Clinical Science (1979) 51,389-396
Changes in arterial blood gases during and after a period of
oxygen breathing in patients with chronic hypercapnic
respiratory failure and in patients with asthma
M. R U D O L F * , J. A. M c M . T U R N E R , B. D. W. H A R R I S O N , J . F. R I O R D A N
A N D K . B. S A U N D E R S
Department of Medicine, The Midlesex Hospital Medical School, London
(Received 1 December 1978; accepted 18 June 1979)
Summary
1. Ten patients with chronic hypercapnic respiratory failure (group 1) and eight patients with
asthma (group 2) breathed pure 0, from an MC
mask for 60 min. Blood gases were measured during
this period and for the subsequent 45 min.
2. In nine of ten patients in group 1 and in all
eight patients in group 2 arterial 0, tension (Pa,o,)
fell to values lower than had been obtained before
0, was given.
3. These undershoots in P40, are unrelated to
changing CO, stores or to hypoventilation, and are
more likely due to persistence of altered ventilationperfusion ratios associated with 0, breathing.
4. Magnitude of the undershoots is usually
small, and periods of less than 15 min off 0, are
unlikely to be harmful.
Key words: asthma, chronic hypercapnia, hypoxaemia, oxygen breathing, oxygen transients, respiratory failure, ventilation-perfusion ratios.
Introductiop
In treating patients with acute exacerbations of
chronic hypercapnic respiratory failure it may be
difficult to increase arterial oxygen tension (Pa.0,)
* Present address: Royal Postgraduate Medical
School, Hammersmith Hospital, Ducane Road,London
W12 OHS.
Correspondence: Dr M. Rudolf, Royal Postgraduate
Medical School, Hammersmith Hospital, Ducane Road,
London W 12 OHS.
without an unacceptable rise in the arterial carbon
dioxide tension (Pa,co,). It is best to give a small
but continuous increase in inspired 0, concentration (Campbell, 1960b; Hutchison, Flenley
& Donald, 1964; Campbell, 1967). Intermittent
administration of 0, (Cohn, Carroll & Riley, 1954)
is illogical because it only intermittently relieves
hypoxia (Massaro, Katz & Luchsinger, 1962).
Another disadvantage of intermittent 0, therapy
in such patients is that some become more
hypoxaemic after supplemental 0, than they were
before they received it (Massaro et al., 1962;
Cullen & Kaemmerlen, 1967) and thus intermittent 0, administration may be dangerous. Two
textbooks of respiratory disease state that 0,
therapy in patients with hypercapnic respiratory
failure should be continuous (Cumming & Semple,
1973; Crofton & Douglas, 1975) and point out that
hypoxaemia may worsen when 0, therapy stops
but give different explanations: the former ascribes
the undershoot in Pa,o, to persistent hypoventilation induced during 0, breathing, the latter
to the different washout characteristics of large
CO, and small 0, stores (Campbell, 1960a).
Although continuous administration of 0,may
be intended, such therapy is often intermittent,
either because the patient removes the mask or
because the physician removes it so that arterial
blood gases can be measured during air breathing.
What then is the time course of Pa,co, and P40,
after a period of 0, breathing and, if there is a
rebound fall in Pa,o,, what is the magnitude and
the cause of it? We have studied these questions in
389
390
M. Rudolfet al.
patients with chronic hypercapnic respiratory
failure and in patients with asthma who were
hypoxic but had a normal or low Pa,co,.
further samples were then taken at 2-5, 5, 10, 15
and 20 min.
The experimental protocol had been approved
by the Hospital’s Ethical Committee, and there
were no adverse effects.
Methods
Blood-gas analysis
Patients
We studied two separate groups of patients.
Group 1. Ten patients (eight male, two female),
mean age 65 years (range 45-78 years). All had
been admitted to hospital with an episode of acute
on chronic hypercapnic respiratory failure, but
were clinically stable at the time of the study. All
had chronic bronchitis (Medical Research Council,
1965), and nine were smokers or ex-smokers. All
had chronic irreversible airflow obstruction (mean
FEV,,, 0.8 litre; range 0-5-1-35 litres). At the time
of study, mean Pa,o, was 49.0 mmHg (6-53 kPa)
with range 39-5-58.4 mmHg and mean Pa,co,
was 58.0 mmHg (7.74 kPa) with range 45.2-73.3
mmHg.
