THE EFFECT OF CHANGE IN CARDIAC OUTPUT ON INTRAPULMONARY SHUNTING

Transcription

1 Br. J. Anaesth. (1974), 46, 337 THE EFFECT OF CHANGE IN CARDIAC OUTPUT ON INTRAPULMONARY SHUNTING G. SMITH, F. W. CHENEY JR AND P. M. WINTER SUMMARY Cardiac output was varied in 6 normal dogs and in 12 dogs with oleic acid induced pulmonary oedema by inducing hyper- and hypovolaemia produced by either transfusing or bleeding ml of blood per kg body weight from the normovolaemic state. In normal dogs, a reduction in cardiac output of 42% was accompanied by a reduction in intrapulmonary shunt (Qs/Qt) from 9.8 to 6.4%. With induced pulmonary oedema, a 35% reduction in cardiac output was accompanied by a fall in Qs/Qt from 29 to 21.6%. In another series of experiments, mixed venous oxygen saturation was varied independently of cardiac output by means of veno-venous shunting of blood through a disc oxygenator. Decreasing Sv 02 from 70 to 50% in normal lungs caused a decrease in Qs/Qt from 10.1 to 7.9% whilst in pulmonary oedema a decrease in Sv 02 from 59.3 to 32.2% produced a decrease in Qs/Qt from 28.8 to 25.7%. Calculated right-to-left shunting (Qs/Qt) is known to be affected in a variety of situations by alterations in cardiac output. This study aims to assess the magnitude of change produced in calculated intrapulmonary shunting in canine lungs by alteration in cardiac output and attempts to assess the effect of the change in Sv O2 produced by alteration in cardiac output. There are conflicting data in the literature on the effect of haemorrhage-induced reductions in cardiac output on pulmonary shunt. Thus, haemorrhagic hypotension has been found to be accompanied by a decrease in pulmonary shunt ratio (Qs/Qt) in normal canine lungs (Leigh and Tyrrell, 1969; Freeman and Nunn, 1963) but an increase in Qs/Qt in the presence of a collapsed lower lobe (Wahrenbrock et al., 1970). In the clinical situation, it has been noted that patients in shock induced by a variety of aetiologies may be hypoxaemic as a result of increased shunting of blood (Border, Tibbetts and Schenk, 1968; Bredenberg et al., 1969) whilst increasing cardiac output in patients with carbon dioxide (Michenfelder, Fowler and Theye, 1966) or catecholamines (Michenfelder, Fowler and Theye, 1966; Sanders et al., 1965) leads to a rise in Qs/Qt. It is possible that these conflicts may be G. SMITH,* B.SC, M.D., F.F.A.R.C.S.; F. W. CHENEY JR, M.D.; P. M. WINTER, M.D.; Department of Anaesthesia and Anaesthesia Research Centre, University of Washington School of Medicine, Seattle, Washington * Present address: University Department of Anaesthesia, Western Infirmary, Glasgow. resolved by postulating variable contributions from the factors which regulate the distribution of blood flow in the lungs. They include the interaction between alveolar, pulmonary arterial and pulmonary venous pressures (West, 1970), the level of alveolar Po 2 and Pco 2 (Barer, 1966; Barer, Howard and Shaw, 1970) and the chemical composition of pulmonary arterial blood (Silove, Inoue and Grover, 1968). In addition, drugs which affect cardiac output may have a direct effect on the bronchioles and pulmonary vasculature. We have studied the effect of change in cardiac output produced by altering blood volume on Qs/Qt in normal canine lungs and the possible contribution of change in mixed venous oxygenation as a mechanism regulating the distribution of pulmonary blood flow. Additionally, we have studied canine lungs with pulmonary oedema produced by the intravenous injection of oleic acid a model which in some ways resembles clinical respiratory failure produced by fat embolus (Ashbaugh and Takeshi, 1968; King et al., 1971). METHODS Normal dogs. Six healthy adult mongrel dogs (wt range kg) were anaesthetized with 30 mg/kg of pentobarbitone, paralysed with 40 mg of suxamethonium and intubated with a cuffed endotracheal tube. The animals were placed in a supine position and ventilated with 100% O, at a tidal volume of 15 ml/kg,

