Direct measurements of alveolar gas in mechanically ventilated humans

Objective.The composition of alveolar gases is based on theoretical assumptions. The role of expired samples at the airway opening in representing the true value of mean alveolar gas remains a subject of debate. We hypothesized that phase III of oxygrams and capnograms obtained at the airway opening must be qualitative similar to the one found at the lung acini. The aim of this study is to test this hypothesis, analyzing the phase III of breath-by-breath samples of airway and alveolar gases from mechanically ventilated patients.Approach.The study compared expired gases at the airway opening and direct gas samples taken from peripheral lung parenchyma in ventilated patients undergoing thoracic surgery. We used video-assisted lobectomies as a model to get direct alveolar gas samples from the lobe that was removed by surgery.Main results.In ten anesthetized patients before lobectomy, the corresponding lung lobe was punctured by a 20 G angiocath at ∼5 mm of depth and then connected via a sampling line to sidestream and mainstream capnographs. An oscillating pattern in synchrony with the mechanical breath cycles was observed in the alveolar CO₂and O₂samples. The alveolar capnograms and oxygrams revealed an absence of phase I, very short phase II, and a similar phase III. The slope of phase III of the expired curves at the airway opening [median 2.20 (inter-quartile range 1.28) mmHg s^-1] were similar to the one of the alveolar mainstream capnograms [median 2.18 (inter-quartile range 1.24) mmHg s^-1;p= 0.76].Significance.Our findings suggest that alveolar gas can be reliably sampled at the airway opening. The similarity in the slopes of phase III between airway and alveolar samples highlights the significance of continuous gas diffusion through the alveolar-capillary membrane as well as the presence of convection- and diffusion-dependent inhomogeneities in their origin.