Group 2. Eight patients (five male, three
female), mean age 51 years (range 18-69 years).
All had recently been admitted to hospital with
severe asthma with hypoxaemia (Pa,o, < 60
mmHg) and normo- or hypo-capnia (Pa,co, < 45
mmHg). All were clinically stable at the time of
study. None had chronic bronchitis and all had
evidence of reversible airflow obstruction. At the
time of study, mean Pa,02 was 63.5 mmHg (8.47
kPa) with range 524-70.5 mmHg, and mean
Pa,co, was 35.3 mmHg (4.71 kPa) with range
25-4-40.8 mmHg.
Protocol
Diuretics and bronchoddators were withheld for
12 h before the study. A brachial artery cannula
was inserted under local anaesthesia, and blood
samples were taken 15 and 30 rnin later while the
patient was breathing air; the mean Pa,o, and
Pa,co, values of these two samples provided
control (pre-0,) values. Pure 0, was then given via
an MC mask at a flow rate of 4 litredmin, giving
an inspired 0, concentration of about 60%. Blood
samples were taken at 30 and 60 min of 0,
breathing, after which the patient again breathed
air and further samples were taken at 2, 4, 6, 10,
20, 30 and 45 min. Throughout the study the
patients maintained a constant semi-recumbent
posture. Patients in group 2 took three deep
inspirations at 45 min after 0, breathing and
Blood samples were taken into heparin-treated
glass syringes and analysed in duplicate with
standard Radiometer electrodes (type E 5046, type
E 5036 and type G297/G2, all with a standard
PHM27 meter), calibrated with standard gases and
buffers immediately before and after every sample.
Arterial 0, saturations were calculated with the
Severinghaus blood-gas calculator (Severinghaus,
1966).
For each group changes in blood-gas data were
assessed for statistical significance by Student’s
paired t-test.
Results
Full data on individual blood-gas analyses are
given in Clinical Science Tables 7914 and 7915,
deposited with the Librarian, the Royal Society of
Medicine, 1 Wimpole Street, London W I M 8AE,
who will supply copies on request.
Group 1
Mean Pa,o, (f1 SD) increased from a control
value of 49.0 rf: 6.9 mmHg (6.53 f 0-92 kPa) to
184 rf: 68-6 mmHg (24.52 f 9.15 kPa) at the end
of the 0, breathing hour, and mean Pa,co, increased
from 58.0 f 8.0 mmHg (7.74 f 1.07 kPa) to
66.8 f 9 . 4 mmHg (8.91 1.26 kPa); the increase
in mean Pa,co, was highly significant (P < O-OOl),
and a rise in CO, tension was seen in all ten patients
(Table 1).
On cessation of 0, breathing, both Pa,o, and
Pa,co, fell to values not significantly different from
control values (P > 0.05) by 10 min. Mean Pa,co,
subsequently remained not significantly different
from control, but mean Pa,o, continued to fall,
being significantly lower than the control at 30 min,
45.5 rf: 5.9 mmHg (6.07 f 0.78 kPa), P < 0.05,
and at 45 min, 44.2 f 6.8 mmHg (5-90 f 0.91
kPa), P < 0.01. Pa,o, fell below control values in
nine of the ten patients. Calculated mean 0,
saturation at 45 min (74.2 f 10.0%) was also
significantly lower than the control value (79.0 f
8.9%, P < 0.05). The largest individual fall in
saturation was 16.3%.
The time course of these events is more clearly
shown by plotting changes relative to the control
49.0
(6.9)
58.0
(8.0)
60
183.9***
(68.6)
66.8***
(9.4)
30
176.3***
(68.7)
65.0***
(9.9)
On 0,
2
4
71.6**
91.
(19.9)
(29.2)
63*7*** 62@**
(8.8)
(9.1)
6
63.3.'
(11.1)
61.3..
(8.6)
30
45.5.
(5.9)
58.6t
(8.6)
20
47.57
(6.2)
59.6t
(8.4)
10
53.0t
(6.6)
59.9t
(8.6)
After cessationof 0,
44.2..