2 338 BRITISH JOURNAL OF ANAESTHESIA a rate of 16/min and an inspiratory: expiratory ratio of 1:1. Catheters were inserted into the aorta and inferior vena cava. Through the right external jugular vein, a No. 5 Swan-Ganz catheter was inserted by the flow-directed technique so that its tip lay just distal to the pulmonary valve. A second No. 5 Swan- Ganz catheter was inserted via the left external jugular vein into the wedge position with the balloon inflated. The first catheter was used for sampling mixed venous blood and recording pulmonary artery pressure. Aortic, pulmonary arterial and pulmonary arterial wedge pressures were recorded with appropriate electronic equipment. The baseline for the pressure transducers was taken as the level of the right atrium. Aortic and pulmonary arterial blood samples were obtained, placed immediately into ice water and analysed as soon as possible for Po 2, Pco 2 and ph on an Instrumentation Laboratories 113 blood-gas analyser using appropriate correction for the temperature of the animal (Hedley-White, Radford and Laver, 1966; Severinghaus, 1966). Oxygen saturation and haemoglobin were measured directly with an Instrumentation Laboratories Co-oximeter model No Oxygen consumption was measured directly using a Collins 9-litre spirometer (with CO, absorption) connected in a closed system with the ventilator. Values were corrected to STPD. Cardiac output was calculated from the Fick equation. The shunt ratio (Qs/Qt) was determined with values of arterial and mixed venous oxygen contents using the standard formula (Comroe, 1965): Qs/<Jt=(Cc' 02 -Ca O2 )/(Cc' 02 -Cv 02 ) Pulmonary vascular resistance was calculated as follows: Pulmonary vascular resistance (dyne sec cm~ 5 )= Pulmonary artery pressure (mm Hg) pulmonary artery wedge pressure (mm Hg) go X 1 Measurements were made in each animal before and after haemorrhage of ml/kg or transfusion of ml/kg of uncrossmatched donor dog blood. Two sets of measurements were made during hypovolaemia, hypervolaemia and "normovolaemia", the order being randomized. Pulmonary oedema. In an additional 12 mongrel dogs (wt range kg), following control measurements, pulmonary oedema was produced by the slow injection of ml/kg of oleic acid into the right ventricle. After 1-H hours, two sets of measurements were made during hypovolaemia of ml/kg, "normovolaemia" and hypervolaemia of ml/kg in a random order. To compensate for fluid losses, normal saline was infused at a rate of 150 ml/hr. Bypass study. To assess the possible effects of change in mixed venous oxygen saturation on Qs/Qt, experiments were performed on 11 healthy adult mongrel dogs (wt range kg). The composition of venous blood returning to the heart was altered by means of an extracorporeal circuit (fig. 1). The withdrawal catheters comprised two large-bore plastic tubes, one inserted via the left jugular vein, the other inserted via the femoral vein so that the tip lay at the junction of the iliac veins. Blood was withdrawn by a roller pump which fed a reservoir, the level of which was maintained to ensure that the volume of fluid in the extracorporeal circuit was constant. From the reservoir, blood was pumped into a disc oxygenator and thence via a Sams pump into a large Portex catheter inserted through the left femoral vein into the inferior vena cava to the level of the diaphragm. F:G. 1. Schematic diagram of veno-venous bypass. The inferior withdrawal catheter terminates at the junction of external iliac vein with inferior vena cava. The infusion catheter terminates at the level of the diaphragm.

