTD) on cardiac patients in the ICU, as well as its precision by performing repeated tests on subjects in a laboratory. Methods. I conducted 3 validation studies involving 89 post cardiac surgery patients in the ICU after extubation. The reduction in V˙A was induced by a step reduction in gas flow to the circuit. The balance of ventilation was composed of previously exhaled gas and no respiratory manoeuvre was required by the patient. QTD was measured simultaneously. I also conducted measurements to evaluate precision by performing 8 repeated tests on 24 subjects in a laboratory setting. Results. In my final validation study, QNI was highly correlated with QTD (R = 0.90) despite a wide range of breathing patterns. The bias and precision of the non-invasive method were 0.08 and 0.80 L/min, respectively. The repeatability of QNI, measured as the SD of repeated tests, was 0.73 L/min, which is slightly less that the repeatability of QTD, which is usually reported as 0.60 L/min. Conclusion. I have developed and tested a new method for measuring cardiac output noninvasively in spontaneously breathing subjects. Although measures to improve accuracy are still needed, our method correlated well with thermodilution and may eventually provide physicians with a new, clinically useful, non-invasive test.Ph.D