(6.8)
58.lt
(7.8)
45
TABLE1. Values of mean Pa.0, and mean Pa,co, in group 1patients
are shown. Significance of difference from control value: *P < 0.05; **P < 0.01; ***P< 0401;
not significant (P> 0-05).
Control
Pre-0,
SD)
(mHs)
pace,
pw,
(mmHg)
Time (min) ...
63.5
(5.7)
35.3
(5.7)
Control
Pre-0,
30
220.4***
(69.7)
39.57
(11.4)
60
2
223.5*** 90.3.'
(73.8) (21.5)
36.4t
38.6t
(9.0)
(7.3)
On 0,
4
72.4.
(11.3)
35.9t
(6.2)
6
66.2t
(8.9)
36.2t
(6.6)
63.27
(7.8)
34.9t
(5.6)
10
(5.1)
(5.8)
30
59@*
(4.9)
36.47
20
60.2t
(5.6)
35.77
After cessation of 0,
45
36.2t
(6.2)
(5.5)
57.7**
2:
59.7tt
(6.7)
37.ltt
(6.7)
58.ltt
(6.1)
36.3tt
(6.5)
5
58.5tt
(7.0)
36.6tt
(6.6)
10
.
59.3tt
(8.7)
36.7tt
(7.3)
15
After three deep inspirations
58.4tt
(7.3)
36.177
(6.5)
20
Tmm 2. Values of mean Pa,o, and mean Pa,co, in group 2 patients
Mean values ( + I SD) are shown. Significance of difference from control value: P < 0.05; ** P < 0.01; *** P < 0401. t Not significantly different from control value (P>
0.05); tt not significantly different from value at 45 m h off 0,[P> 0.05).
(mmHs)
Pa,co,
(mmHs)
pao,
Time (min) ...
Mean values ( + I
z
o
.
(
3
a
M.RudoIfet al.
392
-5
-1
0, breathing
J -1
i
1
0
--
5
10
n
20
15
,
,
25 30
,
35
,
40
45
Time (min)
FIG. 1. Changes in mean Pa,o, and mean Pa,co, (expressed as deviations from control pre-0, values) in
group 1 patients.
0,b;thing
-5k
-10
0 01
~
I
I
I
5
10
15
I
I
1
1
20 25 30 35
l
40
I
45
Time (min)
FIG.2. Changes in mean Pa,o, and mean Pa,co, (expressed as deviations from control pre-0, values) in
group 2 patients.
pre-0, values (Fig. l), when the undershoot at 30
and 45 min is clearly demonstrated.
Group 2
Table 2 shows the alterations in mean Pa,o, and
Pa,co, and Fig. 2 the data plotted as changes from
control pre-0, values. Mean Pa,o, increased from
a control value of 63.5 f 5 . 7 mmHg (8.47 f 0.76
kPa) to 223.5 ? 73.8 mmHg (29.80 ? 9.84 kPa)
at the end of the 0, breathing hour. Mean Pa,o,
then fell to reach a value not significantly different
from control by 6 min, came back to the control
Blood gases during and Bfter oxygen breathing
level by 10 min, and fell further to 57-7 f 5.5
mmHg (7.69 f 0.73 kPa) at 45 min. Both the
values at 30 and 45 min were significantly lower
than the pre-0, value (P < 0.01). An undershoot in
Pa,o, was seen in all eight patients.
There was a small but statistically insignificant
rise in mean Pa,co, during the 0, breathing hour,
from 35.3 f 5.7 mmHg (4.71 0.76 kPa) to 38.6
f 9.0 mmHg (5.15 f 1.20 kPa). None of the
values of mean Pa,co, after cessation of 0,
breathing was significantly different from control.
The three deep inspirations that patients in this
group performed at 45 min after 0, breathing
produced no significant changes in the arterial
blood gases, and in particular this manoeuvre did
not abolish the undershoots in Pa,o,.
lncreasc in Pa,co, after 1 h 0, breathing (mmHg)
'
0
2.0r
5
10
I
I
15
1
i'
7
a
15
10
$
4
j5t
a
a @ a
a
-0.5
Discussion
An undershoot in Pa,o, below the control pre-0,
value was found in nine of the ten patients with
chronic hypercapnic respiratory failure and in all
eight of the asthmatic patients, which confirms the
original observations of Massaro ef al. (1962)and
of Cullen & Kaemmerlen (1967). We believe that a
rebound fall in Pa,o, 30-45 rnin after cessation of
0, therapy is a common through usually unrecognized event.