3 CHANGE IN CARDIAC OUTPUT AND INTRAPULMONARY SHUNTING 339 The reservoir and circuit were primed in advance with 500 ml of Ringer lactate, 500 ml of Dextran 70 in 0.9% NaCl and 1000 ml uncrossmatched donor dog blood, together with 50 m.equiv NaHCO 3 and 4000 units of sodium heparin. In addition, each dog received 3000 units of heparin 3-hourly. A plastic catheter was placed in the aorta and a No. 5 Swan-Ganz catheter in the pulmonary artery. The mode of anaesthesia was similar to that used in the previous experiments and each dog was ventilated with 100% oxygen throughout the experiment which lasted 3-5 hours. Arterial pressure, pulmonary artery and wedge pressure, and arterial and mixed venous blood-gases and ph were measured as before. Cardiac output was measured by the dye dilution technique employing indocyanine green and a Gilford cuvette densitometer. Changes in mixed venous O 2 content were effected by passing either 95% O 2 /5% CO 2 or 95% N,/5% CO 2 through the "oxygenator". At least 10 min at a bypass flow of ml/min elapsed after oxygenation or deoxygenation before measurements were made. Between three and five pairs of measurements were made in each animal. In 7 of these animals, pulmonary oedema was then induced by the slow injection of oleic acid, 0.04 ml/kg through the Swan-Ganz catheter positioned in the right ventricle for this purpose. After 1-11 hours, measurements were made with either O, or N 2 flowing through the oxygenator. In all experiments, the animal's temperature was kept at C by adjustment of the extracorporeal circulation and a heating blanket. Statistical analysis was carried out with Student's f-test for paired data. RESULTS The two measurements made at each blood volume were averaged and the means of these values in all dogs are shown in table I for the 6 normal dogs, and in table II for the 12 dogs that received oleic acid. In normal dogs, hypovolaemia was accompanied by a decrease in cardiac output from 2.33 to 1.35 l./min (table I) and a reduction in Qs/Qt from 9.8 to 6.4%. Pulmonary artery pressure, wedge pressure and mixed venous oxygen saturation all decreased significantly (P<0.05). During hypervolaemia, there was an increase in the value of all these measurements from the normovolaemic level but none reached a level of statistical significance. There were no changes in heart rate or mean arterial pressure. Similar changes were observed in the presence of pulmonary oedema (table II). A reduction in cardiac output was accompanied by a decrease in Qs/Qt from 29 to 21.6% and a reduction in mean pulmonary artery and wedge pressure as well as mixed venous Po 2. There were no significant changes in heart rate but arterial pressure was significantly lower during hypovolaemia (125 mm Hg) than "normovolaemia" (143 mm Hg). Bypass study. The results are summarized in table III. Alteration of the mixed venous oxygen tension produced no significant change in pulmonary artery or wedge pressure or cardiac output. There was no significant TABLE I. Haemodynamic and blood-gas data: control dogs. Values are mean ±SE for 6 normal dogs. Hypovolaemia and hypervolaemia were produced by bleeding or transfusing ml I kg of blood from the normovolaemic state. Mean pulmonary artery pressure Pulmonary wedge pressure Pulmonary vascular resistance (dyne sec cm" 6 ) Pao 2 (mm Hg) Arterial ph (units) Pv O2 (mm Hg) Sv O2 (%) Pvco 2 (mm Hg) Mixed venous ph (unit) Qs/Qt (%) A-a Po 2 difference (mm Hg) Vo, (ml/min) Hypovolaemia 1.35±0.12* 9.7±0.7* 0.7 ±0.3* 555 ±75* 485 ± ± ±2.9* 59 ±3.6* 47± ± ±0.5* 205 ±13 128±11 Normovolaemia 2.33 ± ± ± ± ± ± ± ±4.5 44± ± ± ±15 138±11 Significant difference (P<0.05) between adjacent data. Hypervolaemia 2.95 ± ± ± ±61 492± ± ±2.1 80±2.5 44± ± ± ±10 146±12

