Professional Documents
Culture Documents
THESIS DOCTOR OF PHILOSOPHY KORMAN Bengt 2020
THESIS DOCTOR OF PHILOSOPHY KORMAN Bengt 2020
Doctor of Philosophy
ii Supervisors
KEYWORDS
Keywords iii
iv Keywords
ABSTRACT
There are over 4 million alveoli in the adult lung. They are vital to human
life. Uptake of oxygen and elimination of carbon dioxide takes place in these key
structural elements. Gas exchange occurs between alveolar gas and pulmonary
capillary blood and is controlled by the gas/blood flow-ratio. This ratio, termed the
ventilation-perfusion ratio, is normally about 0.8 for the lung as a whole but varies
from alveolus to alveolus, with a spread of values throughout the lung. Past research
indicates that gas exchange will be most efficient if this ratio is the same in each
alveolus and equal to the value for the lung as a whole. A lung with this property is
physiology is that gas exchange can never be more efficient than in a homogeneous
lung.
investigations. These findings were ignored by many and disputed by at least one
expert in the field who preferred to look for alternative explanations. These reactions
are not surprising, given that the claims appear to violate the fundamental principle
Abstract v
Gas exchange during anaesthesia was modelled by adapting techniques of
mathematical results were further refined for use in individual scenarios. Solutions
solving capabilities. The aims of this research fell into three parts as follows:
PART I
simulations did not contain undetected errors which might have led to false
predictions.
PART II
the previous work on this subject did not take into account the existence of the
new phenomenon.
PART III
Clearly, parts two and three would only be relevant if the 2001 findings were
vi Abstract
The 2001 computer predictions were independently confirmed. This research
showed that the Minimal Alveolar Concentration or MAC reading does not
accurately reflect the depth of anaesthesia during nitrous oxide anaesthesia when
combined with modern volatile anaesthetics. MAC is used to avoid awareness during
anaesthesia and was the only method used widely for this purpose prior to the
results of the formal mathematical analysis needed revision. One unexpected finding
was the possibility of gas uptake seemingly taking place against a partial pressure
gradient.
of gas exchange, namely that it can be represented as the sum of two steps - step one:
gas exchange at constant volume, followed by step two: gas exchange on volume
contraction in volume which occurs during nitrous oxide anaesthesia. Thus, they
the amount of gas presented to step two after uptake in step one. By improving
its counterintuitive effect on gas uptake in a way that is perfectly compatible with
classical teaching.
principle and created a gap in our understanding of gas exchange in the lung. This
gap has now been filled in a way that does not violate the underlying principle. The
research has uncovered several hitherto unrecognized properties of gas exchange and
Abstract vii
led to the discovery of a clinically important feature of nitrous oxide anaesthesia
which can result in patients being too deeply anaesthetized during surgery.
viii Abstract
TABLE OF CONTENTS
: Introduction ........................................................................................... 1
Background .....................................................................................................................1
Context............................................................................................................................3
Purposes ..........................................................................................................................3
Significance, Scope and Definitions ...............................................................................4
Thesis Outline.................................................................................................................6
: Background ............................................................................................ 9
Kety’s theory ..................................................................................................................9
Anaesthetic uptake in peripheral tissues .......................................................................10
Eger’s mathematical model ..........................................................................................11
The second gas effect....................................................................................................13
An additional explanation for the second gas effect .....................................................14
The Eger-Stoelting diagram ..........................................................................................16
The extra-inspired ventilation .......................................................................................18
A matter of relativity ....................................................................................................19
Respiratory pattern during N2O anaesthesia .................................................................19
Respiratory pattern and the Eger-Stoelting diagram.....................................................22
Should Eger have known? ............................................................................................24
A challenge to the validity of the second gas effect .....................................................27
Ventilation-perfusion mismatch appears as a factor .....................................................30
The Severinghaus objection ..........................................................................................31
Summary and Implications ...........................................................................................33
: The Log-Gaussian Distribution .......................................................... 39
Introduction ..................................................................................................................39
Choice of the log normal distribution ...........................................................................40
Correct orientation of graphs ........................................................................................40
The meaning of σ ........................................................................................................40
Measurements of σ in adult subjects ...........................................................................41
The compartmental approach .......................................................................................41
V and Q as explicit functions of log(V/Q)....................................................................43
v and q as explicit functions of (1/σ)log(v/q) ...............................................................46
The case when σ = 0 ....................................................................................................49
Table of Contents ix
Interpretation of gas exchange using Figure 3.4 .......................................................... 49
Interchangeability of x-variables .................................................................................. 50
Lebesgue-Stieltjes integration ...................................................................................... 51
Summary ...................................................................................................................... 53
: Steady-state Equations ........................................................................ 55
The non steady-state model.......................................................................................... 55
The steady-state model................................................................................................. 57
A suitable model to study the second gas effect (SGE) ............................................... 59
Summary ...................................................................................................................... 63
: Persistent Second Gas Effects............................................................. 65
The washin ratio ........................................................................................................... 66
The augmentation ratio ................................................................................................ 67
Simulating persistent second gas effects ...................................................................... 72
Advantages of the mathematical model ....................................................................... 75
Relationship to experimental work .............................................................................. 75
Clinical implications .................................................................................................... 77
Summary and conclusion ............................................................................................. 78
: Partial pressure-solubility diagrams .................................................. 79
The race for MIGET .................................................................................................... 79
Equations...................................................................................................................... 81
The partial pressure-solubility diagram ....................................................................... 86
The augmentation-solubility relationship .................................................................... 92
Alveolar-arterial partial pressure differences ............................................................... 94
The uptake-solubility relationship................................................................................ 98
Partial pressure reversal during gas uptake .................................................................. 99
Alveolar-arterial partial pressure gradient for N2 ....................................................... 102
Conclusion ................................................................................................................. 109
: Elucidating relevant roles ................................................................. 111
Steady-state gas exchange as a two step process ....................................................... 111
Equations for uptake of first gas & diluent gas .......................................................... 116
Equation for inspired ventilation................................................................................ 116
Equation for uptake of second gas ............................................................................. 117
Uptake sequence with a first and second gas ............................................................. 118
Derivation of distributions for diagrams .................................................................... 118
Distributions of inflow, outflow and uptake .............................................................. 122
Second gas uptake as a two-step process ................................................................... 125
Role of the solubility of the second gas ..................................................................... 128
x Table of Contents
Role of V/Q mismatch in relation to the second gas ..................................................129
Effect of V/Q mismatch in relation to the first gas .....................................................130
A third-order contribution to SG uptake .....................................................................132
Location of uptake in the V/Q spectrum.....................................................................133
Effect of ventilation-dependent FG on SG uptake......................................................136
The concentration effect .............................................................................................138
Absorption of gas in airways rather than alveoli ........................................................141
An alternative definition of the Augmentation Ratio .................................................142
Conclusion ..................................................................................................................143
: Influence of Respiratory Pattern...................................................... 145
Introduction ................................................................................................................145
Augmentation Ratios for constant inflow ...................................................................145
The two-step model applied to constant inflow ..........................................................148
Conclusion ..................................................................................................................155
: Conclusions ........................................................................................ 157
Appendix A: Derivation of distribution functions for V and Q ......................... 165
Appendix B: Another version of Kety’s Eqs........................................................ 169
Appendix C: Sample MATLAB routines............................................................. 179
Appendix C.1: ROUTINES USED TO DRAW Figures 6.1, 6.2 and 6.3 .............................179
Appendix C.2 ROUTINES USED TO DRAW Figure 7.4 & Figure 7.5..............................190
Appendix C.3 ROUTINES USED TO DRAW Figure 7.7 ...................................................195
Appendix C.4 ROUTINES USED TO DRAW Figure 8.2 ...................................................198
Bibliography ........................................................................................................... 205
Table of Contents xi
xii Table of Contents
LIST OF FIGURES
Many of the figures in this thesis have been prepared with the help of the Department
of Medical Illustration, Royal Perth Hospital. In spite of their expert efforts, some of
the figures exhibit discontinuities and/or pixilation when viewed on a computer
screen. These effects may be reduced or eliminated by trying a different page
magnification e.g. 75% or 100%.
List of Tables xv
xvi
LIST OF ABBREVIATIONS
General variables:
I Inspired gas
E Expired gas
A Alveolar gas
B Barometric
a Arterial
v Mixed venous
Common abbreviations:
FG First gas
SG Second gas
CE Concentration effect
SGE Second gas effect
these were necessary to facilitate the mathematical treatments. They include the
following:
1. Oxygen and carbon dioxide are excluded from the models because
considered.
thesis.
Publications xix
These publications are appended to the end of this thesis. Permission to include
the first manuscript was granted by ANESTHESIOLOGY and Wolters Kluwer Health
Inc. (publishers of the journal and copyright owners). An editorial on the article
website (Kennedy RR. A Second Look at the Second Gas Effect. Anesthesiology
xx Publications
STATEMENT OF ORIGINAL AUTHORSHIP
The work contained in this thesis is my work, has been completed during
the course of my degree and does not breach any ethical rules with regard to the
conduct of the research. It has not been previously submitted to meet
requirements for an award at this, or any other higher education institution.
This thesis does not violate or infringe any copyright, trademark, patent,
or other rights whatsoever of any person.
This thesis contains published work and work prepared for publication
which has been co-authored.
Alexander P Robertson
xxii Mentors
ACKNOWLEDGEMENTS
mind, humility, kindness, generosity, and a keen sense of humour. Ian had
Ian’s boundless energy and enthusiasm inspired all those who came into contact
(35), one of the great pioneers of anaesthesia as a science. Bill helped me clarify
abstract ideas that led to the publication of two papers in joint name in the British
Journal of Anaesthesia.
Sadly, all three of these great teachers have now passed on. This thesis
Acknowledgements xxiii
Other people influenced my decision to undertake this research and
for the degree of PhD at the university. This proved impossible because I was
still working in full-time private practice in Perth and could therefore not spend
Physiology.
Supervisor. Even though the work was somewhat outside his ambit, he
I could not have wished for a more supportive person in this role.
xxiv Acknowledgements
Dr Barry Marshall, Professor of Medicine at UWA, Nobel Laureate in
Medicine in 2005, graciously agreed to lend his name as Local Supervisor with
examination. This requirement was not satisfied by any of the other Local
Supervisors.
patience and understanding this project would not have been possible.
Acknowledgements xxv
xxvi Acknowledgements
: Introduction
BACKGROUND
the Royal Perth Hospital. I chose to talk about a subject that had formed part of
to explain certain aspects of gas uptake during anaesthesia. These were the
complicated message so clearly. But over time, I began to feel there was
something wrong with the diagram. My suspicions were confirmed when Ian
Eger’s diagram was wrong, then what was the correct diagram? It took me many
years to sort out the puzzle. After publishing a joint paper on the subject with
Professor Mapleson in 1997 (48), Eger’s diagram virtually disappeared from all
textbooks except those written in the United States, his home country.
the same subject in the Journal of Applied Physiology, in which he claimed that
Chapter 1: Introduction 1
mismatching of ventilation and blood flow can lead to improved gas uptake (68-
from Peyton himself and another group based in Belgium (34, 67, 71). Being
busy in a full-time private practice I had not kept up with the literature and was
in the lung, should be perfectly matched to the blood flowing through the lung
(26). If the ventilation is for example, 4 l/min and the pulmonary blood flow is,
say, 5 l/min, this entails the ratio of gas flow to blood flow to be 4/5 or 0.8 in
each gas-exchanging unit throughout the lung. Here was someone claiming the
opposite can be true, that a worsening in the matching of ventilation and blood
flow throughout the lung can actually improve gas exchange. I found this quite
prestigious as the Journal of Applied Physiology could have allowed such heresy
to appear in its pages. I knew only too well how difficult it is to have work appear
I believed that Peyton may have erred either in his original assumptions or
in his computer simulations. I therefore set out with the aim of replicating his
steps and locating his mistake. This would require an algorithm to deal with the
Peyton had included these physiological gases in his model. The algorithms he
used were those of Kelman, published in the 1960s (38, 39). Surely a more up-
to-date algorithm was now available. A quick internet search yielded a paper by
Dash and Bassingthwaighte (13). At first, I struggled with the chemistry and
2 Chapter 1: Introduction
maths in the paper but after reading it several times, I realized that it was quite
became apparent that Peyton’s findings could not be rejected. When I discussed
this with him, Phil suggested I enrol for the degree of Doctor of Philosophy with
CONTEXT
The major focus of this study is the proposition that gas uptake can be
PURPOSES
Chapter 1: Introduction 3
1. To develop a suitable model for investigating the concentration & second
2. To use this model to study the effects on the CE & SGE of varying solubility
The widespread clinical use of nitrous oxide led to the initial discovery of
the concentration and second gas effects. However, the use of this agent has
Nitrous oxide has traditionally been regarded as a very safe agent, having
been used in anaesthesia for over 150 years. Minor issues were considered to be
the possible depletion of vitamin B12 stores associated with long-term exposure
especially in subjects with vitamin B12 deficiency (28), and nausea and vomiting
(15, 32, 87). The ENIGMA trial, published in 2007 (62), led to the conclusion
that the administration of nitrous oxide was associated with increased risk of
Although these fears were not borne out by the larger and more powerful
4 Chapter 1: Introduction
and new trainees in anaesthesia were taught to avoid it wherever possible. The
environmental impact is clearly of concern given that the gas is said to have an
effect on the ozone layer, and a contribution to global warming some 300 times
that of carbon dioxide on a weight for weight basis (77). Moreover, the effect is
1990s regarded the contribution from medical use to be very small (11). The gas
was included in the 2017 WHO List of Essential Medicines and is still available
nitrous oxide is still used widely in operating theatres, labour wards and dental
practices.
is my view that the awkward facts discovered by Peyton cannot merely be swept
under the carpet. They challenge a basic tenet of respiratory physiology and
truth” and if the respiratory physiologists are not prepared to investigate its
cause, then surely it falls to the anaesthetic fraternity to accept responsibility and
The approach used in this thesis is purely theoretical. Since using gas
part of the research for my MD thesis, I have not been involved in any
constitutes a separate experiment and that the skills involved might easily be
Chapter 1: Introduction 5
translated into experimental research. Each of us must, however, recognise our
other hand, over recent years my fascination with mathematics has grown.
Having been exposed to the rigorous development of the theory of linear algebra
and of series in second year Pure Mathematics at honours level at the University
when applied appropriately. What seemed dull and uninteresting in 1966 became
important and relevant in tackling the problems that arose repeatedly. In regard
to the matter at hand, I would argue that the important experimental work had
already been done by Peyton, when he confirmed his prediction that the second
gas effect is greater in blood than in the gas phase and that this is due to the
anaesthesia. In my view, what was required now was not more experimental
work, but a satisfactory explanation of the cause. This could only come from
theory.
THESIS OUTLINE
to explain important aspects of gas uptake was fundamentally flawed but was
uptake. The history of this proposition is traced from inception to the time I first
6 Chapter 1: Introduction
Chapter 5 addresses the issue of persistent second gas effects, claimed to
exist by several workers in the area. I set out to confirm or deny the existence of
these effects.
explanation for the effect of V Q mismatch observed clinically with the low
are usually introduced at the start of the relevant chapter. When the theory begins
with the same assumptions as in previous chapters, the assumptions are often re-
stated so as to avoid the need to move backwards and forwards in the thesis. It
is hoped that this will smooth the reader’s task and prevent unnecessary
Chapter 1: Introduction 7
: Background
the lungs and tissues (42). He identified factors which determine the rate of gas
KETY’S THEORY
Expressed using symbols with which we are familiar today, Kety produced
the following equation for the rate of change with time of the alveolar partial
pressure of an inert gas i.e. a gas which merely dissolves in blood and does not
dPA
VL = VA ( PI − PA ) + λQ t ( P v − Pa ) (2.1)
dt
where,
Chapter 2: Background 9
Although Kety realized that the body tissues could not be treated as a
homogeneous tissue mass, the simplest solution was achieved by making this
assumption with respect to blood flow and solubility, together with the
assumptions that equilibrium between blood and tissue is complete and that the
gives rise to an equation of the following form for events at the tissues:
dP v
VT= Q t ( Pa − P v ) (2.2)
dt
In the period between 1951 and 1963 anaesthetic interest was directed at
cited by Kety, allowed for the division of the body into a number of
of 1 as Kety had done. The blood flow per unit of tissue volume was the same in
10 Chapter 2: Background
analogue. Although several analogues were presented at a gathering of experts
first electric analogue was that of Mapleson (57). His paper was submitted to the
prestigious Journal of Applied Physiology on 4th December 1961 but was not
published until January 1963! At least one of the other participants who
presented an electric analogue at the 1962 conference was also a Reviewer for
the journal. One wonders how this coincidence would be viewed today.
and distribution of anaesthetic gases (18, 21). We can summarize the relevant
VL • Cg n −1 + VI • FI − un + FI • un
Cg n = (2.4)
VL
+u
Cbn −1 • QT
Cbn = n
(2.5)
QT
Cbn
λ= (2.6)
Cg n
where:
Cg n = concentration of anaesthetic agent in gas phase at the end of the nth breath
1
Eger included corrections for temperature, humidification and volume of lung tissue but these are not
critical to the discovery of the concentration effect and resulted in his model becoming unnecessarily
complicated. We do not include them in this thesis as inspired gas is considered to be fully humidified
at 37 deg C and we use the fractional concentration in dry gas as our measure of concentration. Note
that events in VL , the lung volume, are irrelevant in the steady-state model of gas exchange used to
study the concentration and second gas effects in this thesis (see Chapter 4).