What is the cause of this undershoot? One theory
is that, in patients with chronic hypercapnia, the
oxygen breathing may cause a fall in alveolar
ventilation by removing the sole remaining drive to
breathing. When supplementary oxygen is removed,
hypoventilation persists while the patients are
breathing air with an additional fall in Pa,o,
(Crofton & Douglas, 1975). If the metabolic production of CO, remains constant, Pa,co, should
then be elevated above the control value at the
time that the Pa,o, undershoot occurs, which is not
so (Fig. 1).
A second hypothesis is that the Pa,o, undershoot is due to the differing washout characteristics
of large CO, and small 0, stores after 0,
breathing, with persistence of raised PCO,in both
blood and alveolar gas, which is responsible for a
decreased alveolar Po, and hence further arterial
hypoxaemia (Campbell, 1960a;Cumming & Semple, 1973). But although washout is indeed slower
for CO, than for 0, (Fig. l), the CO, washout is
nevertheless completed by the time the Pa,o,
undershoot occurs at 45 min.
In addition, as both the hypoventilation and the
slow-washout hypotheses postulate the presence of
elevated Pa,co, one would expect the largest Pa,o,
undershoots to be associated with the highest rises
393
a
3c
a
I n
L
a
0
0.5
1.0
1.5
2.0
Increase in Pa,co, after 1 h 0, breathing (kPa)
F I ~3.
. Pa,o, undershoot at 45 min after cessation of
0, breathing plotted against increase in Pa,co, atter 60
min of 0, in group 1 patients.
in Pa,co, produced during the 0, breathing hour.
But in group 1 the largest Pa,o, undershoot
occurred in the patient who had the smallest
increase in Pa,co, during 0, administration and
the one patient who did not develop an undershoot
in Pa,o, at 45 min was the very patient who had the
largest rise in Pa,co, brought about by 0, (Fig. 3).
Thus in group 1 the undershoot in P40, after 0,
breathing cannot be explained by either of the conventional theories, for the production of this
rebound hypoxaemia is unrelated to any changes
that occur in Pa,co,. Nor is the presence of chronic
hypercapnia essential, for all patients in group 2
developed undershoots in Pa,o, though P4c0,
was normal or low.
While we can dispose of previous hypotheses
with some confidence, we have considerable difficulty in proposing a plausible new one. We felt it
important to interfere with the subjects' breathing
as little as possible, and particularly to avoid using
a mouthpiece or hood. There is no information on
the effect of breathing through a mouthpiece on
ventilation in patients with chronic airways
obstruction or with asthma, but in patients with
heart failure either hyperventilation or hypoventilation may occur (Saunders, 1966), and it
would certainly have been difficult in this study to
attribute any blood-gas changes to the after-effect
of 0, alone, rather than to the effect of a mouthpiece. This meant that we could not measure CO,
394
M . Rudolfet al.
output, and must assume it to be unchanging in our
dismissal of one previous hypothesis, nor could we
give an accurately controlled inspired concentration of 0,, or measure expired ventilation,
which would have allowed a more precise analysis
of gas exchange with the 0,-CO, diagram (Rahn
& Fenn, 1955).
At the end of the study, Pa,o, was lower than
control values, whereas Pa,co, was the same.
Therefore a change in ventilation-perfusion
relations must have occurred. We now consider
four possible mechanisms.
