A new method of using gas exchange measurements for the noninvasive determination of cardiac output: clinical experiences in adults following cardiac surgery.

New mathematical algorithms have been applied to a computer controlled closed breathing circuit system for non-invasive measurement of cardiac output (COniv). This system has been described in an animal study. Forty patients were studied 5 and 18 hours after cardiac surgery using the thermodilution technique as the reference (COtd). The variables entered into the algorithms for COniv were oxygen uptake, carbon dioxide elimination, end-tidal carbon dioxide partial pressure, tidal volume and arterial oxygen saturation. Mixed venous carbon dioxide partial pressure was obtained from an automatically implemented short rebreathing manoeuvre. Pulmonary perfusion was calculated by a modified Fick equation for carbon dioxide and the shunt flow added to obtain COniv. During mechanical ventilation, there was a good agreement between COtd and COniv (r = 0.8). The bias was -0.14 l/min and the precision was 0.77 l/min. The reproducibility of COniv was 0.03 l/min and for COtd -0.03 l/min with a standard deviation of the difference being 0.35 l/min for COniv and 0.31 l/min for COtd. In awake, but sedated extubated patients, the method proved unsatisfactory on account for uneven tidal volumes and difficulties with leakage around the mouth piece. We conclude that this new technique provides reliable and reproducible measures of cardiac output in sedated, ventilated patients.