Chapter 2: Background 11
Cbn = concentration of anaesthetic agent in blood at the end of the nth breath
The numerator of Eq. 2.4 states that the volume of the anaesthetic gas
present in the lung at the end of the nth breath is equal to the volume of
anaesthetic gas in the lung at the beginning of the nth breath plus the volume of
the anaesthetic gas brought in during the breath minus the volume of anaesthetic
gas absorbed by equilibration with blood plus the volume of extra fresh gas
Eq. 2.6 is the form of Henry’s Law commonly used in Anaesthesia and
Respiratory Physiology. A gas which obeys this law is said to be an inert gas.
From his model, Eger predicted that FA FI would rise more rapidly for higher
particularly if the agent is very soluble in blood. In Figure 2.1, the effect is seen
to be greater for the more soluble agent ether, than for nitrous oxide 3.
2
This is the functional residual capacity or FRC.
3
For copyright reasons, this diagram and several others in this thesis are simulated using inputs from
the original publication in a new computer model described in Appendix B.
12 Chapter 2: Background
Figure 2.1. The concentration effect.
Alveolar concentration as a fraction of inspired concentration for inspired concentrations
of 1%, 40%, 75% & 100% N2O and 1%, 40% and 75% ether during washin. From a
model described in Appendix B with inputs based on the original material by Eger (18).
mixtures of nitrous oxide and halothane, Epstein et al. (24) showed that when
concentration of halothane was the same in both cases, FA FI rose more rapidly
in the presence of 70% nitrous oxide – the second gas effect. This effect was
secondary to the absorption of nitrous oxide at higher concentrations i.e. the term
FI • un in Eq. 2.4 above. The effect was subsequently documented for oxygen
Chapter 2: Background 13
Figure 2.2. The second gas effect.
Simulation of the experiment by Epstein et al. (24) using a model described in Appendix
B, with inputs based on the original investigation in which 0.5% halothane was
administered to dogs anaesthetized with pentobarbital using one of two possible
anaesthetic mixtures. One mixture included 70% nitrous oxide, the other, 10% nitrous
oxide. 4
inspired gas composition was then changed abruptly to a mixture containing 70%
nitrous oxide, the equilibrium concentration of the second gas and oxygen. FA
for the second gas was then observed to rise above FI (see Figure 2.3). The
authors argued that an increase in the inspired ventilation could not explain these
4
The original article may be viewed at:
https://anesthesiology.pubs.asahq.org/article.aspx?articleid=1966256
14 Chapter 2: Background
Figure 2.3. The additional explanation for the second gas effect.
Simulation of experiment by Stoelting and Eger (84) using a model described in Appendix
B with inputs based on the original investigation in which dogs were equilibrated with
low concentrations of ethylene (blue, λ = 0.14), cyclopropane (green, λ = 0.415) and
halothane (red, λ = 2.3) in oxygen. At the time indicated as zero, the inspired gas
composition was abruptly changed to a mixture of 70% nitrous oxide containing the
previous concentration of second gas. The balance of the inspired gas mixture consisted
of oxygen. The alveolar concentration of each second gas was then observed to rise above
the inspired concentration. Ventilation was controlled with a volume-limited ventilator. 5
results as it would oppose the rise of FA above FI. They postulated that an
5
The original article may be viewed at:
https://anesthesiology.pubs.asahq.org/article.aspx?articleid=1965202
Chapter 2: Background 15
THE EGER-STOELTING DIAGRAM 6
The revised diagram was now used to explain both the concentration
and second gas effects. Since its first appearance in ANESTHESIOLOGY in 1969 it
follows:
“The hypothetical lung initially contains 80% nitrous oxide, 19% oxygen
and 1% second gas. A – the concentrating effect. If half the nitrous oxide is taken
up, the remaining second gas now represents 1.7% of the total gas volume, while
before it represented only 1%. Consequently, the second gas has been
maintain lung volume. The inflowing gas contains the same proportions of
nitrous oxide, oxygen and second gas as the gas originally present. Although
this additional
6
The original article containing this diagram may be viewed at
https://anesthesiology.pubs.asahq.org/article.aspx?articleid=1965202
16 Chapter 2: Background
ventilation increases the nitrous oxide concentration from 66.7% to 72%, it
No doubt Eger believed the diagram to accurately reflect the steps in his
mathematical model. Sadly, as we will now demonstrate, this was not the case.
Notice that as the rule for equilibrating nitrous oxide with blood, Eger is
postulating that half the nitrous is taken up. However, the rule is only applied to
the initial gas. The extra gas brought in to replace the nitrous oxide taken up by
blood is left unequilibrated in the lung. Looking at Eq. 2.4, we can see that the
with blood in Eq. 2.6. Therefore, the diagram and the model are NOT the same.
Moreover, the notion of gas sitting for any length of time in the
inert gas exchange, a well-known fact by 1969 (9, 30). Equilibration with
blood leads to the situation shown in Figure 2.5 (44). The starting situation is
shown on the left of the figure. Now however, the extra-inspired ventilation has
proper equilibration with blood according to Eger’s formula that half the
nitrous oxide be taken up, 2/3 V of the nitrous oxide disappears into blood
leaving 2
/3 V behind in the gas phase.
Chapter 2: Background 17
Figure 2.5. Correctly equilibrated version of the Eger-Stoelting diagram.
Reproduced in modified form from Korman (44).
From the time he first gave his explanation of the concentration and second
gas effects, Eger routinely invoked the twin mechanisms of the concentrating
effect and the extra inspired ventilation. Presumably because the concentration
effect was greater with the more soluble agent ether than with nitrous oxide 7, he
postulated that the extra-inspired ventilation is more important for more soluble
agents, while based on the findings shown in Figure 2.3, the concentrating effect
is more important for less soluble agents. Thus, the extra-inspired ventilation
7
This depends on how one measures the concentration effect. In 7.15, we show that when inspired
concentrations are chosen so as to produce a similar contraction in gas volume, the effect is actually
greater for the less soluble agent, nitrous oxide. This involves using 25% ether in a comparison with
70% nitrous oxide.
18 Chapter 2: Background
A MATTER OF RELATIVITY
gas exchange was the subject of many conversations between Ian Ritchie and
me. Ritchie pointed out that it depends where the observer is located. If the
with a gaseous agent, the net movement of anaesthetic gas is into blood. The
observer stationed there will perceive that more anaesthetic gas is drawn down
the airway to replace that taken up by blood. This must always occur at the
interruption.
membrane. We are standing beside our patients and observing events from the
considered.
Kety (42), Mapleson (57) and Eger (21) all treated the functional residual
in volume due to gas uptake can then only be reflected in such models by
differences between the inspired and expired ventilation. In particular, VIA , the
inspired alveolar ventilation, must exceed VEA , the expired alveolar ventilation.
8
These terms were first suggested by Professor Alex Robertson.
Chapter 2: Background 19
2.9.1 CONSTANT INFLOW
In this pattern, the inspired tidal volume is kept constant and the expired
tidal volume allowed to vary, reflecting the gas uptake during the breath.
same for each breath. Expiratory volume varies, depending on the net gas
volume uptake during that breath. The first pair of lines shows the situation
before the anaesthetic gas is introduced when the difference in volume is due to
become apparent as large volumes of nitrous oxide are taken up by blood during
and the expired tidal volume slowly returns to its pre-induction value. The
20 Chapter 2: Background
2.9.2 CONSTANT OUTFLOW
At the other extreme, we have a constant outflow pattern in which the
expired tidal volume is fixed, and the inspired tidal volume allowed to vary.
In Figure 2.7, the second line in each pair represents the constant expired
tidal volume. To maintain the expired tidal volume constant, it is necessary for
the inspired tidal volume to exceed it. The amount by which the inspired tidal
volume exceeds that expired is the volume of nitrous oxide uptake during that
breath. This volume is indicated by the stippled area and may be thought of as
being drawn into the airway to maintain the sum of the FRC and expired tidal
volume at a constant value. We can therefore identify this gas as Eger’s extra-
Chapter 2: Background 21
2.9.3 EGER’S CONSTANT INFLOW MODEL
It is important to note that while Eger produced equations of the form of
Eqs. 2.4 - 2.6 for the constant outflow case (which he equated to a nonrebreathing
system 9), he also produced the following equation for the constant inflow case:
VL • Cg n −1 + VI • FI − un
Cg n = (2.7)
VL + VI • FI − un
Equations 2.5 and 2.6 were applied as before. On rearrangement, this system of
equations gives rise to a quadratic equation which is solved for un . In this system
also, the higher the inspired concentration, the more rapid the approach to the
final concentration but not as fast as with Eqs. 2.4 - 2.6 (18). Eger attributed the
reduction in rate to the limited inflow associated with the use of a circle system 10.
Once gas in the lung is fully equilibrated with blood, we may draw the
difference between the two extreme respiratory patterns (48). In Figure 2.8, we
have again applied the rule that half the nitrous oxide be taken up. Note that the
constant inflow case is associated with a smaller uptake of nitrous oxide than the
equivalent constant outflow case (160 ml compared with 267 ml). This is a
9
This is incorrect as a non-rebreathing system used with a constant volume ventilator actually has a
constant inflow pattern.
10
Again, this is wrong since a subject breathing spontaneously on a circle system with an adequate
fresh gas flow will have a constant outflow pattern of respiration.
22 Chapter 2: Background
Eger. When we apply the rule that half the nitrous oxide be taken up, the final
Once the FRC is fixed, the only possible patterns are constant inflow,
constant outflow and all combinations thereof. Since the concentration effect
occurs in both extreme cases, it follows that it must occur in all combinations
Figure 2.8. Constant inflow and constant outflow using Eger’s diagram.
A-B: Inspired tidal volume is kept constant and equal to 400 ml. C-D: Expired tidal
volume is kept constant and equal to 400 ml. BE = Before Equilibration; AE = After
Equilibration. The gradations on each side of the y-axis occur at intervals of 20% of the
inspired tidal volume for each case, that is at intervals of 80 ml for the constant inflow
case and 133.3 ml for the constant outflow case. Reproduced with permission from
Korman & Mapleson (48).
the concentration and second gas effects. Instead we must look for some other
property common to both constant inflow and constant outflow. That property,
evident in all our figures from Figure 2.4 to Figure 2.8, is the shrinkage in volume
and the associated concentrating effect. Thus, we conclude that the concentration
Chapter 2: Background 23
and second gas effects are ALWAYS caused by the concentrating effect that
Eger’s presentation at the 1962 conference chaired by Papper and Kitz was
he actually presented the constant inflow and constant outflow patterns. He went
Strangely, neither Eger nor Rackow seems to have made the connection.
For year after year, Eger went on explaining the concentration effect repeatedly
Rackow finally described the situation during induction of anaesthesia with the
suspicious of the diagram and asked Ian Ritchie if he could see any flaw in the
reasoning. He came back to me after 2 weeks and pointed out the failure to
equilibrate the extra-inspired ventilation. It took many more years to “join the
dots” and I finally presented a talk entitled “The vacuum theory of anaesthesia”
24 Chapter 2: Background
on the message conveyed by Eger’s diagram, the message that there is always
question “How can this be when most of us use a constant volume ventilator
during anaesthesia?”. Edmond Eger and John Severinghaus 11 were both present.
By this time, I had approached Bill Mapleson for help to bring the
inaccuracy in Eger’s diagram to the attention of both teachers and authors. The
next year our collaboration finally bore fruit when our article entitled
textbooks published outside the United States. Within the United States,
however, the diagram continued to be produced regularly by Eger until his death.
One of the few concessions he made to our criticism is contained in his chapter:
Well-known Anaesthesiology research worker & inventor of the Severinghaus PCO2 electrode (83).
11
Chapter 2: Background 25
Of course, the exact opposite is true. The situation is much simpler –
smaller volume; this increases their partial pressures and accelerates their uptake.
The only complexity has been introduced by Eger himself in persisting with a
diagram that is wrong. So confused has the picture become that Calvey and
Anaesthetists seem too scared to actually offer an explanation. Instead they state:
uncritically. Sometimes the diagram is disguised but the explanation remains the
Other times the diagram remains unchanged, but the explanation is altered
as in this extract from the chapter by Ebert and Naze (17) in Clinical Anesthesia:
26 Chapter 2: Background
“In this hypothetical example, the second gas is set at 2% of a potent
anesthetic and the model is set for 50% uptake of the first gas, nitrous oxide
in the first inspired breath. The second gas is concentrated because of the
uptake of nitrous oxide (middle panel). On replenishing the inspired
second gas in the next breath, the second gas has been concentrated to be
2.7% because of the uptake of nitrous oxide in the previous breath.”
explanation, the second gas concentration starts at 2% in the first panel, rises to
3.1% in the second and finishes up at 2.7% in the third. Note however, that the
dilution in the third panel is now ascribed to the next breath implying that one
Surely this is a classic case of trying to make the facts fit the theory! One can
only feel pity for the poor anaesthetic trainee trying to make sense of such
gobbledygook.
The Eger-Stoelting diagram illustrates exactly what can happen when the
picture is wrong!
that the second gas effect is not a valid concept (85). They studied the effects of
Chapter 2: Background 27
enflurane and on its blood concentration in 14 young healthy males (7 subjects
in each group). They reported that nitrous oxide did not affect the end-tidal or
and that the second gas effect is therefore not a valid concept.
28 Chapter 2: Background
In response, Lee and Sun 13 made the following points (50):
2. They did not believe that they had missed marginal differences or made a
Type II error.
The message agreed on by both sets of authors seemed to be that the second
gas effect is not important clinically. 14 However there remained a clear point of
difference in that Lee and Sun maintained that the second gas effect is difficult
to verify.
The same year, Taheri and Eger demonstrated a significant second gas
effect for desflurane in 65% nitrous oxide and argued against the conclusion
for oxygen in the presence of 66.7% nitrous oxide with the average PaO2
increasing from 172 mmHg to 201 mmHg (65). It was in this setting that the
13
The full response may be seen at:
https://journals.lww.com/anesthesia-
analgesia/Fulltext/1999/11000/The_Second_Gas_Effect_Is_a_Valid_Concept.51.aspx
14
Although this conclusion needs to be revised given the events that followed.
Chapter 2: Background 29
VENTILATION-PERFUSION MISMATCH APPEARS AS A FACTOR
Robinson, one of the Staff Anaesthetists, who was investigating a new non-
throughflow technique (80). This was being compared with the standard
thermodilution method. Robinson had identified that A-a gradients for nitrous
oxide might seriously affect the accuracy of the method. It was decided to
This work ultimately became the subject of Peyton’s MD thesis (66). Data from
the model predicted unexpected findings in relation to the second gas effect
Physiology (68-70).
model with a log normal distribution of V Q ratios and investigated the effect
to mimic the values seen during anaesthesia. The results for the two-
third of the papers, it was shown that concentration and second gas effects may
15
The Abstracts may be viewed at: https://journals.physiology.org/toc/jappl/91/1
30 Chapter 2: Background
persist even as nitrous oxide uptake falls to the low levels reached after the initial
rapid rise of FA FI has been completed and attributed the persistence to the
increase in V Q mismatch.
Peyton directed his efforts at investigating the likely effects on the arterial
oxygen partial pressure - clearly of importance clinically. Thus, the models were
nitrogen, nitrous oxide and a volatile anaesthetic agent were included in the
analysis. Constant inflow and constant outflow models were studied. Absorption
which 30% oxygen was administered either with 70% nitrogen or 70% nitrous
oxide and the change in PaO2 measured using patients as their own controls (71).
second gas effects. Similar findings were obtained by Hendrickx et al. when
commenting on the article by Hendrickx et al. (34) in which the claim is made
that large volume N2O uptake alone cannot explain the second gas effect on
Chapter 2: Background 31
2. This casts suspicion on the accuracy of their analytical methods which
inhomogeneity might explain this effect, and provide some background evidence
uptake of sevoflurane did not sit well with Severinghaus, who preferred to look
for other ways of explaining the results. Moreover, this is not surprising because
it is not at all obvious how such mismatch could actually facilitate the increased
uptake. This then, was the status quo when I first became aware of Philip
Peyton’s work.
16
The full correspondence may be viewed at:
https://bjanaesthesia.org/article/S0007-0912(17)35134-6/fulltext
32 Chapter 2: Background
SUMMARY AND IMPLICATIONS
1. The Eger-Stoelting diagram is often used as a teaching tool. The steps in the
diagram do not match those taken in the mathematical model used by Eger to
The concentration and second gas effects are best explained as follows:
Uptake of significant volumes of gas concentrates each remaining gas in a smaller volume; this
increases its partial pressure and accelerates its uptake.
Chapter 2: Background 33
2. Whenever large volumes of gas disappear into blood during gas uptake a
significant difference may arise between the inspired and expired tidal volume
of each breath. When modelling anaesthetic uptake with a fixed FRC, two
outflow. Each of these exhibits the concentrating effect shown in Figure 2.8.
effect. 17
perfusion mismatch can amplify the concentrating effect of gas volume uptake
produced by the absorption of nitrous oxide. This may lead to the persistence
of significant concentration and second gas effects even when the net gas
the way forward in the research for this thesis now became clear. Its focus
17
Furthermore, it is not possible to have constant outflow in the absence a concentrating effect in
some form or other.