(1) Alveoli collapse faster if N, is replaced by
0, (Burger & Macklem, 1968;Dantzker, Wagner
& West, 1975). Possibly, then, 0, breathing
induced alveolar collapse with shunting which
persisted when air-breathing was resumed. In some
patients with heart failure and pulmonary oedema a
few deep breaths may raise Pa,o,, suggesting reopening of blocked airways or collapsed alveoli
(Saunders, 1965, 1966). Deep breaths in subjects
of the present study caused a small but nonsignificant increase in Pa,o, (Table 2). In normal
subjects (Wagner, Laravuso, Uhl & West, 1974)
and dogs with pulmonary oedema (Wagner,
Laravuso, Goldzimmer, Neumann & West, 1975)
0, breathing increases shunt, but the same group
found no increase in shunt on 0, breathing in four
asymptomatic asthmatic patients (Wagner,
Dantzker, Iacovoni, Tomlin & West, 1978) and
only small increases (mean 0.8% of cardiac
output) in 21 patients with chronic obstructive
pulmonary diseases (Wagner, Dantzker, Dueck,
Clausen & West, 1977), with only minor changes
in ventilation-perfusion distributions. All this work
refers to the effect of 100% 0, for 30 min, whereas
we gave about 60% 0, for 60 min. Assuming a
respiratory quotient of 0.83, an arteriovenous difference in 0, content of 5 m1/100 ml of blood, and
by use of the Severinghaus blood-gas calculator
(Severinghaus, 1966), the alveolar gas equation
and a standard shunt equation (Rahn & Fenn,
1955) we calculate that an increase in anatomical
shunt of about 8% would be required to decrease
Pa,o, from 49 to 44 mmHg (Table l), a very much
larger change than those recorded by the inert-gas
technique.
(2) Both in patients with chronic lung disease
(Pain, Read & Read, 1965;Eiser, Jones & Hughes,
1977) and in asthmatic patients (Field, 1967;
Valabhji, 1968) it has been shown that during 0,
breathing ventilation-perfusion matching is impaired, due to diversion of blood flow away from
well-ventilated to poorly ventilated lung units, pre-
sumably because hypoxic vasoconstriction is
abolished (Fishman, 1961;Cotes, Pisa & Thomas,
1963; Horsfield, Segel & Bishop, 1968). Could the
undershoot be caused by reversal of hypoxic vasoconstriction, with less appropriate ventilationperfusion matching which persisted on resumption
of air breathing? This seems an unlikely explanation since the time course of onset and
reversal of hyposic vasoconstriction in animal
preparations is rapid (see, for example, Fig. 8 of
Grant, Davies, Jones & Hughes, 1976) and it
seems quite unlikely that such an effect could
persist for 45 min after 0,breathing stopped.
(3) The subjects were rested and relaxed, and
cardiac output may well have fallen during the
course of the experiment. If 0, uptake was
constant, it is possible that the resulting fall in
mixed venous Po, might itself cause a fall in Pa,o,
(West, 1977).
(4) Finally, we may speculate whether simple
maintenance of a constant semi-recumbent posture
for 2 h might itself be accompanied by a progressive deterioration in ventilation-perfusion
relations. We know of no mechanism whereby this
might occur, but did not do the relevant airbreathing control experiment.
In summary, we cannot pin down an exact
mechanism from measurementof gas tensions alone
(and the first aim of this work was to demonstrate
whether the phenomenon of Pa,02 undershoot after
0, breathing was a constant one), but we find the
second and fourth hypotheses above unlikely, and
suggest that shunting at the alveolar level, perhaps
with a progressive fall in cardiac output and mixed
venous Po,, could account for our findings.
How long should one wait after discontinuing
supplemental 0, before taking a sample of arterial
blood? In our patients it took on average 15 min
for Pa,o, to return to baseline in patients with
chronic hypercapnic respiratory failure and 10 min
in patients with asthma, which agrees with the
results of studies specifically designed to answer
that question (Cugell, 1975; Howe, Alpert, Rickman, Spackman, Dexter & Dalen, 1975; Sherter,
Jabbour, Kovnat & Snider, 1975). In patients with
chronic airways obstruction (with or without
chronic CO, retention) Sherter er al. (1975) found
that after a short period of 100% 0, breathing it
took on average 20 min for Pa,o, to return to baseline. Howe el al. (1975), in patients with various
cardiac diseases but no lung disease, found that
Pa,o, values returned to control values within 7
min of stopping 0,. In neither of these studies was
Pa,o, followed for as long as 45 min after 0,
Blood gases during and d t e r oxygen breathing
breathing, but their figures suggest that Pa,o, was
still falling and that with longer observation undershoots like those we have seen might have been
observed.
Woolf (1959) suggested that the rate of fall of
arterial 0, saturation measured with an ear
oximeter could be used to diagnose emphysema,
and more recently concluded that it is reasonable to
take a blood sample in patients without chronic airways obstruction after 10 min and in patients with
this condition after 30 min (Woolf, 1976). We
agree with the timing of 10 min for the first group
of patients, but in the second Pa,o, would probably
have undershot at 30 min. Our practice is to take
arterial blood samples 15 min after cessation of 0,
therapy in patients with chronic hypercapnia.
