Measurement of pulmonary capillary blood flow by a nitrous oxide rebreathing technique.

Measurements of pulmonary capillary blood flow by a nitrous oxide rebreathing technique (QN2O) were performed in 21 anesthetized and artificially ventilated minipigs with normal lungs and in nine minipigs with thrombin-induced (75-150 U kg-1 h-1) lung pathology. QN2O was calculated with the Hook-Meyer-formula and compared to cardiac output measurements (thermodilution, QT, or direct Fick's principle, QFick). The coefficient of variation in double QN2O measurements was 0.05. If the tidal volume to dead space ration (VD/VT) is normal, the nitrous oxide method works well, but when the efficacy of ventilation worsens, this gas uptake method fails to detect the circulation of the poorly ventilated parts of the lung. The mean ratio QN2O/QFick in pigs with normal lungs (58 measurements) was 1.00 +/- 0.10 (mean +/- s.d.). During thrombin infusion, the mean ratio QN2O/QT was 0.84 +/- 0.17 (n = 49). After corrections for shunt perfusion (Qs), the mean ration QN2O/(QT-Qs) was 0.89 +/- 0.17 (n = 49). QN2O/QT-Qs) decreased with increasing VD/VT. In measurements during thrombin infusion with VD/VT less than 0.33, the mean ratio QN2O/(QT-Qs) was 0.97 +/- 0.11 (n = 21), with a VD/VT between 0.33 and 0.44, the mean ratio QN2O (QT-QS) was 0.90 +/- 0.08 (n = 20), and with a VD/VT greater than or equal to 0.45, this ratio was 0.62 +/- 0.18 (n = 8). In the presence of only moderate functional inhomogeneities, this noninvasive rebreathing method will offer reliable data on pulmonary perfusion.