Objective: To determine the validity and clinical utility of the partial CO2 rebreathing technique for measurement of nonshunted pulmonary capillary blood flow and cardiac output.
Design: Prospective, controlled animal laboratory investigation and clinical trial.
Settings: Animal research facility and intensive care unit of a university hospital.
Subjects: Fifteen adult sheep, weighting 58 to 78 kg.
Patients: Mechanically ventilated patients with different underlying diseases (n = 12) and with adult respiratory distress syndrome (ARDS) (n = 8).
Interventions: CO2 elimination rate (VCO2) was measured breath-by-breath with a system developed for the study and also by gas collection (validation procedure in patients with different underlying diseases). Partial CO2 rebreathing maneuvers, cardiac output by thermodilution, and blood gas analysis were performed in sheep with lung atelectasis and in patients with ARDS.
Measurements and main results: The degree of correlation between VCO2 measured with the system developed and gas collection was very good (r2 = .95, p < .0001), and bias and precision calculations (1 +/- 9 mL/min) showed close agreement between methods. The overall degree of correlation between partial CO2 rebreathing measurements and cardiac output was moderate (r2 = .54, p < .0001), the noninvasive method tending to underestimate cardiac output, as shown by bias and precision calculations (-1.69 +/- 1.90 L/min). In contrast, the overall degree of correlation between partial CO2 rebreathing measurements and nonshunted pulmonary capillary blood flow was good (r2 = .73, p < .0001). Bias and precision calculations (0.25 +/- 0.83 L/min) showed a tendency for the partial CO2 rebreathing technique to slightly overestimate pulmonary capillary blood flow. Variance differences between partial CO2 rebreathing measurements and cardiac output could be mostly explained by intrapulmonary right-to-left shunt fraction (r2 = .51, p < .0001).
Conclusions: Our results support the use of the system developed for breath-by-breath VCO2 measurements. The lack of agreement between partial CO2 rebreathing measurements and cardiac output was mostly explained by intrapulmonary right-to-left shunt, suggesting that this technique may not be appropriate for monitoring cardiac output in patients with increased venous admixture. In contrast, our results demonstrate that the partial CO2 rebreathing technique is reliable for measurement of the effective nonshunted pulmonary capillary blood flow. This technique may prove useful to guide ventilatory therapy adjustments in an attempt to optimize nonshunted pulmonary capillary blood flow.