34 Chapter 2: Background
2.15.1 INDEPENDENT VERIFICATION
important. From personal experience with the various methods of analysis – gas
finicky and time consuming they can be. They require a degree of skill which I
never quite acquired. Hence, I am somewhat reluctant to place too much reliance
on experimental results based on the use of these techniques. On the other hand,
in-law Dr John Kessell (inventor of the Kessell laryngoscope blade for difficult
intubations). As a result, I know only too well how easy it is to introduce a bug
in computer programs and how difficult it can be to locate and eradicate it. Could
it be that Philip Peyton had inadvertently made a mistake which had gone
undetected? While Phil would undoubtedly have given me access to his program,
I felt that I should tackle the problem afresh in order to avoid overlooking any
In his model, Phil included oxygen and carbon dioxide as well as nitrogen,
nitrous oxide and a volatile second gas. This is understandable given that he was
studying the second gas effect on oxygen, but it added factors that may lead to
make it difficult to draw general conclusions. From previous work with Jim
Bassingthwaighte and Ranjan Dash, I was only too aware of the complex
preferred to omit these gases if possible. The distinction between constant inflow
and constant outflow seemed less important once the universal applicability of
Chapter 2: Background 35
the concentrating effect became obvious, and absorption atelectasis and hypoxic
in the inspired mixture. This supposition has served numerous previous authors
well, including Kety (42), Mapleson (57), Eger (18, 21), West (89-91), Evans et
al. (26), Colburn et al. (10), and Kelman (40). I could see no reason to avoid it.
The San Diego team under the leadership of John West had produced many
assumption 18.
who joined West’s team, had done much to formalize the mathematics of gas
exchange (25). Even today, his elegant treatments are a pleasure to behold!
Unfortunately, their beauty was probably not fully appreciated at the time by
significance has been lost over time. Looking over some of these papers, I
realized I could harness the power of Evans’s mathematics to study the second
gas effect, his derivation of the log normal distribution of V Q ratios could be
calculations. I use his equations for the log normal distribution repeatedly but
freely admit that I cannot follow the steps taken in their derivation. Neither can
18
The team achieved the equivalent effect by assuming that there is no net transfer of gases other than
the one under immediate consideration. “If we imagine a lung in which the test gas is accompanied by
only one other gas, the error in this assumption can be made arbitrarily small by choosing the
solubility of the vehicle gas to be sufficiently low” (10).
36 Chapter 2: Background
any of the mathematicians I have consulted on this question. This missing
of Applied Physiology. Evans passed away in 2016, before I was able to discuss
the matter with him and the derivation is probably now lost forever. The steps
that can be determined are shown in Appendix A, at the end of this thesis. They
has seen fit to investigate the matter. With Leon Farhi passing in 2003 and Albert
his nineties and his co-worker Peter Wagner involved in developing a vineyard.
With the main players out of the picture, it has been largely left up to the
San Diego group from the period in the 1970s when the Multiple Inert Gas
This is the final task I set myself. What use is it to predict numerous
one doesn’t actually believe the phenomenon itself is possible? Having always
Chapter 2: Background 37
efficient gas exchange. If I were to succeed in confirming Peyton’s findings and
extending them, then surely the next step would have to be the search for a
successful.
38 Chapter 2: Background
: The Log-Gaussian Distribution
INTRODUCTION
In this chapter we deal with equations derived by Colburn, Evans and West
(10) which give rise to the density functions for ventilation and blood flow as a
function of log V Q( ) and are central to the mathematical treatment of the
concentration and second gas effects developed in this thesis. This paper seems
to be the only one published by the San Diego research team involving Colburn.
Most likely the key figure in deriving the expressions was John Evans, the
the mathematical papers produced by the team and was clearly instrumental in
providing a sound theoretical basis for the Multiple Inert Gas Elimination
The following equations were derived for the expired alveolar ventilation
VA , and the pulmonary end-capillary blood flow Q (see Appendix A):
V
V = 0 e − x /2
2
(3.1)
2π
Q − x −σ 2 /2
Q = 0 e ( ) (3.2)
2π
Here V0 and Q 0 are used as alternative symbols for the expired alveolar
ventilation VA and the total pulmonary end-capillary blood flow Q t , both
expressed in units of l/min. These symbols are useful when moving from a
West and Wagner (92) give several reasons why a log normal distribution
is used most often, as in Eqs. 3.1 and 3.2. Such distributions are common in
If we were to plot V and Q in Eqs. 3.1 and 3.2 against x, the graph for Q
would lie to the right of the graph for V which does not conform to the usual
way these graphs appear in the literature. However, Colburn et al. point out that
THE MEANING OF σ
blood flow. It is derived from the standard deviation of the log ventilation per
unit lung volume, σV and the standard deviation of the log blood flow per unit
σ σV − σ Q
= (3.3)
and has the property of locating the position of the means for ventilation and
matching is seen in disease states and is also known to occur soon after induction
of anaesthesia, with typical values for σ between 0.75 and 1.75 in healthy adults
(16, 52-54, 78). The changes during anaesthesia are especially relevant given
that the persistent increases of the second gas effect in blood during nitrous oxide
mismatch (68-70).
Although Eqs. 3.1 and 3.2 allow ventilation and perfusion to each be
as continuous log normal distributions), this route does not appear to have been
under West, or the group in Buffalo under Hermann Rahn. There are several
reasons why the compartmental approach seems to have been preferred when
investigating the effects of V Q mismatch. First and foremost is the fact that
their primary interest lay in describing the exchange of oxygen, carbon dioxide
exchange are most easily dealt with by treating gas exchange in the lung as
ventilation or blood flow. West and Wagner give a second reason: “the division
compartmental nature of the real lung” (92). In other words, the use of a
per unit volume. This is assumed to follow a log normal distribution with mean
value μV and SD σV and has been drawn using an expression of the form:
1 ( x − μ )2
V
=yV exp − (3.4)
2πσV2 2σV
2
per unit volume. This is assumed to follow a log normal distribution with mean
x−μ
( )
2
1 Q
yQ = exp − (3.5)
2πσQ2 2σQ2
up which effectively covers the whole range of values of the ventilation per unit
volume and blood flow per unit volume. By matching compartments going from
left to right across Figure 3.1(a) and Figure 3.1(b), it is possible to produce the
Figure 3.1(c). Essentially, this involves plotting the average value of yV and yQ
for each compartment as a function of the simultaneous value of log yV yQ . 20 ( )
The ease of applying this technique is probably a third reason it was favoured.
In applying Eqs. 3.1 and 3.2, West makes no attempt to derive explicit
( )
expressions for V and Q in terms of log V Q (91). Colburn et al. also fail to
progress this possibility further (10). This will now be investigated in more
detail.
19
Interestingly, the first negative sign inside the curly brackets in Eq. 3.4 was omitted in the original
paper by West (91). The mistake was repeated in a chapter by West and Wagner in a reference text
(92). The same error features in a scene in the film Gifted, about an 8-year-old girl set a mathematical
problem involving the incorrect expression.
20
This involves an application of the trapezoidal rule (4,36).
21
The equations presented here and in the next section are modified versions of those appearing in
reference (45). Those in 3.8 have been corrected. In the reference, v and q are each shown as
functions of time whereas they are in fact dimensionless.
( Q
log V= ) ( )
log V0 Q 0 + σ 2 2 − σ x (3.6)
Solving for x:
σ
x=
2
( )
+ (1 σ ) log V0 Q 0 − (1 σ ) log V Q ( ) (3.7)
We now introduce two new variables v and q which are the fractional
z = log ( v q ) Eq. 3.7 can then be written in the more compact form:
σ z
x= − (3.8)
2 σ
Substituting for x in Eqs. 3.1 and 3.2 we obtain (see also footnote 23):
2
σ2
V − z −
2
2σ 2
V = 0 e (3.9)
2π
2
σ2
Q − z +
2
2σ 2
Q = 0 e (3.10)
2π
( ) ( )
since z log V Q − log V0 Q 0 , a plot of V or Q as a function of z
Note that=
(
log V0 Q 0 ) units to the left if V0 > Q 0 and to the right if V0 < Q 0 . Plotting
against z has the advantage of always locating the graphs for σ = 0 at zero. When
σ > 0, the mean for V is situated exactly σ 2 2 units to the right of 0 and that
can now be drawn and are shown in Figure 3.2 for values of σ = 0, 0.5, 1 and 2.
for ventilation represents a flow of 4 l/min, the red line for blood flow represents
that for ventilation moving to the right and that for blood flow moving to the left.
Three changes are taking place simultaneously. First, the graphs for
ventilation and blood flow are moving away from each other in proportion to the
increasing degrees of V! Q! mismatch (clockwise from top left σ = 0, 0.5, 1 and 2). Scales
on the z and y axes have been kept constant to allow the effect of changes in σ to be seen
easily.
located at the maximum value of ventilation (blue) and blood flow (red), in those
cases where σ is greater than 0. Secondly, the spread or dispersion of the graphs
changes make it difficult to detect and interpret alterations in other variables such
x = (1 σ ) log ( v q ) on the x-axis. This unit is suggested by Eq. 3.7. When this
2
V0 − x − 2
σ
2
V= e (3.11)
2π
2
Q 0 − x + 2
σ
2
Q= e (3.12)
2π
results in the graphs for ventilation and perfusion being the same shape and size
for every value of σ . All that changes as σ is increased is the location of these
graphs relative to the y-axis and each other. The maximum for v is located
exactly σ 2 units to the right of the y-axis while that for q is located exactly σ 2
22
In this thesis, we generally set these at 4 l/min and 5 l/min respectively.
23
The scaling occurs because of the change of variable, so when integrating Eqs. 3.1 and 3.2 using the
substitution shown in Eq 3.8, dx = − ( d z σ ) .
= 0. 0.5, 1, 2. The distributions are now symmetrical about the vertical line at x = 0. The
curve for ventilation is blue; that for blood flow is red. All curves now are identical in
shape and size and have a standard deviation of 1 unit. The means for ventilation and
blood flow are located σ 2 units to the right and left of the y-axis (blue and red dash-
dotted lines respectively). When σ = 0 , the curves overlap. The area under each curve is
equal to 1.
0.4. This is also the mean value for the distribution. 24 The results are shown in
Figure 3.3. The process of gas exchange can be visualized further by rotating the
graph for blood flow by 180° around the x-axis. This is shown in Figure 3.4 for
σ = 0, 0.5, 1 and 2.
Consider now the case when σ = 2 in Figure 3.4. To the left of the vertical
24
It can be seen that σ has the property of locating the position of the mean as stated in 3.2 above,
rather than being a SD. In fact, we use the symbol μ for σ 2 in parts of this thesis.
Figure 3.4. Distributions of fractional ventilation (v) and blood flow (-q).
Fractional ventilation (v) and fractional blood flow (q) distributions as a function of
(clockwise from top left). The curve for ventilation is blue; that for blood flow is red. The
curve for blood flow has been reflected in the x-axis to show the effect of mismatch more
clearly. The area between each curve and the x-axis is equal to 1.
blood flow ratios but conforms with the classical view that V Q mismatch must
When ventilation and perfusion are perfectly matched, i.e. σ = 0, the red
curve for blood flow in Figure 3.4 is the mirror image of the blue curve for
which ventilation and blood flow are located on a single vertical line, as to
treated by finding lim (1 σ ) log ( v q ) , which from Figure 3.4 clearly results in
σ →0
the red and blue vertical lines approaching each other more and more closely
not exactly equal to zero, say σ = 0.001. Clinically, we would not be able to
distinguish the situation from the case when σ actually equals 0 but
the values of log ( v q ) are increased a thousand fold to produce curves of the
same size and shape as those in Figure 3.3 and Figure 3.4. At each x-value, v =
log ( v q ) =
q, so z = ( ) (
log V Q − log V0 Q 0 = )
0 giving the required condition
that V Q = V0 Q 0 .
low V Q gas-exchanging units to the left. If we wish to show actual gas uptakes
VA and the fractional blood flow q, by Q t . In this thesis, we will often be using
which to study the uptake of anaesthetic gases. Any questionable results can be
( )
investigated further by reverting to log ( v q ) or log V Q as the independent
variable.
INTERCHANGEABILITY OF X-VARIABLES
flow V and Q , and the derived variables v and q. Traditionally, these have been
plotted as functions of log (V Q ) or log10 (V Q ) . In later parts of this thesis we
are also interested in the distribution of gas uptake i.e. where gas uptake occurs
mirrors a similar trend in the distribution when plotted against log (V Q ) . In
other words, are we likely to miss some trend by using the new variable on the
, it follows that a trend in D(y) does indeed reflect a similar trend in D(x).
LEBESGUE-STIELTJES INTEGRATION
his colleagues when equilibrating alveolar gas with pulmonary capillary blood
of the lung. The principle is illustrated in Figure 3.5. Equilibration between gas
output for each phase calculated by adding together the appropriately flow-
courses, in which the area under a curve is taken as the integral of the function
involved in generating the curve and is approximated more and more closely by
of the x-axis is made finer and finer (4). It should therefore come as no surprise
Colburn et al. (10) although there is no evidence that it was put to use by them
many mathematical texts e.g. McShane (60). While the theory is somewhat
(
Gi = f Vi , Q i ) (3.13)
n
G = ∑ f (Vi , Q i ) (3.14)
i =1
which Vi and Q i are replaced by expressions given in Eqs. 3.1 and 3.2 or variants
∞
V − x − σ
2 2
Q 0 − x + 2
n σ
( ) ∫
2 2
=G lim
= ∑
f Vi , Qi f 0 e 2
, e dx
(3.15)
−∞ 2π 2π
n →∞
i =1
problem.
SUMMARY
for ventilation and blood flow as functions of the logarithm of a suitable variable.
( )
log V Q , the logarithm of the V Q ratio. We have proposed an alternative
and fractional blood flow respectively. The use of this variable provides a stable
platform for the graphical interpretation of gas exchange in later chapters of this
thesis. The suitability of this variable lies in the fact that the SD of the curves for
ventilation and blood flow is always 1 regardless of the value chosen for σ .
In the 1970s, West and Wagner described the standard method of treating
to dispense entirely with the compartments and treat the distribution as a density
about the behaviour of the distribution” (92). This possibility was not
investigated further by them. In this chapter, we have laid the groundwork for
In this chapter we derive the equations used to study the exchange of the
first and second gases in any gas-exchanging unit of the lung, assuming that
studying the second gas effect. This involves excluding any factor which does
therefore does not play an essential role in its production. Thus, we exclude the
physiological gases oxygen and carbon dioxide from consideration. Clearly, they
may themselves be the subject of a second gas effect and can also make a small
contribution to the net gas volume change during nitrous oxide anaesthesia.
important results that would otherwise be obvious. This simplification has been
used by many previous workers in the field (18, 21, 42, 57, 58). Absorption
atelectasis (12), which was included in the modelling of Peyton et al. (68-70),
has also been excluded as it is not an essential ingredient in the production of the
second gas effect. The equations being sought are those describing the fractional
concentration, partial pressure and uptake of the gases involved. These are
further.
mismatch are respectively VIA and VEA , 26 then a mass balance equation of the
and
= ( , F v Pv
Fa Pa PB − PH 2O= ) (P
B )
− PH 2O , 27 where Pa , P v , PB , P H 2O are
the arterial partial pressure and mixed venous partial pressure of the gas,
centigrade respectively. Equations of this type were solved first with VIA = VEA
by Kety (42) and Mapleson (57), and then with differing inspired and expired
using the chain rule, this gives rise to the terms VL FA + FAVL . Although the
this was not the case when the equations were first solved so the term FAVL was
25 Concentrations are defined by convention as fractional concentrations in dry gas and may be converted to
partial pressures through multiplication by (P − P ) .
B H 2O
is allowed to vary with constant inflow and also results from the varying inputs
Each gas present will have its own mass balance equation. Those for
oxygen and carbon dioxide are more complicated because there is no single
also a mass balance equation for the net gas volume change. All of these
from consideration. This approach is not new and has been adopted by previous
workers in the field (10, 27, 48, 68-70). Several benefits follow immediately:
2. The composition of gas in the FRC remains constant from the end of one
breath to the end of the next so that FRC gas can be completely ignored and
the inspired alveolar tidal volume equilibrated directly with the pulmonary
blood flow in much the same fashion as was done by Riley and Cournand
and may also be regarded as applying over very small periods of time during the
initial phase of rapid rise. Both phases are shown in Figure 4.1.
Furthermore, Peyton has shown that a “square root of time” rule applies to
the change in uptake of all gases and volatile agents (66). This is consistent with
Severinghaus’s formula (83) and led Peyton to conclude that “after 5 minutes,
the rate of change of uptake of any gas will be only 10% of its value at that time
and that this declines to 5% after 10 minutes and 1% after 50 minutes”. This low
rate of change provides further support for the use of the steady-state approach.
pressures of the first gas, (generally nitrous oxide in most simulations in this
thesis), it is possible to mimic the net gas volume changes occurring at any
specified stage of anaesthetic induction. Since the concentration and second gas
effects are generated by the net change in gas volume and do not depend on
whether we are dealing with steady-state gas exchange or not, and given the
especially attractive and will be used almost exclusively throughout this thesis.
1. A first gas (FG). Alveolar-capillary uptake of this gas generates the volume
changes responsible for the SGE. It must be soluble in blood and should
SGE. We use the subscript SG/FG to indicate that the SG is used together
3. The remaining gas. This acts purely as a vehicle and for the purposes of the
28
As mentioned in a previous footnote, the San Diego team usually achieved the same outcome by
specifying no net exchange of the remaining gases (10, 27). In our 3-gas system only the first gas is
permitted to contribute significantly to gas volume contraction.