4 340 BRITISH JOURNAL OF ANAESTHESIA TABLE II. Haemodynamic and blood-gas data: dogs with induced pulmonary oedema. Values are mean ±SEfor 12 dogs with normal blood volume before inducing pulmonary oedema {control) and with pulmonary oedema during bleeding or transfusing mljkg of blood from the normovolaemic state. Mean pulmonary artery pressure Pulmonary wedge pressure Pulmonary vascular resistance (dyne sec cm" 5 ) Pao 2 (mm Hg) Paco 2 (mm Hg) Arterial ph (units) Pvo 2 (mm Hg) Sv 02 (%) Pvco 2 (mm Hg) Mixed venous ph (unit) Os/Qt(%) A-a Po 2 difference (mm Hg) Vo 2 (ml/min) Hypovolaemia 1.37 ±0.19* 16±1.5* 3 ±0.9* 883 ±127* 187 ±28 38±2.1* 7.29 ±0.02* 36 ±2.8* 49.5 ±3.0* 51 ± ± ±3.2* 496 ± ±8.8* Pulmonary oedema Normovolaemia 2.11 ± ±1.6* 7±1.1* 653 ±60* 191 ±40 43 ± ± ± ± ± ± ± ± ±10 *Significant difference (P<0.05) between adjacent data. Hypervolaemia 2.28 ± ± ± ± ±31 41 ± ± ± ±3.4 51± ± ± ± ±9.5 TABLE III. Haemodynamic and blood-gas data: bypass study. Values are mean ±SEfor 11 dogs after instituting veno-venous bypass. Measurements were made before (normal lungs) and after (pulmonary oedema) the injection of oleic acid into the right ventricle and data collected during oxygenation (O 2 ) or deoxygenation (N t ) of the extracorporeal blood. Mean pulmonary artery pressure Pulmonary wedge pressure Pulmonary vascular resistance (dyne sec cm" 5 ) Arterial Po 2 (mm Hg) PcOj (mm Hg) ph (units) Mixed venous Po 2 (mm Hg) Pco, (mm Hg) ph (units) Svo,(%) Qs/Qt(%) o ± ± ± ± ± ± ± ±1.1* 38.2 ± ±0.02* 70.5±3.1* 10.1±1.5* change in mixed venous Pco 2 in the normal lungs, but there was a small change in mixed venous ph in normal and abnormal lungs. Pulmonary shunt ratio increased significantly from 7.9 to 10.1% (mean difference 2.17%, SE 0.33) on changing mixed venous oxygen saturation from 50 to 70% (mean difference 19.9%, SE 3.12) in the normal lungs, and from 26 to 29% (mean difference 3.1%, SE 1.1) on changing Sv 02 from 32 to 59.3% (mean difference 26.8%, SE 3.4) in oedematous lungs. Normal lungs N ± ± ± ± ± ± ± ± ± ± ± ±1.5 Bypass Pulmonary oedema o ± ± ± ±88 210± ± ± ±2.3* 42.5 ±4* 7.28 ±0.05* 59.3 ±5* 28.8±5.7* * Significant difference (P<0.05) between adjacent data. N ± ± ± ± ± ± ± ± ± ± ± ±3.1 DISCUSSION In the present study, we have shown that a reduction in cardiac output by hypovolaemia in dogs both with normal lungs and with experimental pulmonary oedema is associated with a considerable fall in Sv O2 but little change in arterial Po 2 and a reduction in Qs/Qt. This is confirmatory of previous work by Freeman and Nunn (1963), Gerst, Rattenborg and Holaday (1959) and Wahrenbrock et al. (1970). If oxygen consumption remains constant, as

5 CHANGE IN CARDIAC OUTPUT AND INTRAPULMONARY SHUNTING 341 cardiac output falls, mixed venous oxygen content falls. In the presence of a constant Qs/Qt and a constant alveolar and pulmonary end capillary Po 2, the oxygen content of arterial blood would fall also. This concept has lead to models for predicting the change in Pa 0, resulting from changes in cardiac output at a given Qs/Qt (Kelman et al., 1967; Marshall and Wyche, 1972). In addition, it has been demonstrated clinically that increases in cardiac output induced by CO, have been associated with rises in Pa 02, as predicted by the model (Prys-Roberts et al., 1967). In contrast, a fall in cardiac output, induced by pentolinium, was not associated with a fall in Pa 02 in dogs, implying that Qs/Qt decreased concomitantly (Leigh and Tyrrell, 1969). Several studies have produced data on the effect of change in cardiac output on intrapulmonary shunting but interpretation is often complicated by possible direct effects of pharmacological agents on the pulmonary vasculature. In man, a substantial increase in cardiac output produced by elevation of Pa co, was without effect on Qs/Qt during halothane (Prys-Roberts et al., 1968) and enflurane (Marshall et al., 1971) anaesthesia. However, Michenfelder, Fowler and Theye (1966) found that increased P C02 could lead to a fall in Qs/Qt and rise in Pa 02 in the absence of cardiac output changes and also that increased cardiac output was accompanied by similarly directed change of shunting "interacting with and at times over-riding the apparent direct 'effect' of CO 2 ". An explanation for these discrepancies may be derived from pulmonary perfusion experiments in the bovine lung in situ in which it was shown that acidaemia accentuates the pulmonary vasoconstriction induced by hypoxia (Silove, Inoue and Grover, 1968). Similar findings have been obtained in excised dog lungs (Lloyd, 1966). In the abnormal lung, a positive correlation has been found between Qs/Qt and cardiac output in patients with existing high intrapulmonary shunting due to chronic obstructive