Apart from one patient in group 1 whose arterial
0, saturation fell 16%, the undershoots we have
observed are small and possibly not of clinical
importance. All our patients were in a clinically
stable state at the time of study and it cannot be
assumed that our results necessarily apply to
patients with acute exacerbations of respiratory
failure, especially those with very low Pa,o, and
low pH, who seem particularly likely to get CO,
narcosis (Bone, Pierce & Johnson, 1978). On the
other hand, in such cases experimental
manipulations of inspired 0, tension of the type
used in this study would be completely unethical.
Perhaps in acutely ill patients who are more
severely hypoxaemic and thus on the steep slope of
the oxyhaemoglobin dissociation curve, small
undershoots in Pa,o, might be important, especially
if the patients are unable to improve 0, delivery to
the tissues by increasing their cardiac output.
In conclusion, we have shown that the
phenomenon of a rebound fall in arterial 0, tension
30-45 min after cessation of 0, breathing is a
common occurrence, and takes place irrespective
of the presence of CO, retention. We believe that it
is due to the persistence of areas of low ventilationperfusion ratio, brought about by the period of 0,
breathing. Short periods of up to 15 min off 0, are
not harmful, and indeed are recommended if it is
necessary to obtain blood-gas data representative
of the patient’s condition breathing room air. The
size of the undershoots are likely to be of major
clinical importance only in acutely ill patients.
References
BONE,R.C., PIERCE,A.K. & JOHNSON,
R.L. (1978) Controlled
oxygen administration in acute respiratory failure in chronic
obstructive pulmonary disease. American Journal of
Medicfne, 65,896-902.
BURGER,EJ. & MACUM, P.T. (1968) Ainvay closure:
demonstration by breathing 100% 0,at low lung volumes
395
and by N, washout. Journal of Applied Physiology, 25,
139-148.
CAMPBELL,
EJ.M. (19606 Respiratory failure: the relation
between oxygen concentrations of inspired air and arterial
blood. Lancer, U, 1&11.
CAMPBELL,
ELM. (1960b) A method of controlled oxygen
administration which reduces the risk of carbon dioxide
retention. Lancet, ii,12-14.
CAMPBELL,
E.J.M. (1967) The management of acute respiratory
failure in chronic bronchitis and emphysema American
Review of Respiratory Diseases, 96,626439.
COHN,J.E., CARROLL,D.G. & RILEY,R.L. (1954) Respiratory
acidosis in patients with emphysema. American Journal of
Medicine, 17,447463.
COTES, J.E., PISA, Z. & THOMAS,AJ. (1963) Effect of
breathing oxygen upon cardiac output, heart rate, ventilation,
systemic and pulmonary blood pressure in patients with
chronic lung disease. Clinical Science, 25,305-321.
CROPTON,
1. & DOUGLAS,
A. (1975) Respiratory Diseases, 2nd
edn., p. 375. Blackwell Scientific Publications, Oxford.
CUGELL,D.W. (1975) How long should you wait? Chest, 67,
253.
J.T. (1967) Effect of oxygen
CULLEN,J.H. & KAEMMERLEN,
administration at low rates of flow in hypercapnic patients.
American Review of Respiratory Diseases, 95,116-120.
CUMMING,
G. & SEMPLE,S.J.G. (1973) Disorders of the
Respiratory System, p. 209. Blackwell Scientific Publications,
Oxford.
DANTZKER,D.R., WAGNER,P.D. & WEST, J.B. (1975)
Instability of lung units with low V,/Q ratios during 0,
breathing. Journal of Applied Physiology, 3 4 88-95.
EISER,N.M.,JONES,H.A. & H U G ~ SJ.M.B.
,
(1977) Effect of
3096 oxygen on local matching of perfusion and ventilation in
chronic airways obstruction. Clinical Science and Molecular
Medicine, 53,387-395.
FIELD, G.B. (1967) The effects of posture, oxygen, isoproterenol, and atropine on ventilation-perfusion relationships in the lung in asthma. Clinical Science, 32,279-288.