(1 − FI )VI =
(1− F ) VA (4.3)
where
F = concentration of the first gas in the dry fraction of the outflow which is also
To this system we now add a second soluble gas, denoted by the subscript
“SG” to allow it to be distinguished from the first gas. (In the absence of a “FG”
or “SG” subscript, the variable refers to the first gas.) The second gas also obeys
Based on the conservation of mass, Eqs. 4.2 and 4.4 state that the amounts
of the first and second gases leaving the gas phase must equal the amounts of
these gases gained by the blood phase. The remaining gas is treated as if it were
and second gases. Equation 4.3 expresses the conservation of the remaining gas.
The second gas is assumed to be present in such a low concentration that its
contribution to the net volume change may be ignored. Thus, it does not appear
recirculation. This maximizes the uptake of the second gas making the detection
traces of second gas at intermittent intervals. From Eq. 4.4 we then have:
VI FISG
FSG = (4.5)
VA + Qλ SG
This case is obtained by solving Eq. 4.3 for VI and eliminating this
variable from all equations.
1− F
VI = VA (4.6)
1 − FI
Substituting for VI in Eq. 4.2 we may solve this equation for F .
VI FI + λQ (1 − FI ) F v
F= (4.7)
VA + λQ (1 − FI )
We denote the uptake of the first gas N2O by the symbol VN2O , the uptake
of the second gas in the absence of the first gas by VSG and the uptake of the
second gas in the presence of first gas as VSG / FG . VN2O is given by the left-hand
side of Eq. 4.2; VSG is given by the left-hand side of Eq. 4.4 when VI = VA ; and
λVA( FI − F v )
VNhlO = (4.8)
2
λ(1 − FI ) + (VA Q )
Note that Eq. 4.8 was derived by Colburn, Evans and West (10). Except
term, Eq. 4.10 is similar to Eq. 4.8. Because FI SG 1 , it has been omitted from
“nhl”, composed of n functional units. Let Vj and Q j be the alveolar ventilation
SG in blood phase
By analogy with Eq. 4.10, the net uptake of the second gas in the presence
29
As was pointed out in 3.1 we use the terms V0 & VA and Q 0 & Q t interchangeably, where VA
refers here to the total expired alveolar ventilation per minute.
By analogy with Eq. 4.9, the net uptake of the second gas in the absence
n V λ FI
VSG
nhl
= ∑ j SG SG (4.12)
j =1 λSG + V j Q j
SUMMARY
variables include the mixed alveolar concentrations and partial pressures of the
first and second gases, the rate of uptake by pulmonary capillary blood of these
gases and the partial pressures of the gases in arterial blood. The derivation has
mean the levels associated with plateauing of the washin curve for nitrous oxide
shown in Figure 4.1. The value of 100 ml/min has been given by Severinghaus
(83).
The existence of this phenomenon was first predicted by Peyton et al. (68),
oxygen and sevoflurane (34, 67, 71, 72). The explanation given was that the
(82) who believed that the SGE cannot be divorced from significant nitrous
oxide uptake (see 2.14). The idea that a phenomenon traditionally believed to
be caused by the large reductions in gas volume associated with high N2O
uptake, can persist at clinically significant levels even when those volume
Since 2006, there has been no real progress in resolving this dispute. The
current chapter attempts to provide useful answers. We set out to further explore
The material in this chapter is used with the permission of ANESTHESIOLOGY and Wolters Kluwer
30
The sequence adopted in this chapter is to define the Washin Ratio (WR),
which we use to simulate the degree of nitrous oxide washin. We need this
nitrous oxide washin. We next define our measure of the SGE, the Augmentation
Ratio (AR). Equations for the augmentation ratio are then derived using results
anaesthesia.
where:
VN2O
WR = 1 − (5.1)
VI FI N2O
VN2O = VI FI N2O , its maximum possible value, WR = 0, a situation that can only
Stage of
Washin Ratio N2O Uptake FA FI
Anaesthesia
End of washin 1 0 1
of VN2O for which 0.9 ≤ WR ≤ 1.0 , i.e. when washin is 90 to 100% complete.
Applying Severinghaus’s equation for the rate of uptake of nitrous oxide (83),
this should occur within 10 to 15 minutes from the start of induction, at which
time the rate of change in the uptake of nitrous oxide ( VN2O ), has declined to less
than 5% of its current rate of uptake. The washin ratio is roughly equivalent to
second gas in the presence of N2O, over that which would have existed in the
absence of N2O, with all other variables kept constant. With no SGE, the
augmentation ratio is 1. Once the SGE becomes evident, the augmentation ratio
exceeds 1. The term “augmentation ratio” was first used by Epstein, Rackow,
Salanitre and Wolf (24). Although they only calculated augmentation ratios in
end-tidal gas, the ratio may also be calculated for blood. The ratio is defined as
follows:
oxide, and hence it cannot be used to measure the concentration effect. Other
disadvantages are discussed elsewhere in this thesis (see 7.17). In this chapter
we obtain equations for the augmentation ratio as defined in Eq. 5.2 and use them
to investigate the SGE as nitrous oxide washin plateaus and uptake reaches
and then extend the derivation to a non-homogeneous lung with a log normal
distribution of V Q ratios.
to find that:
AR ( blood ) = (1 − F ) (1 − FI )
VSGhl/ FG / VSG hl = (5.3)
AR in gas phase
Substituting for VI from Eq. 4.3 in Eq. 4.2, we obtain an expression for FSG
hl
/ FG :
(VA / Q ) FISG 1− F
FSGhl / FG = (5.4)
(VA / Q) + λSG 1 − FI
When no first gas is present, VI = VA , and hence the expression for FSG
hl
can be
AR ( gas ) =
F hl / F hl
SG / FG SG
=(1 − F ) (1 − FI ) (5.5)
augmentation ratios in the gas phase and in the blood phase are both equal to
(VA / Q ) + λ(1 − F v )
AR(gas) = AR(blood) = (5.6)
(VA / Q ) + λ(1 − FI )
n Vj 1 − Fj
∑
j =1 λSG + V j Q j 1 − FI
AR ( blood ) = (5.7)
n Vj
∑
j =1 λSG + V j Q j
Applying Eq. 4.7 to the jth compartment and simplifying, we obtain the following
n Vj V0
∑
λ ( FI − F v ) j =1 λ (1 − FI ) + Vj Q j − 1
AR ( blood ) =
1+ (5.8)
λSG − λ (1 − FI ) Vj V0
n
∑
λ + V Q
j =1 SG j j
Note here that with 70% N2O ( λ = 0.47 ) and FI = 0.7 , the expression
λSG − λ (1 − FI ) is positive for all λSG > 0.141 which includes all the volatile
Assuming now that these series are both applied using a log normal
λ ( FI − F v ) I
AR ( blood ) =
1+ − 1 (5.9)
λSG − λ (1 − FI ) I SG
where
1
∞
dx n Vj V0
I ∫ λ (1 − FI ) e ≡ ∑
( ) j =1 λ (1 − FI ) + V j Q j
2
2π −∞
x 2 /2
+ V0 Q 0 e( x −σ ) /2
(5.10)
1
∞
dx n Vj V0
I SG ∫−∞ λ e x2 /2 + V Q e( x−σ )2 /2 ≡ ∑ (5.11)
2π SG ( 0 0 )
j =1 λSG + V j Q j
of N2O equilibration). However, Peyton and co-workers actually use the N2O
uptake, VN2O , as an input variable (68-70). This enables us to eliminate the term
the uptake in each compartment to obtain the total uptake, we obtain the
following equation:
n λVj ( FI − F v )
VNnhlO = ∑ (5.12)
2
j =1 (
λ (1 − FI ) + V Qj j )
31
Note that we have used the Eqs. 3.1 and 3.2 here rather Eqs. 3.11 and 3.12 in forming the integrals
but as mentioned previously in 3.12, either set of equations will do.
VNnhl
O
FI − F v =2 (5.13)
λV I 0
The final expression relating the SGE to the uptake of N2O and the degree of
1 VNnhl 1 1
AR ( blood ) =
1 + 2O
− (5.14)
λSG + λ (1 − FI ) V 0 I SG I
We use WR the previously defined washin ratio as a variable on the x-axis. Note
that given the assumption that only the first gas, N2O contributes to changes in
or not N2O is present, this is not the case in the gas phase. Here the second gas
is contained in a smaller volume whose size depends on the rate of N2O uptake.
The equations describing this phase are therefore more complicated. Only the
1 VNnhl λSG I SG − λ (1 − FI ) I
AR ( gas ) = 1 +
2O
λSG + λ (1 − FI ) I V0 1 − λSG I SG
(5.15)
Note that Eqs. 5.14 and 5.15 may both be simplified to the following
approximate form:
1 + ψ (1 − WR )
AR = (5.16)
effect of σ . Perhaps the most useful form of Eq. 5.15 involves the substitution
of (VI − VA ) for VN2O . This gives rise to the following equation:
1 + k (VI − VA )
AR = (5.17)
When AR = 1, there is no second gas effect so when the SGE occurs, we can
write SGE = k (VI − VA ) . 32 This indicates how the second gas effect can be both
a function of the nitrous oxide uptake (VI − VA ) and the degree of ventilation-
perfusion mismatch (which appears in the term k along with the solubility of the
first and second gases and the inspired concentration of the first gas).
Figure 5.1 shows the results for the blood and gas phases for desflurane
( λ = 0.42), isoflurane ( λ = 1.4), and diethyl ether ( λ = 12.1). The arrow in each
diagram indicates the direction in which σ is increasing. The x-axis shows the
proportion of nitrous oxide washin, and the y-axis shows the augmentation ratio.
The most consistent feature of these diagrams is the divergence of the lines from
the point (1,1), the point on the x-axis representing complete washin of the
inspired alveolar ventilation, so that according to Eq. 5.17, the SGE disappears
completely.
32
In the ANESTHESIOLOGY article, the term VA was included in the denominator. Since we are dealing
with the constant outflow case in this simulation, the equations for AR and SGE have been simplified
further by incorporating the constant expired ventilation VA into the variable k.
σ = 0 , which represents a homogeneous lung, is the same for both gas phase and
blood phase, as expected for a “single compartment” lung with full equilibration
between gas and blood phases. In addition, it is the same for all three gases, i.e.
to be opposite for the blood and gas phases. In blood, an increase in σ increases
the augmentation ratio, while in the gas phase, the augmentation ratio decreases.
Moreover, in blood, this amplification is greatest for the least soluble gas,
desflurane, and least for the most soluble gas, diethyl ether, while in the gas
phase, the opposite is true. Results obtainable for sevoflurane ( λ = 0.67) are
Equation 5.17 provides an explanation of how the SGE can continue, even
after nitrous oxide uptake decreases to levels seen in the “maintenance phase”.
k (VI − VA ) is responsible for the SGE. Here the difference (VI − VA ) is produced
“shrinkage” need not be the only factor involved in producing the SGE. We have
shown that k is affected by the degree of V Q mismatch and that the effect of
than 1. In the gas phase, as σ increases, k decreases and is less than 1. The term
have shown that it is possible for V Q mismatch to influence the SGE through
its contribution to k, this still does not explain exactly how this occurs. This
is the investigation of the effect of varying V Q ratios across the lung. These
et al. (71), that the magnitude of the SGE on arterial sevoflurane partial pressures
minutes afterward. However, these authors did not attribute the finding to a
concomitant diminution in SGE in the gas phase. Our model provides a more
findings of Taheri and Eger (86), and those of Hendrickx et al. (34) for end-
It is evident from Figure 5.1 that the SGE in blood and its augmentation
by increasing V Q scatter is greater for the least soluble second gas, desflurane.
nitrous oxide to the inspired gas mixture may maintain the efficiency of the lung
While the SGE in the gas phase is most easily detected using end-tidal gas
theatres today, augmentation in the blood phase is not normally measured but is
This represents a new factor that should be taken into consideration in addition
to other relevant factors (3, 23, 37, 61), when interpreting a MAC reading during
anaesthesia.
33
This is defined as the concentration of the anaesthetic agent in 100% O2, that prevents movement in
response to skin incision in 50% of unpremedicated subjects at sea level (1 atm). Prior to the invention
of the BIS and Entropy depth-of-anaesthesia monitors, this was the commonly used measure of depth
of anaesthesia. The more potent the anaesthetic agent, the lower the MAC value. This measure of
anaesthetic depth was introduced into clinical anaesthetic practice by Merkel & Eger (60).
simulate the second gas effect as nitrous oxide uptake falls to the low levels seen
which are based on end-tidal gas analysis, may underestimate the depth of
anaesthesia when volatile agents are used together with nitrous oxide.
In the last chapter we verified Peyton’s claim that the second gas effect
can persist into the maintenance phase of nitrous oxide anaesthesia and
past, it seems appropriate to revisit some of the previous results which might be
During the 1970s, John West’s group in San Diego was engaged in a race
the Multiple Inert Gas Elimination Technique or MIGET (88), a test which is
available today in various specialized centres around the world. West recruited
problems which confronted the team and possessed the mathematical skills to
chapters.
diagrams in the case of a single soluble inert gas present in the inspired gas
mixture together with other gases not undergoing net transfer 34. Their study
focused on the elimination from the lung of a gas supplied at constant partial
pressure in mixed venous blood. Colburn et al. (10) explored further the
ventilation and blood flow must reduce the pulmonary uptake of any gas with a
linear dissociation curve i.e. any gas that obeys Henry’s Law. This conclusion
their complex interaction with haemoglobin, neither oxygen nor carbon dioxide
obeys Henry’s Law. Evans et al. did investigate situations in which oxygen and
carbon dioxide might not obey this rule, but these were extremely unlikely to be
met under normal conditions. Using a computer model which generated a wide
variety of V Q distributions, Neufeld et al. also concluded that inert gas a-A
principle easily understood from Figure 3.4. However, these authors did not
consider possible effects associated with net gas volume changes during gas
exchange. In this chapter we use N2O uptake as a tool to investigate the effect of
34
As explained on page 36, this is equivalent to assuming that the soluble gas is accompanied by
an insoluble filler gas.
EQUATIONS
delivered in the inspired gas mixture at a rate VI FI and removed at a rate VA FA
VI FI − VA FA . This must equal the uptake by blood which is given by
Pv (P B )
− PH 2O respectively, so that we have the following equation (analogous
to Eq. 4.2):
have:
these equations give rise to the following general equation for the alveolar and
P=
a P=
A
( VA Q ) PI + λ (1 − FI ) Pv (6.3)
( VA Q ) + λ (1 − FI )
( VA Q ) + λ (1 − F v ) VA Q
(
VI Q = ) VA Q + λ (1 − FI ) ( ) (6.4)
( )
N
VI = ∑ VI j (6.5)
j =1
We have previously shown (see 3.10) that with a log normal distribution
λ ( FI − F v )
VA
{e }
+∞
=VI ∫
− x 2 /2
+ dx (6.6)
( )
2
VA Q e + λ (1 − FI ) e
2π −∞ − ( x − σ ) /2 x 2
/2
The left-hand term in the integrand produces the log normal distribution of
nitrous oxide. Once VA , Q , σ , and F v are specified, Eq. 6.6 provides a one-to-
one relationship between VI and FI , and hence allows the calculation of the
inspired concentration of nitrous oxide associated with the desired total inspired
V
alveolar ventilation. Letting Δ= (VI − VA ) we may now determine the required
35
Eq. 6.6 may take several alternative forms if the variable x is substituted as discussed in 3.8.