pulmonary disease or acute infection (Hedley-White, Pontoppidan and Morris, 1966) and in patients with acute oedematous cor pulmonale (Penman, Howard and Stentiford, 1968). In the present study, a similar finding has been obtained in induced pulmonary oedema, whereas in the presence of acute collapse of one lower lobe, an increase in Qs/Qt was found to occur with haemorrhagic hypotension (Wahrenbrock et al., 1970). It was suggested that in the atelectatic lobe, a fall in left atrial pressure induced by hypovolaemia allowed relatively improved perfusion through the lobe (Wahrenbrock et al., 1970). There are many factors which are known to regulate the calibre of pulmonary blood vessels including the pattern of distribution of pulmonary blood flow, pulmonary arterial, left atrial and alveolar pressures, together with hypoxia, CO 2, ph, temperature and catecholamines. In the present study, both pulmonary artery wedge pressure and pulmonary artery pressure fell with hypovolaemia. However, there was a relatively poor correlation between driving pressure (Pa-Paw) and Qs/Qt. It is possible that hypovolaemia causes a reduction in mean airway pressure and improved perfusion in Zone II (West and Dollery, 1965; West, Dollery and Naimark, 1964), thereby accounting for a fall in Qs/Qt with a reduction in cardiac output. However, in a similar study on artificially ventilated greyhounds, subjected to changes in blood volume of greater magnitude than in the present study, hypovolaemia was accompanied by a non-significant change in mean airway pressure of 0.3 cm H 2 O whilst Qs/Qt fell significantly (Sykes et al., 1970). In the bypass studies described here, an attempt was made to assess the effect of the mixed venous Po 2 on Qs/Qt by altering the Pv 02 experimentally whilst stabilizing as many other factors as possible. Fortunately, perhaps, cardiac output and vascular pressures were unaffected by changes in Pv 02 though phv increased slightly. Increased oxygenation of mixed venous blood was associated with an increase in Qs/Qt and a small fall in ph in both normal and abnormal lungs. It is unlikely that the change in ph was responsible for the change in shunt as acidaemia has been shown to potentiate the vasoconstriction of hypoxaemia (Silove, Inoue and Grover, 1968) which could cause an effect opposite to that seen. The acidaemia might be expected to diminish blood flow through areas acting as functional shunts, but in fact the shunt increased with the slight reduction in mixed venous ph. Our bypass studies showed that a change of 20% in mixed venous saturation accounted for a change of 2.2% in shunt in normal lungs. In abnormal lungs, a 27.1% change in Sv O2 was accompanied by a 3.1% change in Qs/Qt (table III). During hypovolaemia, however, a change in Sv 02 of 16% was accompanied by a change of 3.4% in Qs/Qt in normal lungs, whilst in pulmonary oedema a change of 12.6% in Sv O2 was associated with a change of 7.4% in Qs/Qt. This suggests that changes in mixed venous oxygenation may contribute a small, perhaps constant amount to the regulation of distribution of pulmonary blood flow during altered cardiac output. We postulate that the decrease in Sv 02 increased

6 342 BRITISH JOURNAL OF ANAESTHESIA pulmonary vasoconstriction and diminished the amount of blood flow through atelectatic areas. An increase in Sv 02 caused vasodilatation of the otherwise constricted blood vessels in atelectatic hypoxic areas of the lung thereby increasing perfusion of the atelectatic areas. If this hypothesis is correct, it is necessary to postulate that a 3% change in shunt was not of sufficient magnitude to be accompanied by measurable changes in resistance as there was no change in pulmonary vascular resistance on altering mixed venous saturation in the bypass study. ACKNOWLEDGEMENTS The authors thank Drs Thomas F. Hornbein and J. W. Butler for reviewing the manuscript and Messrs R. Tuck and R. Henry for their technical assistance. Graham Smith was in receipt of a U.K. Medical Research Council Travelling Fellowship. Peter M. Winter was in receipt of a Research Career Development Award No. 5 KO4 HL Supported by Public Health Service Grant No. 5 RO1 1 + L and GM from the National Institute of General Medical