Cause of Low Arterial Oxygen Saturation in Pulmonary Fibrosis.

Since the description of 'alveolar-capillary block' by Austrian, McClement, Renzetti, Donald, Riley, and Cournand (1951), impairment of diffusion across the alveolar-capillary membrane has been regarded as an important factor in the production of low arterial oxygen saturations in patients with diffuse pulmonary disease. However, Luchsinger, Moser, Buhlmann, and Rossier (1957) suggested that restriction of the capillary bed might be responsible, and the demonstration by Roughton and Forster (1957) that the relatively slow chemical reaction with haemoglobin was a major resistance to the absorption of respiratory gases lent support to this view. More recently, Motley (1958) and Finley, Swenson, and Comroe (1962) have shown the importance of uneven distribution of ventilation and perfusion in producing hypoxaemia. In view of the doubt as to the importance of diffusion in gas absorption, Cotes (1963) has suggested the non-committal term 'transfer factor' instead of 'diffusing capacity' (DL) to describe the process. Roughton and Forster (1957) showed that it was possible to separate the red cell and the membrane components of gas transfer by making use of the competition between carbon monoxide and oxygen for the available haemoglobin (Fig. 1). This technique gives an estimate of the volume of the pulmonary capillaries (Vc) if the rate of combination of the gas with blood (9) is known. However, both the capillary volume (Vc) and the membrane component (DM) estimated in this way are probably influenced by uneven distribution effects (Hamer, 1963b) which limit the contact between alveolar gas and capillary blood. The relation between the factors is given by the formula I I I DL =DM #V-C In the present study, the technique of Roughton and Forster (1957) was used to analyse the process of gas transfer in seven patients with diffuse