Noninvasive cardiac output measurement using partial carbon dioxide rebreathing is less accurate at settings of reduced minute ventilation and when spontaneous breathing is present.

BACKGROUND Although evaluation of cardiac output by the partial carbon dioxide rebreathing technique is as accurate as thermodilution techniques under controlled mechanical ventilation, it is less accurate at low tidal volume. It is not clear whether reduced accuracy is due to low tidal volume or low minute ventilation. The effect of spontaneous breathing on the accuracy of partial carbon dioxide rebreathing measurement has not been fully investigated. The objectives of the current study were to investigate whether tidal volume or minute ventilation is the dominant factor for the accuracy, and the accuracy of the technique when spontaneous breathing effort is present. METHODS The authors enrolled 25 post-cardiac surgery patients in two serial protocols. First, the authors applied three settings of controlled mechanical ventilation in random order: large tidal volume (12 ml/kg), the same minute ventilation with a small tidal volume (6 ml/kg), and 50% decreased minute ventilation with a small tidal volume (6 ml/kg). Second, when the patient recovered spontaneous breathing, the authors applied three conditions of partial ventilatory support in random order: synchronized intermittent mandatory ventilation-pressure support ventilation, pressure support ventilation with an appropriately adjusted rebreathing loop, and pressure support ventilation with the shortest available loop. After establishing steady state conditions, the authors measured cardiac output using both partial carbon dioxide rebreathing and thermodilution methods. The correlation between the data yielded by the two methods was determined by Bland-Altman analysis and linear regression. RESULTS Cardiac output with the carbon dioxide rebreathing technique correlated moderately with that measured by thermodilution when minute ventilation was set to maintain normocapnia, regardless of tidal volumes. However, when minute ventilation was set low, the carbon dioxide rebreathing technique underreported cardiac output (y = 0.70x; correlation coefficient, 0.34; bias, -1.73 l/min; precision, 1.27 l/min; limits of agreement, -4.27 to +0.81 l/min). When there was spontaneous breathing, the correlation between the two cardiac output measurements became worse. Carbon dioxide rebreathing increased spontaneous tidal volume and respiratory rate (20% and 30%, respectively, during pressure support ventilation) when the rebreathing loop was adjusted for large tidal volume. CONCLUSIONS During controlled mechanical ventilation, minute ventilation rather than tidal volume affected the accuracy of cardiac output measurement using the partial carbon dioxide rebreathing technique. When spontaneous breathing is present, the carbon dioxide rebreathing technique is less accurate and increases spontaneous tidal volume and respiratory rate.