Sciences. REFERENCES Ashbaugh, D. G., and Takeshi, U. (1968). Respiration and haemodynamic changes after injection of free fatty acids. J. Surg. Res., 8, 417. Barer, G. R. (1966). Reactivity of the vessels of collapsed and ventilated lungs to drugs and hypoxia. Circ. Res., 18, 366. Howard, P., and Shaw, J. W. (1970). Stimulus response curves for the pulmonary vascular bed to hypoxia and hypercapnia. J. Physiol. (Lond.), 211, 139. Border, J. R., Tibbetts, J. C, and Schenk, W. G. (1968). Hypoxic hyperventilation and acute respiratory failure in the severely stressed patient: massive pulmonary arterio venous shunts? Surgery, 64, 710. Bredenberg, C. E., James, P. M., Collins, J., Anderson, R. W., Martin, A. M., and Hardaway, R. M. (1969). Respiratory failure in shock. Ann. Surg., 169, 392. Comroe, J. H. (1965). Physiology of Respiration, p Chicago: Year Book Medical Publishers Incorporated. Freeman, J., and Nunn, J. F. (1963). Ventilation-perfusion relationships after haemorrhage. Clin. Sci., 24, 135. Gerst, P. H., Rattenborg, C, and Holaday, D. A. (1959). The effect of haemorrhage on pulmonary circulation and gas exchange. J. Clin. Invest., 38, 524. Hedley-White, J., Pontoppidan, H., and Morris, M. J. (1966). The response of patients with respiratory failure and cardiopulmonary disease to different levels of constant volume ventilation. J. Clin. Invest., 45, Radford, E. P., and Laver, M. B. (1966). Nomogram for temperature correction or electrode calibration during Po, measurements. J. Appl. Physiol., 20, 785. Kelman, G. R., Nunn, J. F., Prys-Roberts, C, and Greenbaum, R. (1967). The influence of cardiac output on arterial oxygenation: a theoretical study. Br. J. Anaesth., 39, 450. King, E. G., Wagner, W. W., Ashbaugh, D. G., Latham, L. P., and Halsey, D. R. (1971). Alterations in pulmonary microanatomy after fat embolism. Chest, 59, 524. Leigh, J. M., and Tyrrell, M. F. (1969). Arterial oxygenation during induced hypotension in the dog. Br. J. Surg., 56, 380. Lloyd, T. C. (1966). Influence of blood ph on hypoxic pulmonary vasoconstriction. J. Appl. Physiol., 21, 358. Marshall, B. E., Cohen, P. J., Klingenmaier, C. H., Neigh, J. L., and Pender, J. W. (1971). Some pulmonary and cardiovascular effects of enflurane (Ethrane) anaesthesia with varying Pa C o 2 in man. Br. J. Anaesth., 43, 996. Wyche, M. Q. jr (1972). Hypoxaemia during and after anesthesia. Anesthesiology, 37, 178. Michenfelder, J. D., Fowler, W. S., and Theye, R. A. (1966). CO 2 levels and pulmonary shunting in anaesthetised man. J. Appl. Physiol., 21, Penman, R. W. B., Howard, P., and Stentiford, N. H. (1968). Factors influencing pulmonary gas exchange in patients with acute edematous cor pulmonale due to chronic lung disease. Am. J. Med., 44, 8. Prys-Roberts, C, Kelman, G. R., Greenbaum, R., and Robinson, R. G. (1967). Circulatory influences of artificial ventilation during nitrous oxide anaesthesia in man. Br. J. Anaesth., 39, 533. Kain, M. L., and Bay, J. (1968). Haemodynamics and alveolar-arterial Po 2 differences at varying Pa C o 2 in anaesthetised man. J. Appl. Physiol., 25, 80. Sanders, C. A., Harthorne, J. W., Heitmann, H., and Laver, M. B. (1965). Effect of yasopressor administration on blood gas exchange in mitral disease. Clin. Res., 13, 351. Severinghaus, J. W. (1966). Blood gas calculator. J. Appl. Physiol, 21, Silove, E. D., Inoue, T., and Grover, R. F. (1968). Comparison of hypoxia, ph and sympathomimetic drugs on bovine pulmonary vasculature. J. Appl. Physiol., 24, 355. Sykes, M. K., Adams, A. P., Finlay, W. E. I., Wightman, A. E., and Munroe, J. P. (1970). The cardiorespiratory effects of haemorrhage and overtransfusion in dogs. Br. J. Anaesth., 42, 573. Wahrenbrock, E. A., Carrico, C. J., Amundsen, D. A., Trummer, M. J., and Severinghaus, J. W. (1970). Increased atelectatic pulmonary shunt during haemorrhagic shock in dogs. J. Appl. Physiol., 29, 615. West, J. B. (1970). Ventilation/Blood Flow and Gas Exchange, 2nd edn. Oxford: Blackwell. West, J. B., and Dollery, C. T. (1965). Distribution of blood flow and the prlksure-flow relations of the whole lung. J. Appl. Physiol, 20, 175. Naimark, A. (1964). Distribution of blood flow in isolated lung: relation to vascular and alveolar pressures. J. Appl. Physiol, 19, 713.