These are equivalent to each other. As mentioned previously in 3.12, we choose the most
suitable form for use in each particular situation.
with σ , the degree of mismatch between ventilation and blood flow. We use the
terms “volume shrinkage”, “volume contraction” and “net gas volume uptake”
equal to 0. Alveolar ventilation VA is set at 4 l/min and pulmonary blood flow
demonstrate differences when these occur, while also lying within the range
∑VA PA
j =1
j j
PA = N
(6.7)
∑VA
j =1
j
P AFG
=
VA Q t (
I1
) (6.8)
PI FG 2π
∞
dx
where I1 = ∫ ( ) + x
(6.9)
λ (1 − FI ) e (VA Qt ) e
x 2 + 2 σx − σ 2 /2 2
/2
−∞
∑ Q PA
j =1
j j
Pa = N
(6.10)
∑ Q
j =1
j
P aSG
=
(
VA Q t
I3
) (6.11)
PI SG 2π
∞
dx
where I3 = ∫ λ (1 − FI ) e (6.12)
( )
2
+ VA Q t e( x −σ )
2 /2
x /2
−∞
distinguish the second gas from the first gas, nitrous oxide, whose abbreviations
From Eq. 6.3, and setting P v = 0 for the second gas we have:
FjSG Vj Q j
= (VI VA )j (6.13)
FI SG λSG + Vj Q j
where (VI VA ) is the ratio of the inspired ventilation to the expired ventilation
j
in the jth compartment. This ratio may be obtained from Eq. 6.4 and results in the
following equation:
=
FjSG ( λ − λSG ) (Vj Q j) −
( λ − λ' ) (Vj Q j) (6.15)
FI SG ( λ'− λSG ) ( λSG + Vj Q j) ( λ'− λSG ) ( λ' + Vj Q j)
integrals we obtain:
P ASG (VA Q t )
= {( λ − λ ) I − ( λ − λ') I } (6.16)
( λ' − λSG )
SG 2 1
PI SG 2π
∞
dx
where I2 = ∫ (x ) /2 + V x
(6.17)
( A Qt ) e
2 2
+ 2 σx − σ 2
−∞ λSGe /2
P aSG (
VA Q t )
and =− {( λ − λSG ) I 4 − ( λ − λ') I3 } (6.18)
PI SG ( λ' − λSG ) 2π
∞
dx
where I4 = ∫λ (6.19)
( )
2
e x /2 + VA Q t e( x −σ ) /2
2
−∞ SG
PASG P aSG
− 0
= (6.20)
PI SG PI SG
and the result converted to a value for FI N2O by inverting the relationship
λ'
FI N2O = 1 − (6.21)
λ
Solving Eqs. 6.16 and 6.18, graphs may now be drawn for P A PI and
described by West, Wagner and Derks (89). This is achieved in our model by
setting the inspired nitrous oxide concentration to zero, so that with the very low
change in net gas volume. The resulting curves are shown in Figure 6.1A.
pressures are equal. This is indicated by the black stippled curve. Moving along
this curve towards the left end of the x-axis, we are dealing with less soluble
gases – very little is transferred to blood so that the alveolar (and hence arterial)
partial pressure approaches the partial pressure in the inspired gas mixture and
( ) with no net volume change i.e. ΔV = 0 and σ = 1.5 The stippled black curve
lower curve ΔV = 0 ( ) and upper curve ΔV = 600 ml/min ( ); C: the
curves in A and B are summed for gas ( ) and blood ( ). In each case, the
stippled green vertical line on the left indicates the position of sulphur hexafluoride, the
stippled green vertical line on the right indicates the position of methoxyflurane.
axis, we are dealing with more soluble gases – more and more of the alveolar
gas is transferred to blood so that the alveolar (and hence the arterial) partial
less gas is removed from the alveolar gas mixture producing a higher partial
pressure in the gas phase and a lower partial pressure in blood. Thus, the curves
in Figure 6.1A for alveolar gas and arterial blood separate, producing an
V Q mismatch i.e. with σ = 0. The black stippled curve is the same as in Figure
6.1A and represents the situation where there is no net gas volume change
( ΔV = 0 ) ; the solid black line represents the situation with a net gas volume loss
of 600 ml/min ( ΔV = 600 ) . The difference between the two curves reflects the
magnitude of the second gas effect. The effect is greater with gases of low
solubility, a fact noted by Stoelting and Eger (84). Until the work of Peyton et
If the effects were merely additive, they would produce the result shown in
Figure 6.1C.
increasing ΔV from 0 ml/min to 600 ml/min on the alveolar and arterial partial
mismatch ( σ = 1.5).
ΔV = 600 ml/min. Now it is the increase in P A PI of the continuous blue curve
relative to the continuous black curve that reflects the impaired uptake in the
With ΔV = 600 , the graph of P a PI with σ = 1.5 (continuous red curve)
greatly exceeds that for σ = 0 (continuous black curve) for much of its course,
especially at low solubilities ( λ < 0.05 ) . This reflects the increased uptake which
occurs with significant gas volume shrinkage. The effect of the contraction in
comparing the stippled curve with the continuous curve for each colour. The
greater magnitude of the SGE in blood than in the gas phase is indicated by the
larger area between the two red curves in Figure 6.2B when compared with the
(AR) which was defined in 5.2 as the partial pressure of the second gas in the
relevant phase (gas or blood) in the presence of N2O divided by the partial
pressure the second gas would have in that phase in the absence of N2O, all other
factors being held constant (24, 45). Augmentation ratios may be obtained for
the gas phase and for blood by dividing P A PI and P a PI at any given value
of ΔV with their equivalent values when ΔV = 0 . In each case, all other
variables must be held constant. AR(gas) and AR(blood) may then be graphed
as functions of λ .
gas volume shrinkage may be derived in Figure 6.2 by dividing the value of
P A PI or P a PI when ΔV = 600 by its value when ΔV = 0 for the value of
λ associated with that gas. In Figure 6.3A we have shown the augmentation ratio
for gas (blue) and blood (red) for ΔV = 200 ml/min (stippled curve) and 600
ml/min (continuous curve) with σ = 1.5; augmentation, (and hence the SGE) is
greater at the higher value of ΔV , an effect most noticeable as the solubility of
Figure 6.3B shows the augmentation ratio for gas and blood for a range of
σ values with ΔV set at 100 ml/min. In the absence of V Q mismatch, the
mismatch has opposite effects on the gas phase (blue) and blood (red). As σ
increases, the SGE increases in blood, but decreases in the gas phase (black
arrows). This in line with our previous findings (45) as shown in Figure 5.1.
inspired gas mixture. For example, when 1% sevoflurane is present in the dry
drawn with σ = 0 (stippled black line) and σ = 1.5 (continuous lines, blue = gas,
(clockwise) σ = 1.5, A: ΔV = 0, B: ΔV = 200, C: ΔV = 400, and D: ΔV = 800
ml/min. The stippled black curve in each diagram ( ) is associated with no
mismatch of ventilation and perfusion i.e. σ = 0.
curve in each figure regardless of the value of ΔV . Notice that in each case, the
blue line representing P A PI always lies above or coincides with the black line.
However, while the red line representing P a PI lies below the blue line for
P A PI on the right of the solubility axis, it crosses the blue line as one moves
left along the solubility axis. This represents the situation where the partial
gas. The point of intersection is affected by the shrinkage factor and moves to
of V Q mismatch) for ΔV = 0, 200, 400, 600, 800 and 1000 ml/min. In each case, the
direction of increase in ΔV is shown by the red arrow. Note the change in the range of
values shown on the y-axis for sulphur hexafluoride.
The red arrow in each diagram shows the direction of the increase in ΔV
In the case of sevoflurane and ethylene, the A-a partial pressure gradient is
reduced and moves towards zero, the value seen when σ = 0 and there is no
pressure difference is less than zero when ΔV is 400 ml/min or greater, with the
higher than that in mixed alveolar gas, when σ = 1.2 and ΔV = 600 ml/min.
36
We use both the Australian spelling “sulphur hexafluoride” and the US spelling “sulfur
hexafluoride” in this thesis.
change. The red arrows indicate the direction of increase of σ from 0 to 4 in increments
of 0.5; B: with a net volume change of 1000 ml/min (BTPD), so that ΔV VA = 0.25. Red
arrows as in A. The blue arrow indicates the point where the uptake ratio of sulphur
volume change has been increased in the direction of the red arrows from 0 ml/min to
1000 ml/min (BTPD), in increments of 200 ml/min. In each diagram, the solid vertical
green line is drawn at log10 ( 0.8 ) ; the position of sulphur hexafluoride is shown by the
stippled green vertical line on the left, that of methoxyflurane by that on the right.
As an index of gas exchange, Colburn, Evans and West studied R(λ), the
ratio of the gas exchange rate of a compartmental model over the exchange rate
of the homogeneous model with the same total expired alveolar ventilation and
blood flow (10). In their model, R(λ) is always less than 1, indicating that gas
homogeneous lung. We use their expression for R(λ) and apply it to an inert gas,
present in the inspired gas mixture in low concentrations in the presence of net
Many of the above results are consistent with gas uptake in the presence
Figure 6.6, we compare the situation which applies when V Q mismatch occurs
in the absence of net gas volume changes with two scenarios in the presence of
such changes.
Figure 6.6A shows a series of such graphs for values of σ ranging from 0
to 4 in increments of 0.5. With very low inspired concentrations of inert gas there
is no net volume change during gas exchange and the family of curves is
increased, R(λ) falls as predicted by Colburn et al. (10) and is never greater than
1.
This is the classical result described by Colburn et al. (10), with which we
now compare our two scenarios. Figure 6.6B shows the effect of increasing the
value of σ from zero with the net volume uptake fixed at 1000 ml/min (BTPD),
the degree of mismatch fixed at σ = 2. In each case, once ΔV is greater than
zero, the curves are no longer symmetric about the vertical line x = log10 (VA Q )
; the minimum for each curve has been shifted to the right; each curve begins
above the R(λ) = 1 line and ends below it. At the left end of each curve in Figure
6.6B, R(λ) increases as σ increases, while at the right end, R(λ) decreases as σ
increases. At the left end of each curve in Figure 6.6C, R(λ) increases as ΔV
exhibit this property (see blue arrow in Figure 6.6B) and only then under extreme
conditions, which are probably not attainable clinically ( ΔV = 1000 ml/min, σ
= 4).
According to West, Wagner and Derks, during gas uptake, the alveolar
component and must therefore always be positive (89). However, this rule
clearly cannot apply at the lower end of the solubility scale in Figure 6.4B-D.
Our findings are shown in Figure 6.5. With VA = VI (i.e. ΔV = 0), the A-a partial
but less so as the solubility of the gas is decreased. As the difference between
VA and VI is increased with the less soluble gases shown in Figure 6.5B-D, the
of the gases, the pressure gradient is reversed, with the partial pressure in blood
predicted to be higher than that in the gas phase, even though the movement of
the sulphur hexafluoride is in the direction from gas phase to blood. One would
normally expect the partial pressure of a gas to be lower at all points downstream
oxygen from inspired air to the mitochondrion - the so-called “oxygen cascade”
(5).
values of λ, σ , and (VI − VA ) , the factors controlling augmentation in Eq. 5.17.
its effect on the term k in Eq. 5.17, it impedes nitrous oxide uptake and therefore
can reduce augmentation through its effect on the “shrinkage” term, (VI − VA ) .
Increasing the inspired nitrous oxide concentration can compensate for this but
These competing effects are shown in Figure 6.7. Each point on the blue
coloured curved surface in Figure 6.7 represents a point at which the alveolar
and arterial partial pressures are equal. This surface is shown as a function of the
gas solubility (x-axis) and the degree of V Q mismatch (y-axis). On the z-axis
is shown the FI N2O corresponding to each point on the surface. The surface
slopes upwards and backwards from the bottom right to the top left. A yellow
horizontal plane is drawn for FI N2O = 0.7, the highest inspired concentration
region between the yellow horizontal plane and the dark-blue curved surface compatible
with the phenomenon of A-a partial pressure reversal. When σ = 0, P A = P a for all
values of FI N O so the graph has been offset slightly from σ = 0 to facilitate discussion.
2
which can be used safely in most healthy subjects without producing hypoxia
(55). In the region beneath the blue surface (red arrow), the inspired nitrous oxide
concentration is too low to produce the required volume change. The region
above the yellow plane on which FI N2O = 0.7 (black arrow) is associated with
an “unsafe” FI O2 . In the region above the dark blue surface and below the yellow
FI N2O = 0.7 plane (green arrow), it is possible for nitrous oxide uptake to
generate a reversal of the alveolar-arterial partial pressure gradient for the second
the highest solubility for which A-a reversal can occur is approximately 0.35 so
that none of the volatile agents in common use can be expected to demonstrate
A-a partial pressure reversal during uptake. However at least 6 other gases may
0.0076). With σ = 1.5, the highest solubility for which A-a reversal can occur is
although A-a partial pressure reversal is theoretically possible, the force driving
So far, we have only considered a second gas with a mixed venous partial
differences in anaesthetized dogs being ventilated with air (8). Here the
oxygen uptake and carbon dioxide output but there is no net exchange of
nitrogen. They showed the mean arterial pressure of nitrogen was greater than
throughout the lung. The situation may be mimicked in our model by setting P v
equal to P a for the inert gas involved as described in the next section.
venous partial pressure of N2 in the mass balance and setting it equal to the
still necessary to determine PI for N2O uptake to produce the desired value of
ΔV . After equilibration in the jth compartment, for the second gas we now have
P ASG N
Pj
So, since = ∑ SG (Vj VA ) , the mean alveolar partial pressure is given
PI SG j =1 PI SG
by:
P ASG N
Vj Q j λ + Vj Q j λSG ( P v PI ) SG
∑ (Vj VA ) λ
+
PI SG j =1 SG + Vj Q j λ (1 − FI ) + Vj Q j λSG + Vj Q j
(6.23)
P aSG N
Pj
and since ( )
= ∑ SG Q j Q t , the mean pulmonary end-capillary blood
PI SG j =1 PI SG
P aSG N
Vj Q j λ + Vj Q j λSG ( P v PI ) SG
∑ ( Q j Q t ) λ
+
PI SG j =1 SG + Vj Q j λ (1 − FI ) + Vj Q j λSG + Vj Q j
(6.24)
Defining I5 as follows:
∞
dx
I5 = ∫ e( x )
(6.25)
( )
2
2 − 4 σx + 2 σ 2 /2
−∞ λSG e(
x − σ ) /2
+ VA Qt
λ′ λ (1 − FI ) , as defined previously:
Lebesgue-Stieltjes integrals in which =
Since P vSG = P aSG , we can eliminate the term containing P v from Eq. 6.27 and
2π
I= I= (6.29)
1 3
(
λ'+ VA Qt )
2π
I= I= I= (6.30)
2 4 5
(
λSG + VA Qt )
So, Eq. 6.28 becomes:
P aSG
=
(
λ + VA
Qt ) (6.31)
PI SG (
λ' + VA
Qt )
arterial pressures are equal and both greater than the inspired partial pressure. If
λ ' = λ , alveolar, arterial and inspired partial pressures are all equal.
The results are shown in Figure 6.8. This graph may be compared with the
graphs in Figure 6.5. The red arrow shows the direction of increase in ΔV . This
is associated with an increase in the size of the a-A partial pressure gradient.
horizontal green line in Figure 6.8. The curve for ΔV = 50 ml/min intersects this
line with σ approximately equal to 1.8. This is close to the upper limit of σ
Figure 6.8. A-a partial pressure difference for nitrogen when R < 1.
Alveolar-arterial partial pressure difference for nitrogen as a function of σ , the degree of
ventilation-perfusion mismatch, for ΔV = 0, 25, 50, 75, 100 ml/min. These values of
ΔV are associated with a Respiratory Exchange ratio less than or equal to 1. The
direction of increase in ΔV is shown by the red arrow.
alveolar gas is shown in Figure 6.9. Here we have set PI to zero for N2O and
determined the value of P v associated with the specified value of ΔV . This
gives rise to the following set of equations for the second gas, nitrogen:
2π ΔV
F vN2O = − (6.33)
λN2O VA I v
(6.34)
(6.35)
∞
dx
H1 = ∫ λe
−∞
( x 2 + 2 σx − σ 2 )/2
(
+ VA Q t e x /2
2
)
(6.36)
∞
dx
H2 = ∫λ
−∞ e ( x 2 + 2 σx − σ 2 )/2
(
+ VA Q t e x /2
2
)
(6.37)
SG
∞
dx
H3 = ∫λ
−∞ e x 2 /2
(
+ VA Q t e( x −σ ) /2
2
)
(6.38)
SG
∞
dx
H4 = ∫ λe
−∞
x 2 /2
( )
+ VA Q t e( x −σ ) /2
2 (6.39)
∞
dx
H6 = ∫λ ( x −σ ) 2
/2
( )
+ VA Q t e( x − 4 σx + 2 σ )/2
2 2 (6.41)
−∞ SG e
Since P vSG = PaSG , we can proceed as before and eliminate the term containing
λ* λ 1 − F v N O .
Note that in Eqs. 6.34 - 6.42, λ = λN2O and = ( 2
)
Figure 6.9. A-a partial pressure difference for nitrogen when R > 1.
A-a partial pressure difference for nitrogen as a function of σ , the degree of ventilation-
perfusion mismatch, for ΔV = 0, -25, -50, -75, -100 ml/min. These values of ΔV are
associated with a Respiratory Exchange ratio greater than or equal to 1. The red arrow
shows the direction of decrease in ΔV .
partial pressure of nitrogen in dry air, to produce the curves in Figure 6.8 and
Figure 6.9, the quantitative results are only approximate since our analysis is
to the simplifying assumption that VIA FIN2 = VA FAN2 in each gas-exchanging
longer applies, Eq. 6.2 must be modified to include the second gas, giving rise
to 2 equations that must be solved simultaneously for each gas to yield the mean
integration purposes. We have not gone down this path as the other conclusions
drawn from Figure 6.8 and Figure 6.9 regarding the direction of the alveolar-
arterial partial pressure gradient still apply as they are due to the assumption that
P v = P a for nitrogen.
inert gas pressure is greater than that of the mean alveolar gas” is borne out in
must occur in the presence of a reduction in the system volume i.e. ΔV must be
greater than zero. Thus, the line for ΔV = 0 in Figure 6.8 is associated with no
difference between mean alveolar and mean arterial partial pressure for nitrogen.
Furthermore, if the volume change occurs in the reverse direction, i.e. ΔV < 0,
example further demonstrates the importance of taking net gas volume changes
CONCLUSION
In this chapter, we have investigated the effect of net gas volume loss such
gases present. It was shown that the classical partial pressure-solubility diagram
of West, Wagner and Derks (89) must be modified. At the lower end of the
solubility range, the partial pressure-solubility curve of the second gas in arterial
gas effect in blood compared with alveolar gas. Uptake of the second gas may
of mismatch and the rate of nitrous oxide uptake are sufficiently large and can
occur with highly insoluble gases such as sulphur hexafluoride but is less likely
with agents such as desflurane and sevoflurane in common use today. We have
Rahn (8) for nitrogen in steady-state air breathing, we show that for nitrogen, the
direction of the alveolar-arterial gradient follows the direction of net gas volume
movement.
respiratory physiology, the effect on gas transfer should be the opposite (89, 92).