FISHMAN,
A.P. (196 1) Respiratory gases in the regulation of the
pulmonary circulation. PhysiologicalReviews, 41,214-280.
GRANT,B.J.B., DAVIES,E.E., JONES,H.A. & HUGHES,J.M.B.
(1976) Local regulation of pulmonary blood flow and
ventilation-perfusion ratios in the coatimundi. Journal of
Applied Physiology, 40,216-227.
HORSFIELD,K., SEGEL, N. & BISHOP, J.M. (1968) The
pulmonary circulation in chronic bronchitis at rest and during
exercise breathing air and 8096 oxygen. Clinical Science, 43,
473-483.
D.G.,
How& J.P., ALPERT,J.S., RICKMAN,F.D., SPACKMAN.
DEXTER,L. & DALEN,J.E. (1975) Return of arterial Po,
values to baseline after suppkmental oxygen in patients with
cardiac disease. Chest, 67.256-258.
HUTCHISON.D.C.S., FLENLEY,D.C. & DONALD,K.W. (1964)
Controlled oxygen therapy in respiratory failure. Brirish
Medical Journal, ii, 1159-1 166.
P.C. (1962) Effect of
MASSARO,
DJ., KA- S. & LUCHSINGER,
various modes of 0, administration on the arterial gas values
in patients with respiratory acidosis. Brirish Medical Journal,
ii,627629.
MBDICALRESEARCH
COUNCIL(1965) Delinition and classification of chronic bronchitis for clinical and epidemiological
purposes. Lancet, I, 775.
PAIN, M.C.F., READ,DJ.C. & READ, J. (1965) Changes of
arterial CO, tension in patients with chronic lung disease
breathing 0,. Australasian Annals of Medicine, 14,195-204.
RAHN,H. & FENN,W.O. (1955) A Graphical Analysis of rhe
Respiratory Gas Exchange. The American Physiological
Society, Washington, D.C.
SAUNDERS,
K.B. (1965) Alveolar-arterial gradient for oxygen in
heart failure. Lancer, U, 160-162.
SAUNDERS,
K.B. (1966) Alveolar gas exchange in heart failure.
M.D. Thesis, Cambridge University.
SEVERINGHAUS,
J.W. (1966) Blood gas calculator. Journal of
Applied Physiology, 21,1108-1 116.
396
M . Rudovet al.
SHERTER,C.B., JABBOUR,
S.M.,KOVNAT,D.M. & SNIDER,G.L.
(1975) Prolonged rate of decay of arterial Po, following
oxygen breathing in chronic airways obstruction. Chest, 67,
259-26 1.
VALABHJI, P. (1968) Gas exchange in the acute and asymptomatic phases of asthma breathing air and oxygen. Clinical
Science, 34.43 1-440.
WAGNER,P.D., DANTZKER,D.R., DUECK,R., CLAUSEN,J.L.
& WEST. J.B. (1977) Ventilation-perfusion inequality in
chronic obstructive pulmonary disease. Journal of Clinical
Investigation, 59,203-2 16.
WAGNER P.D.. DANTZKER,D.R., IACOVONI,V.E., TOMLIN,
W.C. & WEST, J.B. (1978) Ventilation-perfusion inequality
in asymptomatic asthma American Review of Respiratory
Diseases, 1 1 4 5 11-524.
WAGNER,P.D., LARAVUSO,
R.B., GOLDZIMMER,E.,NAUMA",
P.F. & WEST, J.B. (1975) Distribution of ventilationperfusion ratios in dogs with normal and abnormal lungs.
Journal of Applied Physiology, 3 4 1099- 1109.
WAGNER,P.D., LARAVUSO,R.B., UHL, R.R. & W e s s J.B.
(1974) Continuous distribution of ventilation-perfusion ratios
in normal subjects breathing air and 100% 0,. J o u m l of
Clinical Investigation, 54,54-68.
WEST, J.B. (1977) Ventilation-perfusion relationships.
American Review of Respiratory Diseases, 116.919-943.
WOOLF,C.R.(1959) The diagnosis of emphysema: a physical
sign, a roentgenographic sign, and an oximeter test. American
Review of Respiratory Diseases, 80,705-715.
WOOLF, C.R. (1976) Arterial blood gas levels aRcr oxygen
therapy. Chest, 69,808.