Search

We develop tidal-ventilation pulmonary gas-exchange equations that allow pulmonary shunt to have different values during expiration and inspiration, in accordance with lung collapse and recruitment during lung dysfunction (Am. J. Respir. Crit. Care Med. 158 (1998) 1636). Their solutions are tested against published animal data from intravascular oxygen tension and saturation sensors. These equations provide one explanation for (i) observed physiological phenomena, such as within-breath fluctuations in arterial oxygen saturation and blood-gas tension; and (ii) conventional (time averaged) blood-gas sample oxygen tensions. We suggest that tidal-ventilation models are needed to describe within-breath fluctuations in arterial oxygen saturation and blood-gas tension in acute respiratory distress syndrome (ARDS) subjects. Both the amplitude of these oxygen saturation and tension fluctuations, and the mean oxygen blood-gas values, are affected by physiological variables such as inspired oxygen concentration, lung volume, and the inspiratory:expiratory (I:E) ratio, as well as by changes in pulmonary shunt during the respiratory cycle.

The Grodins model of respiratory control (Grodins et al., 1967) describes cardio-respiratory control for a lung with homogeneous gas concentrations. In this study we modify the Grodins model to take account of the inhomogeneities in gas concentration within the lung that are seen in many subjects with respiratory illnesses. This modification has the effect of lowering arterial oxygen partial pressure significantly. We investigate the effect on cardio-respiratory control of this low arterial oxygen signal and find that the governing equations may be reduced to a single delay-differential equation. This reduced model is found to be a good approximation to the full model and gives predictions that are similar to reported clinical data.

The equations governing oxygen transport from blood to tissue are presented for a cylindrical tissue compartment, with blood flowing along a co-axial cylindrical capillary inside the tissue. These governing equations take account of: (i) the non-linear reactions between oxygen and haemoglobin in blood and between oxygen and myoglobin in tissue; (ii) diffusion of oxygen in both the axial and radial directions; and (iii) convection of haemoglobin and plasma in the capillary. A non-dimensional analysis is carried out to assess some assumptions made in previous studies. It is predicted that: (i) there is a boundary layer for oxygen partial pressure but not for haemoglobin or myoglobin oxygen saturation close to the inflow boundary in the capillary; (ii) axial diffusion may not be neglected everywhere in the model; (iii) the reaction between oxygen and both haemoglobin and myoglobin may be assumed to be instantaneous in nearly all cases; and (iv) the effect of myoglobin is only significant for tissue with a low oxygen partial pressure. These predictions are validated by solving the full equations numerically and are then interpreted physically.

Background: Measures of impairment of oxygenation can be affected by the inspired oxygen fraction., Methods: We used a mathematical model of an inhomogenous lung to predict the effect of increasing inspired oxygen concentration (FIO2) on: (1) venous admixture (Qva/Qt); (2) arterial oxygen partial pressure (PaO2); (3) the PaO2/FIO2 index of hypoxaemia; and (4) sulphur hexafluoride (SF6) retention (often taken to be true right-to-left shunt). This model predicts whether or not atelectasis will occur., Results: For lungs with regions of low V/Q, increasing the inspired oxygen concentration can cause these regions to collapse. In the absence of atelectasis, the model predicts that Qva/Qt will decrease and arterial oxygen partial pressure increase as FIO2 is increased. However, when atelectasis occurs, Qva/Qt rises to a constant value, whilst PaO2 falls at first, but then begins to rise again, with increasing FIO2. The SF6 retention increased markedly in some cases at high FIO2., Conclusions: Venous admixture will estimate true right-to-left shunt at high FIO2, even when oxygen consumption is raised. This model can explain the way that the Pa/Fl ratio changes with increasing inspired oxygen concentration.

In this study we investigate the equations governing the transport of oxygen in pulmonary capillaries. We use a mathematical model consisting of a red blood cell completely surrounded by plasma within a cylindrical pulmonary capillary. This model takes account of convection and diffusion of oxygen through plasma, diffusion of oxygen through the red blood cell, and the reaction between oxygen and haemoglobin molecules. The velocity field within the plasma is calculated by solving the slow flow equations. We investigate the effect on the solution of the governing equations of: (i) mixed-venous blood oxygen partial pressure (the initial conditions); (ii) alveolar gas oxygen partial pressure (the boundary conditions); (iii) neglecting the convection term; and (iv) assuming an instantaneous reaction between the oxygen and haemoglobin molecules. It is found that: (a) equilibrium is reached much more rapidly for high values of mixed-venous blood and alveolar gas oxygen partial pressure; (b) the convection term has a negligible effect on the time taken to reach a prescribed degree of equilibrium; and (c) an instantaneous reaction may be assumed. Explanations are given for each of these results.

The coupled conservation of mass equations for oxygen, carbon dioxide and nitrogen are written down for a lung model consisting of two homogeneous alveolar compartments (with different ventilation-perfusion ratios) and a shunt compartment. As inspired oxygen concentration and oxygen consumption are varied, the flux of oxygen, carbon dioxide and nitrogen across the alveolar membrane in each compartment varies. The result of this is that the expired ventilation-perfusion ratio for each compartment becomes a function of inspired oxygen concentration and oxygen consumption as well as parameters such as inspired ventilation and alveolar perfusion. Another result is that the "inspired ventilation-perfusion ratio and the "expired ventilation-perfusion ratio differ significantly, under some conditions, for poorly ventilated lung compartments. As a consequence, we need to distinguish between the "inspired ventilation-perfusion distribution, which is independent of inspired oxygen concentration and oxygen consumption, and the "expired ventilation-perfusion distribution, which we now show to be strongly dependent on inspired oxygen concentration and less dependent oxygen consumption. Since the multiple inert gas elimination technique (MIGET) estimates the "expired ventilation-perfusion distribution, it follows that the distribution recovered by MIGET may be strongly dependent on inspired oxygen concentration., (Copyright 2000 Academic Press.)

We investigate the effect of the width of ventilation-perfusion distributions on arterial blood oxygen content. We assume that the perfusion within the alveolar volume is a continuous function of ventilation-perfusion ratio, known as the continuous ventilation-perfusion distribution, and then write down the conservation of mass equations in the lung incorporating the nonlinear relationship between oxygen concentration in the gas phase and blood oxygen content. We solve these equations for various unimodal and bimodal ventilation-perfusion distributions believed to occur in practice and calculate the arterial blood oxygen content in each case. When a subject has a unimodal ventilation-perfusion distribution we show that the fraction of cardiac output to that mode (i.e. the fraction of non-shunted blood) has a large effect on arterial oxygen blood content. However, the width of the distribution has only a negligible effect on arterial oxygen blood content. For a bimodal ventilation-perfusion distribution the location and fraction of cardiac output to each mode has a large effect on arterial oxygen blood content. Again, the width of each mode of the distribution has little effect on arterial oxygen blood content. As a result there is little point, from a clinical perspective, in developing techniques for investigating the width of modes of these distributions since all relevant clinical information is contained in the nature (i.e. unimodal or bimodal) and in the location of the modes., (Copyright 1999 Academic Press.)