In this chapter, we use the graphical representation of the log normal distribution
of V Q ratios in the lung developed in Chapter 3 to study more closely the roles
of the first and second gases and how these interact in the presence of V Q
mismatch to produce the SGE. The reader is advised to read Chapter 3 and
practical model for studying the concentration and second gas effects. In this
exchange which can be used to explain how a low solubility and high degree of
lung. If such a gas is delivered in the inspired gas mixture at a rate VI FI and
removed at a rate VA FA , then the uptake of the gas is VI FI − VA FA . This must
equal the uptake by blood which is given by λQ ( Fc ' − F v ) , where Fc ' and F v
( )
are equal to Pc ' PB − P H 2O and P v (P B )
− P H 2O respectively, with Fc ' the
the capillary blood draining the unit and Pc ' the partial pressure of that gas, so
that:
In the case of the physiological gases O2 and CO2, the expression on the
with haemoglobin. Either side of Eq. 7.1 may be taken as the rate of transfer of
gas. Let us therefore denote the rate of transfer of some gas G by the symbol
constant volume VI , the second step during the volume correction from VI to
VG (=
01) VI FI − VI F ′ (7.3)
VG (=
12 ) VI F ′ − VA FA (7.4)
In Eq. 7.3, FI and F ′ are the fractional concentrations in dry gas at the
start and end of step 1 and VG ( 01) is the uptake (or output) of the gas in this
step. Similarly, in Eq. 7.4, F ′ and FA are the fractional concentrations in dry
gas at the start and end of step 2 and VG (12 ) is the uptake (or output) of the gas
in this step. It follows from the addition of Eqs. 7.3 and 7.4 by comparison with
VG ( 02 ) .
Note that the output of Eq. 7.3 becomes the input to Eq. 7.4, i.e. the
output of step 1 becomes the input to step 2. Moreover, this treatment makes no
between the gas G and blood and is therefore always applicable to gas exchange
the right-hand side of Eq. 7.1, where V is the volume of this individual gas-
exchanging unit, we obtain the following equation which now includes the extra
37
Rahn and Fenn used a pathway in explaining the divergence of the gas R-lines on the O2-CO2
diagram (75) but, in doing so, did not apply the principle enunciated here. This is the first time this
principle has been stated explicitly. Note that Kety’s theory of inert gas uptake (see 2.1), only
included step 1 and therefore could not be expected to predict any of the phenomena associated with
step 2.
time:
This manoeuvre allows us to extend the property described in Eqs. 7.3 and
property is that it is always possible to find F ′ such that Eqs. 7.3 and 7.4 are
satisfied.
permissible to sum the terms on both sides of Eq. 7.5. i.e. we must show that the
following applies:
N N N
j
=j 1 =j 1
∑ (VIFI − VA F=
A) ∑ λ Q ( Fc '− F v ) + ∑ d (VFA )
j
=j 1
j
dt (7.6)
where the subscript j refers to the jth gas-exchanging unit and there are N such
units. The value of each term is to be the value applicable in the jth unit. For
example, VI refers to the inspired ventilation in the jth unit, VI j and Q refers to
The sum on the left side of Eq. 7.6 requires that the uptake of gas G
from the inspired alveolar ventilation for the whole lung should equal the sum
of the uptakes from the inflow into the individual gas-exchanging units which is
clearly true. The first sum on the right side of Eq. 7.6 requires that the uptake of
gas G by the total pulmonary blood flow should equal the sum of the uptakes
from by the individual blood flows to each unit which again is obviously true.
( ) ( )
N N
d ∑ V j FA j dt = ∑ d V j FA j dt (7.7)
=j 1 =j 1
We can illustrate that this is true by considering the changes in the four-
Compartment 1 2 3 4 Lung
Units before 1 5 10 4 20
Units after 2 1 2 2 7
Change +1 -4 -8 -2 -13
The number of units of the gas in each compartment before exchange is shown
in row 2, the number remaining after exchange is shown in row 3, the change in
number of units is shown in yellow in row 4. The sixth column shows what is
litre at STPD. It can be seen that the sum of the changes for the 4 compartments
in row 4 (in yellow) is equal to the change in the sum for the lung as a whole as
shown in the last entry in column 6 (in green). As the third sum in Eq. 7.6 applies
to the lung as a whole, we conclude that gas exchange in the lung may always
G ( 02 )
appear in relation to VG in the equation V= VG ( 01) + VG (12 ) :
so that (01) denotes step 1, (12) denotes step 2 and (02) denotes the sum of steps
concentration and second gas effects in the system we have used in previous
chapters, i.e. a system consisting of first gas, second gas and insoluble filler gas.
The first gas (FG) is assumed to be inert and so will obey Eq. 7.1. For the
Assuming complete equilibration between the first gas and blood so that
Pa = PA , these equations give rise to a general equation for the alveolar and
( VA Q t ) + λ (1 − F v )
(VI VA ) = (7.9)
( VA Q t ) + λ (1 − FI )
With a log normal distribution of V Q ratios, Eq. 7.9 may be expanded to give
(7.10)
In Eq. 7.10, the left-hand term in curly brackets generates the distribution of the
expired ventilation while the right-hand term in curly brackets generates the
distribution of the first gas uptake. Both of these distributions are generated as
∞
expressed as the equivalent Lebesgue-Stieljes integral, ∫ V ( x) dx
−∞
I yields the
total inspired alveolar ventilation VI . Once λ , VA , Q t , σ and F v are specified,
the integral provides a one-to-one relationship between VI and FI , and hence
allows the calculation of the inspired concentration of the FG associated with the
Eq. 7.10 equal to 0. Alveolar ventilation VA is set at 4 l/min and pulmonary blood
FG uptake. Graphs of the FG uptake may now be drawn on the same x-axis as
When present as the sole soluble inert gas, the uptake profile of the second
gas (SG) is also given by the right-hand term of Eq. 7.10 in curly brackets. Since
this gas is present in low concentrations (we use a concentration of 1% in the dry
inspired gas mixture), the FI term in the denominator may be ignored so that
equation for the uptake of SG must be multiplied by Eq. 7.9, the correction factor
for the difference between the inspired and expired ventilation of each gas-
exchanging unit.
When second gas uptake occurs in the presence of a soluble first gas, we
set F v for each gas to zero and use the following sequence, which mimics the
0. Both gases are introduced via the inspired alveolar ventilation to each
gas-exchanging unit
1. The SG is then equilibrated with the blood flow to that unit (step 1)
Splitting the gas exchange process into 2 steps in each gas exchanging unit
distributions of inflow, uptake, retention and outflow of first and second gases.
We proceed as follows.
parts. Since we have specified that P v equals zero for each gas under
consideration, we need only consider gas being brought into the lung by the
partition on the interval [-4, 4] can be used throughout. Choosing n = 800 will
then provide 101 points in each unit sub-interval, e.g. from -4 to -3, each of the
n intervals will be 0.01 units wide. This will generally provide sufficient
coverage. We now have a partition { xk } of the interval [-4, 4] with n+1 points
such that −4 = x1 < x2 < x3 < ....xk .... < xn < xn +1 = 4 and define the vector
7.1, we define the following parameters and the associated vectors which are
listed in Table 7.2
x [ x , x , x , x ]
1 2 n n +1
Partition on x-axis with -4 ≤ x ≤ 4
I
V VI1, VI 2 ,VI n −1, VI n Gas inflow for each x
A
V VA1,VA2,VAn −1,VAn
Gas outflow for each x
Q Q1 , Q 2 , Q n −1 , Q n Blood flow for each x
IG (01)
V I FI
V Gas G inflow in step 1 for each x
G
G ( 01)
V VG (01)1 ,VG (01) n Gas G uptake in step 1 for each x 38
G ( 02 )
V VG (02)1 ,VG (02) n Gas G uptake in steps 1 & 2 for each x
A G(01)
V IG (01) - V
V G (01) Gas G outflow from step 1 for each x
G ( 12 )
V G ( 02 ) - V
V G (01) Gas G uptake in step 2 for each x
38
VG(01) is obtained from Eq. 7.14. Note that the sum of the values for all xi has not been converted
to the equivalent Lebesgue-Stieltjes integral. To do this, we would have to substitute for VIk using Eq.
7.10 in Eq. 7.14. This is far too complicated. Instead, we calculate the volume of N2O taken up for
each value of x using the Lebesgue-Stieltjes form of Eq. 4.8 and add it to the value of VAk calculated
for that value of x from Eq. 7.12. This is far more efficient.
VA 2
− ( xk − μ ) 2 VA λ ( FI − F v )
=VIk e +
2π 2π λ (1 − FI ) e( xk − μ ) ( )
+ VA Q t e( xk + μ ) 2 (7.11)
2 2
2
I = VI1,VI2,VIn −1,VIn
V
VA − ( xk − μ )2
VA k =
2
e
2π (7.12)
A = VA ,VA ,VA
V ,VAn
1 2 n −1
Q t − ( xk + μ )2
Q k =
2
e
2π (7.13)
= Q , Q , Q
Q , Q n
1 2 n −1
λVI k ( FI − F v )
VG (01) k =
(
λ + VI Q t k k ) (7.14)
G (01) = VG (01)1 VG (01)n +1
V
VA λ ( FI − F v )
VG (02) =
{
2π λ (1 − FI ) e( xk − μ ) ( ) }
2
+ VA Q t e( xk + μ )
2
2 2
(7.15)
G ( 02 )
V = VG (02)1 , VG (02) 2 ,VG (02) n −1 , VG (02) n
The gas retained at the end of step 1 becomes the outflow of the gas G at the end
V G ( 02 ) - V
G ( 12 ) = V G (01) .
required parameter. The total uptake of gas over the whole range of x values may
MATLAB but this involves such a complicated expression for the integrand in the
case of Eqs. 7.14 and 7.15 that the much simpler option of applying the
trapezoidal rule (4, 36) has been used instead here and for other calculations,
where appropriate.
necessary, equations were solved using the fsolve nonlinear algebraic equations
solver. When required, the integral function was used to integrate improper
integrals. For individual figures involving N2O, values of ΔV were chosen so
as to demonstrate differences when these occur, while also lying within the range
believed to occur clinically during N2O anaesthesia. When the uptake of other
from the sum of the terms in both sets of curly brackets in Eq. 7.10. The expired
ventilation is obtained from the bracketed left-hand term. The resultant inspired
and expired ventilations are shown in the left column of Figure 7.1. The upper
about the line x = σ 2 . The curve for VI is symmetrical about the line x = σ 2
only when σ = 0; for higher values of σ it is asymmetrical and bulges to the left.
= 0, B: σ = 1, C: σ = 2. In each case, the graph for VA is symmetrical about the line x =
σ 2 (vertical green lines). The yellow-shaded area between the two graphs shows where
gas volume uptake occurs. VI and VA are in l/min. Total expired alveolar ventilation is
equal to the area under the lower curve and is set at 4 l/min. RIGHT COLUMN: Inspired
and expired flows of N2O vs. x. The inspired N2O concentration has been adjusted so as
to produce the same total uptake of N2O in each figure. The yellow-shaded area between
the two graphs shows where N2O uptake occurs and is the same size as in the left column.
The inflow and outflow of 1% sevoflurane is shown in red and is visible at the bottom of
the diagram, just above the x-axis. Note that the y-axis has been shifted to the left in B
and C so that the expired alveolar ventilation is aligned with that in A.
In the right column, we have drawn graphs of the inflow and outflow of
the first gas, N2O ( λ = 0.47) and a second gas, in this case sevoflurane ( λ =
0.65). The graphs for N2O are smaller than those for VI in the left column
because the concentration of N2O has been allowed to vary and has been chosen
to maintain the uptake at 575 ml/min, equal to the value that occurs when σ = 2
and FI N2O = 0.7 , so the volume of N2O taken up is always less than the
increased, a higher FI N2O is required to maintain the same net volume uptake.
The yellow-shaded area between the inflow and outflow curves is the same
as in the left column both in size and in its location with respect to the x-axis and
is equal to the volume contraction in l/min due to N2O uptake. The red inflow
and outflow graphs for 1% inspired sevoflurane appear at the bottom of this
figure and are so small as to seem insignificant when compared both with the
curves for 70% N2O and the net volume uptake (yellow area). If we are to show
ventilation unless these are drawn using an appropriate change in scale. Note
how, as the degree of V Q mismatch increases, the uptake of N2O shifts further
and further to the left, following the movement of the blood flow distribution
Let us now consider the uptake of 1% sevoflurane and 70% N2O as a two-
step process in a lung with perfect matching between ventilation and blood flow,
i.e. σ = 0. This is shown in Figure 7.2. In the left panel, sevoflurane exchange is
allowed to occur before the uptake of N2O. The distribution of the inflowing
sevoflurane is shown in blue above the x-axis; that transferred to blood is shown
in red below the x-axis. The distribution of sevoflurane remaining after uptake
is shown in black above the x-axis. The grey area between the black curve and
the x-axis is equal to the volume per minute of sevoflurane which remains after
uptake in step 1. This forms the outflow from step 1 which becomes the inflow
of sevoflurane to step 2. The area of the red region below the x-axis is equal to
the uptake of sevoflurane per minute and matches the area of the white region
between the blue and black curves above the x-axis. The yellow area shows the
distribution of the N2O uptake which is about to occur and the results of which
For reasons of scale discussed above, the yellow area is not equal to the
uptake of N2O per minute but is proportional to it in size. Its value is maintained
constant in the figures that follow. N2O uptake is now permitted to take place
size of the yellow area. This produces an increase in the partial pressure of
shown in red below the x-axis. The sequence is similar to that shown in step 1,
however the black curve in step 1 now becomes the starting point for sevoflurane
uptake in step 2 and is therefore shown in blue. Notice how, in this case of perfect
matching between ventilation and blood flow, the incoming sevoflurane in steps
1 and 2 is perfectly aligned with the uptake of N2O. The right-hand diagram
depicts the SGE which has now been isolated from the contribution to the uptake
concentration will now be examined in the presence of 70% N2O ( λ = 0.47) with
the other gases are included for the sole purpose of investigating further the role
of solubility in producing the SGE. While there is evidence that highly soluble
gases such as acetone are mainly absorbed in the airways rather than in the
alveoli (see 7.16), we use it here purely as an example of a highly soluble gas
to demonstrate the underlying principles. The process is shown for each gas in
two steps in Figure 7.3 and follows the same pattern as in Figure 7.2, with the
SGE shown in the right-hand column. The largest transfer of SG occurs in Figure
7.3 with sevoflurane. Using these diagrams, we can now investigate the factors
in net gas volume in step 2 must be as great as possible. This gas is shown in
each left-hand diagram in grey. Maximization of the grey area will occur if the
loss of SG to blood during the first step is minimized i.e. if the solubility of SG
in blood, λSG , is as low as possible. Thus, in the case of acetone, so much of the
gas present initially has been transferred to blood during step 1, that there is
that for any given set of the other input variables, ( λFG , VA , Qt , σ , FI , F v )
there is a value of λSG for which the SGE is maximized. In the sequence of gases
shown in Figure 7.3, this occurs with sevoflurane. Using the conditions which
apply in Figure 7.3 in Figure 7.4, we have drawn a graph of the total SG uptake
in step 2 vs. log10 ( λSG ) . It can be seen that the maximum occurs for log10 ( λSG )
= -0.41 which corresponds to a value of 0.4 for λSG , consistent with the position
by σ = 2.
the solubility of the second gas, λSG , needing to be as high as possible in step 1
the second step. We can therefore expect that for each value of λSG , there is a
N2O uptake. One fact, however, remains clear - notably that in neither of the two
that V Q mismatch per se can accelerate rather than hinder gas transfer in the
lung. In step 1, we exploit the fact that a high degree of V Q mismatch impedes
gas uptake while in step 2, we exploit the fact that a low degree of V Q
Not only is the magnitude of the volume contraction important but also its
in the case of acetone. Although most of the acetone has been transferred to
blood in step 1, a small amount remains in the high V Q range for exposure to
low V Q region (yellow area at the top of Figure 7.3C). It does not overlap the
remaining acetone and there is therefore virtually no SGE at all. In Figure 7.1,
mismatch on N2O is to shift the uptake curve to the left. Clearly, the effect of
of log10 ( λSG ) with σ varied from 0 to 2 in increments of 0.01. The N2O uptake is kept
constant at 575 ml/min by changing the inspired concentration of N2O. The red curve
represents a value of σ = 0, the blue curve represents a value of σ = 2, as in Figure 7.4.
The direction from red to blue of a vertical line drawn at any given solubility indicates
whether SG uptake in step 2 increases or decreases as σ is increased. Red and blue curves
intersect at points X and Y. To the right of Z, gas exchange occurs almost exclusively in
the airways so no V Q effect will be detectable. The large black arrow points to a small
log10 ( λSG ) for values of σ ranging from 0 to 2. The volume of N2O taken up
has been kept constant at 575 ml/min by varying the inspired concentration of
N2O. The red curve corresponds to the situation when σ = 0; the blue curve
corresponds to the situation when σ = 2. The two curves cross at the points
the likeliest cause of the increased SGE is the improved alignment between the
N2O uptake and the SG retained at the end of step 1. To the right of Y, an increase
here, due to the wide separation of the retained gas and the volume contraction
increased (see Figure 7.3C), the increased SGE must be due to the effect of V Q
right of Z, gas uptake occurs almost exclusively in the airways and is therefore
Thus, we conclude that as we move from left to right along the solubility
axis, the importance of correct alignment wanes while that of retention increases.
In the region between X and Y, these processes are not powerful enough to
In Figure 7.5, the large black arrow points to a small area between X and
Y which has been enlarged in Figure 7.6 in which λSG and σ are shown on the
x-axis and y-axis respectively with the SG uptake on the z-axis. The effect of
increasing the degree of V Q mismatch in this region varies and may involve
further, as shown by the red arrows. The effect is small and probably
insignificant.
concentrations in the inspired gas mixture. We then examine what will happen
further and further into the low V Q zone. We can describe sulphur
The situation for a highly soluble gas is illustrated in Figure 7.7B using
methoxyflurane. It can be seen that the uptake profile shifts further and further
become asymmetrical as σ is increased so the vertical line in each graph is located at the
median value of y. The uptake profile of sulphur hexafluoride shifts to the left as σ
increases; an example of perfusion-dependent uptake; the uptake profile of
methoxyflurane shifts to the right as σ increases; an example of ventilation-dependent
uptake (see direction of arrows in A and B).
by the direction of the arrow in each of these figures. It is possible for the uptake
increased. This is shown in Figure 7.7C and occurs when λ = VA Q t which in
our case is equal to 0.8. Note that we have used the fraction of the total uptake
as the variable on the y-axis rather than its absolute value as we are interested
here in the location of the uptake rather than the fact that the uptake must
solubility of the gas effectively becomes λ (1 − FI ) so, for example 70% N2O
behaves like a gas with λ = 0.47*(1 - 0.7) = 0.14, present in very low
concentrations. As a result, the uptake distribution curve always shifts to the left
for any gas as its inspired concentration is increased. N2O-driven SGEs therefore
physiological gases. However the solubility of these gases is not constant but
varies according to numerous factors which must be taken into account by using
a suitable algorithm for their interaction with haemoglobin, for example (14).
For gases present in very low concentrations, the solubility and effective
It follows from the above discussion that the magnitude of the SGE might
the effect of diethyl ether on sevoflurane in step 2 is compared with that of N2O
in Figure 7.8. 39 It can be seen that the SG uptake is reduced in Figure 7.8B
(diethyl ether) compared with Figure 7.8A (N2O). The shape of the inspired SG
distribution (blue curve) is slightly different in Figure 7.8B than in Figure 7.8A
different locations of the FG uptake in the case of ether and N2O (yellow region
39
Note that according to Anderson et al. (1, 2), diethyl ether absorption occurs to a large extent in the
airways, rather than in the alveoli but we use it here to illustrate the possible effect of a ventilation-
dependent FG by assuming that it is exchanged solely at alveolar level (see 7.16). This assumption
has also been made with respect to methoxyflurane in Figure 7.7B.
We now apply these principles to the first gas, whose uptake is itself
known as the concentration effect (CE) and was first described by Eger (18, 21).
Treating the first gas uptake as a two-step process as before, with the first
step at constant volume, while the second step incorporates all volume changes,
we obtain the results shown in Figure 7.9 for N2O ( λ = 0.47) and diethyl ether
( λ = 12).
uptake at constant volume. The gas transferred to blood is shown in red below
the gas phase, which is corrected in step 2, during which an additional amount
The sum of the red areas in the left and right diagrams in Figure 7.9A and
Figure 7.9B is identical but the volumes arising in each step differ. According to
our previous argument in relation to the SGE, it is the transfer in step 2 that is
alignment of the contraction with the gas retained at the end of step 1 is no longer
a factor but its effect on the retention of FG in step 1 occurs as before. It can be
seen that for equal net gas volume uptake, the CE is greater for the less soluble
agent N2O.
Figure 7.10 shows similar plots to Figure 7.5 and Figure 7.6. It clearly
with the increase being greater at lower solubilities. Thus looking at the gas with
log10 ( λFG ) = 1, (i.e. λ = 10) in Figure 7.10B, the CE increases as we move left
more and more as we move left along the solubility axis and is maximal at the
left end of the solubility axis when log10 ( λFG ) = -0.8, (i.e. λ = 0.16). There are
no decreases in CE similar to those between X and Y with the SGE in Figure 7.5
step 2 on the z-axis. This confirms that the concentration effect for any gas always
It has been shown by Anderson et al. that with λ > 10, gas exchange begins
to take place in the airways, as well as in the alveoli (1, 2). V Q scatter therefore
does not play the same role as it does with gases of lower solubility. The effect
of V Q mismatch in increasing the SGE shown in Figure 7.5 for gases with λ
greater than 10 will therefore only be of importance for the alveolar component
of gas exchange and may not be significant clinically. This serves to further
explain why the effects of V Q mismatch on the SGE were more likely to have
currently in use.
increases. With λ greater than 100, exchange takes place almost exclusively in
the airways so that our use of acetone in the discussions above is purely
hypothetical. Because of the wide separation between the location of the gas that
remains at the end of step 1 and the location of the contraction of volume in step
with a gas such as halothane or methoxyflurane. This is the rationale for using it
augmentation ratio (AR) defined by Epstein et al. (24) as the partial pressure of
the SG in the presence of the FG, divided by that which would have existed in
the absence of the FG, with all other variables kept constant (45, 46). This
between the combined uptake of gas in steps 1 and 2, divided by the uptake in
step 1, i.e. the portion of uptake that is independent of the volume contraction.
This second definition is easily extended to the concentration effect. In the case
of the SGE, it ensures that the divisor is based on the number of molecules of
gas in VI FI rather than the number in VA FI , and therefore prevents the SGE
exchange on volume correction. This applies irrespective of the way the gas
dissolves in blood i.e. it applies whether we are dealing with the physiological
gases which combine with haemoglobin or those gases that simply obey Henry’s
Law. This property of steady-state gas exchange has not been explicitly stated in
the past. Using this property, we have investigated the transfer of gases in the
lung using a two-step model which allows the contribution from net gas volume
volume, which occurs in step 1. Gas uptake in step 2 is identifiable as due to the
concentration or second gas effect depending on whether the gas being studied
effects depends on sufficient inflow of the gas to step 2, which itself depends
critically on how much gas is retained after uptake in step 1. Both the effective
Table 7.3. Factors affecting gas uptake in each step of a mechanistic model of gas exchange.
Low effective solubility Favors retention in gas phase Impedes gas transfer to blood
High effective solubility Favors loss of gas to blood Favors gas transfer to blood
Low degree V Q mismatch Favors loss of gas to blood Favors gas transfer to blood
High degree V Q mismatch Favors retention in gas phase Impedes gas transfer to blood
effective solubilities of the first and second gases. This controls the extent to
which the SG retained at the end of step 1, is exposed to the contraction in gas
The information in Table 7.2 allows us to conclude that both solubility and V Q
for a second gas whose solubility in blood exceeds 10 may not be as important
because airway exchange for such gases becomes progressively more important
mismatch on the SGE to the right of point Y in Figure 7.5 may not be
demonstrable clinically.
represents another new property of inert gases albeit one that may change with
the values of alveolar ventilation and pulmonary blood flow. It indicates whether
the uptake of an inert gas, present in low concentrations in the inspired gas
inert gas is increased from these low levels, its uptake always moves to a lower
V Q region as indicated by the formula for effective solubility given above, i.e.
INTRODUCTION
The astute reader will have noticed by now that apart from a brief mention
on page 68, the terms constant inflow and constant outflow have not been
outflow, it is the total expired ventilation which has its value pre-determined at
some fixed rate (we set it at 4 l/min) while in constant inflow, it is the total
inspired ventilation which has its value pre-determined at the same fixed rate
(see Figure 2.8). With these constraints applied, the inflow of fresh gas is always
less in the constant inflow model at any stage of the washin process. Hence it
comes as no surprise to find that with the constant inflow model, the graph of
equivalent constant outflow case for any gas undergoing transfer from lungs to
blood (58). So far in this thesis, we have been using the constant outflow
discrepancy between the two extreme patterns, in the first instance by using the
the augmentation ratio in the constant inflow case. These are cumbersome and
equations for concentration and uptake is given in the next section. However,
graphs equivalent to those in Figure 5.1 are shown above. These again illustrate the
persistent nature of the second gas effect, even as uptake of the first gas, nitrous oxide,
1. Divergence of the lines from the point (1,1), the point on the x-axis
for both gas phase and blood phase, as expected for a “single
compartment” lung with full equilibration between gas and blood phases
seen to be opposite for the blood and gas phases. In blood, an increase in
the least soluble gas, desflurane, and least for the most soluble gas,
The effects are not as great as in Figure 5.1. This partly because of the
restriction placed on gas inflow but also because, as was alluded to in 7.17,
second gas molecules brought in with the extra inflow of gas are not included in
the reference. This involves a correction of VA VI which must be applied to the
results in Figure 5.1. This correction did not become apparent until the idea of a
two-step model of gas uptake was conceived almost 18 months later. The
obtained by comparing the SGE when all factors except the respiratory pattern
are kept the same. To achieve this, we utilize the two-step model developed in
Chapter 7. The following derivation should indicate to the reader how one deals
8.3.1 EQUATIONS
First Gas
We begin, as in 4.3 with the following two equations which describe the
steady-state uptake of the first gas in a single compartment model of the lung:
(1 − FI )VI =
(1− F ) VA (8.2)
VA =
(1 − FI ) VI (8.3)
(1 − F )
Letting ψ = VI ( λQ t ) , we obtain the following quadratic equation 40 for the
40
Eger also solved a quadratic expression for the constant inflow case, describing it as cumbersome.
With minor exceptions, he virtually ignored this case subsequently whenever he discussed the
concentration and second gas effects (19, 22). Mapleson incorporated both cases into his electric
analogue and gave equal weight to each in subsequent discussions (57).
(1 + ψ + F v ) ± (1 + ψ + F v ) − 4 ( ψFI + F v )
2
F= (8.5)
2
(1 + ψ ) − (1 + ψ )
2
− 4ψFI
F= (8.6)
2
The uptake of the first gas is obtained from either side of Eq. 8.1, and with F v
= 0 is given by:
VFG
=
λQ
2 {
(1 + ψ ) − (1 + ψ )
2
− 4ψFI } (8.7)
In a lung model consisting of n parallel compartments, each with its own inspired
ventilation Vj and blood flow Q j , where j = 1 to n, and defining ψ j = Vj ( λQ )
j
we have the following expression for the total uptake of the first gas:
λ n
VFG ∑ Q j (1 + ψ j ) − (1 + ψ )
2
= j − 4ψ j FI (8.8)
2 j =1
V0 λQ 0
=VA II − IU (8.9)
2π 2 2π
{(1 + g ( x)) − }
2
∞ σ
− x + 2
(1 + g ( x) )
2
=IU
−∞
∫e 2
− 4 FI g ( x) dx (8.11)
Lebesgue-Stieltjes associated with first gas uptake. Letting V0 be the total
2
e −( x − μ ) 2
g ( x) η=
= 2 ηe 2 μx (8.12)
−( x + μ ) 2
e
so that I U becomes:
∞
(1 + ηe ) − (1 + ηe ) 2 μx 2
2
−( x + μ ) 2
IU ∫e
2 μx
= − 4ηFI e 2 μx dx (8.13)
−∞
IU = I 1 - I2 (8.14)
∫ (e ) dx
∞
2 2
−( x + μ ) 2
+ ηe −( x − μ ) 2
=I1 (8.15)
−∞
(e ) − 4ηF e (
∞
2
−( x + μ ) 2
2
−( x − μ ) 2
2
− x2 + μ2 )
I2 = ∫
−∞
+ ηe I dx (8.16)
gas uptake for any value of σ , in much the same way as was done in 7.3.
Moreover, Eq. 8.9 provides the distribution of the gas inflow and outflow from
the lung and gas uptake by blood in a similar manner to that used to determine
V − x − μ 2
VI ( x) = 0 e ( )
2
(8.17)
2π
λQ 0
(
− ( x + μ )2 2
) ( ) ( )
2
2 2 2 − x2 + μ2
VFG ( x)
−( x − μ ) 2 −( x + μ ) 2 −( x − μ ) 2
= e + ηe − e + ηe − 4ηFI e
2 2π
(8.18)
V=
A( x ) VI ( x ) − VFG ( x ) (8.19)
Second Gas
For the second gas, which is present in very low concentrations so that
VIFI SG − VAF=
SG λSG Q t ( FSG − F vSG ) (8.20)
This gives the following solution for the uptake of the second gas:
λSG
=VSG
λSG + VA Q
(VI FI SG − VA F vSG ) (8.21)
step 1:
λSG FI SG
VSG (01) = VI ( x) (8.22)
λSG + VI ( x) Q ( x)
λSG FI SG
VSG (02) = VI ( x) (8.23)
λSG + VA( x) Q ( x)
We now construct the vectors shown in Table 7.1, using the new definitions for
VI ( x ) , VFG and VA ( x ) . These are used to generate results similar to those of
Figure 7.3.
combined both steps into one diagram in which we compare constant outflow
(upper diagram) and constant inflow (lower diagram). The desflurane inflow into
step 1 is shown in stippled green. The region between the stippled green line and
the solid blue line corresponds to the uptake of desflurane in step 1. The region
between the solid blue line and the solid black line corresponds to the uptake of
desflurane in step 2; this distribution is also shown in red below the x-axis. The
left vertical black line is positioned at the maximum in the N2O uptake; the right
hand vertical black line is positioned at the maximum in the distribution of the
It can be shown that when all other factors are kept equal, the main effect
of the constant inflow respiratory pattern is to restrict the migration of the nitrous
oxide uptake to the left when it attempts to follow the movement of blood flow
degree of overlap between the second gas retained at the end of step 1 and the
contraction in volume that takes place in step 2. Although not shown here,
and constant outflow for second gases with solubilities in the range 0.01-100, we
obtain the graphs shown in Figure 8.3. These confirm the predictions based on
Figure 7.3 and Figure 8.2, that augmentation increases for both constant outflow
and constant inflow as the solubility of the second gas decreases, with the effect
more marked for constant inflow, the difference increasing as the degree of V Q
Figure 8.3. Augmentation ratios for constant inflow and constant outflow.
Augmentation ratios for constant inflow and constant outflow with σ = 1 and σ = 2. λSG
has been varied from 0.01 to 100. The variable on the x-axis is log10 ( λSG ) . Inspired
ventilation has been made the same for constant inflow and constant outflow.
inflow and constant outflow come closer together and eventually coincide.
CONCLUSION
Thus, we conclude that when all other factors except V Q mismatch are
eliminated, the constant inflow pattern is associated with a greater SGE than the
constant outflow pattern provided the solubility of the second gas is low enough
Otherwise, the two patterns produce identical augmentation. The cause of this
difference is the better alignment between the SG retained at the end of step1
and the contraction in volume taking place in step 2. The improvement in overlap
follow the movement of the perfusion distribution to the left as the degree of
V Q mismatch is increased.
I set out initially to try and prove Philip Peyton was wrong in his
The very idea seemed preposterous! Yes, there was experimental evidence
supporting the notion but I believed that other factors might have been at play
and that these would become apparent in time. The theoretical basis for his
to me. Like Eger (18, 21), Peyton had used a model with many built-in features
interaction of O2 and CO2 with haemoglobin and the two extreme patterns of
ventilation first identified by Frumin et al. (31) (termed constant inflow and
model, the greater the likelihood of a mistake buried deep in computer code.
The search for a suitable model ended when I came across an old
manuscript by Colburn, Evans and West (10). Their work had lain dormant for
over 40 years. Although it focused on a two-gas system, the other gases merely
behaving as if they were an insoluble filler acting purely as a vehicle for the gas
under study which in anaesthetic terminology becomes the first gas, it was
δx amounts, without interfering with the basic equations of gas exchange. This
together with a second gas present in very low concentrations, mimics the
provide an efficient way of varying the degree of V Q mismatch, one of the key
results are described below as the conclusions of each chapter of this thesis are
presented.
In Chapter 2, previous work on the concentration effect and second gas effect
gas uptake when modern volatile agents are used together with nitrous oxide, is
presented. The important lesson from this chapter is the following explanation
Uptake of significant volumes of gas concentrates each remaining gas in a smaller volume;
this increases its partial pressure and accelerates its uptake. In the case of CO2, its elimination
The preoccupation with the flawed diagram produced by Eger to explain his
author after another. Eger’s reluctance to abandon the diagram, even after Bill
The use of (1 σ ) log ( v q ) as the variable on the x-axis and v or q on the y-axis
providing a stable platform for studies of the concentration and second gas
progression which, when applied, allows one to take full advantage of the
by my colleague Ranjan Dash, that was used in most of the simulations. These
Although the equations themselves are well known, the technique of tagging a
second gas onto these equations has not been suggested before. The minimum
number of components necessary for a study of the SGE are identified. These
are: a first gas which produces the volume changes whose effects are being
gas that acts as a vehicle for delivering the first and second gases.
Chapter 5 addresses the issue of persistent second gas effects. It is shown that
not only does the SGE persist even as N2O uptake falls to maintenance levels
but that as V Q mismatch increases, the partial pressure of the currently used
volatile agents actually increases in blood but decreases in the gas phase. This
has an obvious clinical corollary, namely that MAC readings, based on the
Chapter 6 reassesses theoretical work emanating from the San Diego group,
taking into account previously ignored effects of volume change. The main
finding is the possibility of partial pressure reversal during gas uptake. By this
we mean that second gas uptake seems to occur against a partial pressure
gradient between alveolar gas and arterial blood. This is predicted for agents of
very low solubility with a large contraction in gas volume, in the presence of a
unreported property of gas exchange: gas exchange in the lung can always be
split gas uptake into two steps. Uptake of the SG is permitted to take place in
both steps but FG exchange is only permitted in step 2 where it produces the
contraction in step 2 and that this is the primary site where SG solubility and
mismatch can exert an influence through its impact on the alignment of the
contraction in volume with the remaining SG. Thus, retention and alignment
contraction to gas uptake and both agent solubility and V Q mismatch exert
their effects through these two factors. In the case of the first gas, retention
restriction of gas inflow in this case means that there are always fewer
contraction because it produces better alignment but only for agents with a low
The work in this thesis is a direct continuation of that which came out of
San Diego from West’s group in the 1970s. Until the anomalous behaviour of
the SGE was predicted by Peyton et al. in 2001 (68-70), and the existence of
being unlocked, was over. It began in the laboratory of Wallace Fenn, Hermann
Rahn and Arthur Otis in Rochester, New York during the World War II, and was
York City. Leon Farhi joined the Rochester group, which moved to Buffalo, New
York in 1956, where he was later joined by Albert Olszowka. Meanwhile, John
Diego in 1969 and led a research team which included Peter Wagner and John
Evans. The epoch came to a close with the successful development of the
Multiple Inert Gas Elimination Technique by West’s group in the 1970s and
early 1980s. When Peyton’s predictions were published in the Journal of Applied
Physiology, most of the members of these teams had either died or redirected
was left to the Anaesthetic fraternity to sort out the mechanism that underpins
uptake when he did. Otherwise, what followed may never have come to light.
computer simulation tool does not facilitate a study of the effects described in
this thesis. Philip Peyton has repeatedly urged me to undertake the production of
a more realistic simulation tool. Knowing that the underlying model will require
the simultaneous solution of first order differential equations for oxygen, carbon
Notwithstanding this, I cannot rule out the possibility of such a tool being
provided by MATLAB. Other areas that may warrant further exploration include:
anaesthesia.
research to the two years I spent in the Science Faculty at Melbourne University
before transferring to the Medical Faculty. The quality of the lecturers has not
been surpassed. Mentorship also plays an important role – Ian Ritchie and Alex
Robertson – no one could ask for support from more qualified experts. Reaching
out to people one doesn’t know and have never met, can be a daunting prospect.
41
This would require the simultaneous solution of some 600 equations, 500 of them first order
differential equations, 200 of these non-linear due to the interaction between O2, CO2 and
haemoglobin. A further 300 may become non-linear in the constant inflow case.
possessed by many medical doctors. Neither are they everyone’s cup of tea. With
opportunity to pursue interests other than those directly relevant to their career
physics and mathematics will take the opportunity to broaden their knowledge
in these areas. One never knows when such skills can be put to good use.
and Q in the case of a lognormal distribution is riddled with typos. The steps in
the derivation are so impenetrable as to possibly render the final result “unsafe”.
We therefore present here a step by step derivation to clarify how the final result
Suppose a lung of total volume V has the property that given any two real
numbers x1 < x2 , the portion of the lung with the log of the ventilation per unit
with the log of the blood flow per unit volume between m2 + σ 2 x1 and m2 + σ 2 x2
V x2
∫
2
dV = e − x /2 dx (A.1)
2π x1
Let = x
y m1 + σ1 x , so that = ( y − m1 ) σ1 and dx = dy σ1 .
When x = x1 , =
y m1 + σ1 x1 ; when x = x2 , =
y m1 + σ1 x2
V m1 +σ1 x2 2
dV = ∫ e −( y − m1 ) /2σ12
dy (A.2)
σ 1 2π m1 +σ1 x1
42
I am indebted to Professor Snezhana Abarzhi for her assistance in tracing the steps used by Colburn
et al. John West was unable to cast any further light on the derivation. Sadly, John Evans passed away
in 2016.
V m1 +σ1 x2 2
V = e y ∫ e −( y − m 1 ) /2σ12
dy (A.3)
σ1 2π m1 +σ1 x1
V m1 +σ1 x2 2
V = ∫ e y e −( y − m 1 ) /2σ12
dy (A.4)
σ 1 2π m1 +σ1 x1
V m1 +σ1 x2 2
V = ∫ e y −( y − m1 ) /2σ12
dy (A.5)
σ1 2π m1 +σ1 x1
2
1 y − m1 1
y− =
2 σ1
2σ 12
{2 yσ 12 − ( y − m1 )2 } (A.6)
2
1 y − m1 1
2 {
y− = −( y 2 − 2m1 y + m12 ) + 2σ 12 y} (A.7)
2 σ1 2σ 1
2
1 y − m1 1
2 {
y− = − y 2 + 2m1 y − m12 + 2σ 12 y} (A.8)
2 σ1 2σ 1
2
1 y − m1 1
2 {
y− = − y 2 + 2 y (m1 + σ 12 ) − m12 } (A.9)
2 σ1 2σ 1
(A.10)
2
1 y − m1
y−
2 σ1
=
1
2σ 1
2 { 2
− y − (m1 + σ 12 ) + (m1 + σ 12 ) 2 − m12 } (A.11)
2 2
1 y − m1 − y − (m1 + σ 12 ) (m1 + σ 12 ) 2 − m12
y− = + (A.12)
2 σ1 2σ 12 2σ 12
2 2
1 y − m1 − y − (m1 + σ 12 ) (m1 + σ 12 − m1 )(m1 + σ 12 + m1 )
y− = +
2 σ1 2σ 12 2σ 12
(A.13)
2 2
1 y − m1 − y − (m1 + σ 12 ) σ 12
y −= + m1 + (A.14)
2 σ1 2σ 12 2
2
σ12 − y − ( m 1+σ12 )
V m1 +σ1 x2 m1 +
V =
2 2σ12
σ 1 2π ∫ m1 +σ1 x1
e
e dy (A.15)
σ12
2
− y − ( m1 +σ12 )
V m1 +σ1 x2
2σ12 dy m 1+ 2
V = ∫ e
σ 1
e (A.16)
2π
m1 +σ1 x1
z
Now let = ( y − m1 ) σ1 , =
y m1 + σ1 z , dz = dy σ1 , and substitute in Eq. A.16 to
obtain:
σ2
V x2 −σ1
− x2 m 1+ 21
V = ∫ e 2
dx e
(A.18)
2π x1 −σ1
2
V m 2 + σ22
2
−[ z −σ 2 )]
x2
Q = ∫ e 2
dz e
(A.19)
2π x1
Let =
a σ 2 − σ1 , so that x − a = z − σ 2 to give the result:
σ2
V x2 −σ 2
− ( x − a )2 m 2 + 22
Q = ∫ e 2
dx e
(A.20)
2π x1 −σ 2
These are the results given by Colburn et al. (10) in 6 steps not the 20 steps
x2 −σ1
∫
2 2
How do we obtain the result e − x 2 dx = e − x /2
used in deriving the density
x1 −σ1
σ2
− m 1 + 1
2
Why does the total ventilation of the lung equal Ve
which is implied in
original material used in Figures 2.1 – 2.3, even though the diagrams are freely
available online at the journal site. It was therefore decided that in addition to
pointing the reader to the relevant site, a model of anaesthetic uptake would be
constructed and used to illustrate the point under discussion. Since the
complicated matter of copyright was being investigated while the thesis was
It was decided to use the correct form of Kety’s equations (42) and extend
the model along the lines of Mapleson’s electric analogues (57, 58). Thus Kety’s
dPA
VL =VIPI − VAPA + λQ t ( P v − Pa ) (B.1)
dt
Assuming the balance gas is insoluble in blood, we have the following mass
(1 − FI )VI =
(1 − FA )VA (B.2)
Since
= (
FI PI PB − PH 2O and
= ) ( )
FA PA PB − PH 2O , Eq. B.2 simplifies to:
VI =
(P − P
B H 2O − PA ) VA (B.3)
(P − P
B H 2O − PI )
With constant outflow, we solve Eq. B.3 for VI , with constant inflow, we solve
Eq. B.3 for VA . Note that we take PB as 760 mmHg and PH 2O as 47 mmHg at 37
it makes a negligible contribution to the net gas volume change and is therefore
omitted from Eqs. B.3 and B.4. We use inspired concentrations of 0.5% for the
compartments for the lung, the very rich group of organs in terms of blood supply
(heart, brain, kidneys, liver etc), the muscle group and fat. He does not bother
with the tissues such as tendon and bone which have a poor blood supply.
Here VL is the effective lung volume, VA the alveolar air volume, V p the lung
parenchymal tissue volume, Qb the volume of the lung blood pool in equilibrium
with alveolar gas and λp , λb are the parenchymal/gas and blood/gas partition
volume = 2.75 litre, Vp = 0.6 litre and Qb, = 2 litre so that VL becomes equal to
( 2.75 + 0.6 λp + 2 λb )
Vi λiTi + λbQi
= (B.5)
coefficient and Qi is the volume of blood which is in equilibrium with the tissue.
Where Pi is the partial pressure of agent in the gas phase with which the tissue is in
equilibrium and Q i is the blood flow in l/min to the tissue. Mapleson suggests the
Table B.1. Tissue volume, blood pool and blood flow for various compartments
Compartment Ti Qi Q i
Mapleson takes the expired alveolar ventilation as 4 l/min and the total blood
flow, obtained from the fourth column, is 5.41 l/min. The tissue/gas partition
coefficients for various agents in each compartment are given in the next table.
They have been taken from various sources including Mapleson (56, 57) and
https://en.wikipedia.org/wiki/Blood%E2%80%93gas_partition_coefficient.
The model does not allow for different circulation times to various organs
Gaseous agent λp λb λv λm λf
diethyl ether 12 12 12 12 63
methoxyflurane 13 13 24 18 800
We next present the MATLAB code which solves the equations. For this I
required much assistance from Ranjan Dash as I had not previously used
produced the desired result. Note that all green material must be commented out.
% Model parameters
Pb = 760; % Atmospheric barometric pressure (mmHg)
Ph2o = 47; % Pressure of water vapour (mmHg)
% Constant inflow = ci; Constant outflow = co
ci = 1; co = 2;
% Blood flow to each tissue compartment
Q_v = 4.08; % vrg (L/min)
Q_m = 1.07; % muscle (L/min)
Q_f = 0.26; % fat (L/min)
Q_ = [Q_v Q_m Q_f]; % blood flow vector
Q = sum(Q_);% Pulmonary capillary blood flow (l/min)
% FG = 7;
% SG = 3;
% Parameters = [Pb Ph2o FG SG];
% Initial conditions used in all but Fig 2.3 where the SG partial pressures initially
equal Fi_SG*(Pb-Ph2o)
Pa0_FG = 0; % Partial pressure of FG in mixed alveolar gas at t = 0
Pvrg0_FG = 0; % Partial pressure of FG in MVB of VRG at t = 0
Pmus0_FG = 0; % Partial pressure of FG in MVB of muscle at t = 0
Pfat0_FG = 0; % Partial pressure of FG in MVB of fat at t = 0
P0 = [Pa0_FG,Pvrg0_FG,Pmus0_FG,Pfat0_FG,Pa0_SG,Pvrg0_SG,Pmus0_SG,Pfat0_SG]'; %
Initial conditions
options = [];
SG = 3;
tspan = 0:0.01:40;
flow = co;
blue = [173 216 230; 0 191 255; 30 144 255 ;0 0 255]'/255;
red = [250 128 114; 255 0 0; 220 20 60; 255 0 0 ; 128 0 0]'/255;
Fig_21_FG = [1 7];
for j = 1:length(Fig_21_FG)
FG = Fig_21_FG(j);
num_curves = 5 - j;
Parameters = [Pb Ph2o FG SG flow];
for i = 1:num_curves
Fi_FG = Fi_FG_vec(i);
% State variables
Pa_FG = P(:,1); % Partial pressure of FG in alveoli (mmHg)
ax = gca;
ax.LineWidth = 1.5;
Font_size = 14;
text(25,0.85,'Nitrous oxide','Interpreter','Tex','Fontsize',Font_size)
text(25,0.50,'Ether','Interpreter','Tex','Fontsize',Font_size)
pause
figure(2);
FG = 1;
SG = 3;
tspan = 0:0.01:5;
flow = co;
Parameters = [Pb Ph2o FG SG flow];
Fi_FG_vec = [0.1 0.7];
Fi_SG = 0.005;
for i = 1:length(Fi_FG_vec)
Fi_FG = Fi_FG_vec(i);
% State variables
Pa_SG = P(:,5); % Partial pressure of SG in alveoli (mmHg)
Pvrg_SG = P(:,6); % Partial pressure of SG in MVB of VRG (mmHg)
Pmus_SG = P(:,7); % Partial pressure of SG in MVB of muscle (mmHg)
Pfat_SG = P(:,8); % Partial pressure of SG in MVB of fat (mmHg)
Pv_SG = (Q_v*Pvrg_SG + Q_m*Pmus_SG + Q_f*Pfat_SG)/Q;
xticks([0 1 2 3 4 5])
ylim([0 0.5])
ylim manual
yticks([0 .1 .2 .3 .4 .5])
yticklabels({'0','.1','.2','.3','.4','.5'})
xlabel('Time (min)');
% Prepare y-axis label
fmt='\\fontsize{%d} %s';% format string for piece
fnLbl=@(FS,S) {[sprintf(fmt,FS,S(1)) sprintf(fmt,FS,S(2)) sprintf(fmt,FS-3,S(3))
sprintf(fmt,FS,'/') sprintf(fmt,FS,S(4)) sprintf(fmt,FS-3,S(5))
sprintf(fmt,FS,S(6))]};
lbl=fnLbl(12,'(FAFI)');
str = ['Fraction of Inspired Concentration ', lbl];
newstr = join(str);
ylabel(newstr,'FontSize',12)
box off
ax = gca;
ax.LineWidth = 1.5;
Font_size = 9;
s1 = '0.5% halothane with 70% N_2O';
s2 = '0.5% halothane with 10% N_2O';
text(0.5,0.37,s1,'Interpreter','Tex','Fontsize',Font_size,'Rotation',0)
text(3.1,0.26,s2,'Interpreter','Tex','Fontsize',Font_size,'Rotation',0)
pause
flow = co;
tspan = 0:0.01:10;
Kolours = {'b','g','r'};
for j=1:length(SG_)
SG = SG_(j);
Parameters = [Pb Ph2o FG SG flow];
% Initial conditions
Pa0_FG = 0; % Partial pressure of FG in mixed alveolar gas at t = 0
Pvrg0_FG = 0; % Partial pressure of FG in MVB of VRG at t = 0
Pmus0_FG = 0; % Partial pressure of FG in MVB of muscle at t = 0
Pfat0_FG = 0; % Partial pressure of FG in MVB of fat at t = 0
Pa0_SG = Fi_SG*(Pb-Ph2o); % Partial pressure of FG in mixed alveolar gas at t =
0
Pvrg0_SG = Fi_SG*(Pb-Ph2o); % Partial pressure of FG in MVB of VRG at t = 0
Pmus0_SG = Fi_SG*(Pb-Ph2o); % Partial pressure of FG in MVB of muscle at t = 0
Pfat0_SG = Fi_SG*(Pb-Ph2o); % Partial pressure of FG in MVB of fat at t = 0
P0 = [Pa0_FG,Pvrg0_FG,Pmus0_FG,Pfat0_FG,Pa0_SG,Pvrg0_SG,Pmus0_SG,Pfat0_SG]';
% State variables
Pa_SG = P(:,5); % Partial pressure of SG in alveoli (mmHg)
Pvrg_SG = P(:,6); % Partial pressure of SG in MVB of VRG (mmHg)
Pmus_SG = P(:,7); % Partial pressure of SG in MVB of muscle (mmHg)
Pfat_SG = P(:,8); % Partial pressure of SG in MVB of fat (mmHg)
Pv_SG = (Q_v*Pvrg_SG + Q_m*Pmus_SG + Q_f*Pfat_SG)/Q;
colour = Kolours{j}
plot(t,(Pa_SG/(Fi_SG*(Pb-Ph2o))),colour,'LineWidth',2); hold on
end
box off
ax = gca;
ax.LineWidth = 1.5;
xticks([-5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10])
xticklabels({'-5','-4','-3','-2','-1','0','1','2','3','4','5','6','7','8','9','10'})
Font_size = 12;
s1 = ['70% N_2O' newline 'Added'];
text(-4,1.2,s1,'Interpreter','Tex','Fontsize',Font_size,'Rotation',0)
annotation('arrow',[4.2 5.7]/15,[1.6 0.59]/5,'LineWidth',4);
pause
figure(4)
% Commonly used inspired concentrations of N2O, des, sevo, iso and halo as first gas
FG_vec = [1 6 5 4 3]; % identifies first gas so appropriate
inspired concentration is chosen for calculations in ode15s
Fi_FG_vec = [0.75 0.06 0.01 0.015 .01]; % Lists the FG concentration commonly
used
SG = 3;
tspan = 0:0.01:30;
flow = co;
Kolours = ['r' 'b' 'y' 'c' 'g'];
num_curves = length(Fi_FG_vec);
for i = 1:num_curves
FG = FG_vec(i);
Fi_FG = Fi_FG_vec(i);
Parameters = [Pb Ph2o FG SG flow];
% State variables
Pa_FG = P(:,1); % Partial pressure of FG in alveoli (mmHg)
Pvrg_FG = P(:,2); % Partial pressure of FG in MVB of VRG (mmHg)
Pmus_FG = P(:,3); % Partial pressure of FG in MVB of muscle (mmHg)
Pfat_FG = P(:,4); % Partial pressure of FG in MVB of fat (mmHg)
col = Kolours(i);
plot(t,Pa_FG/(Fi_FG*(Pb-Ph2o)),'Color',col,'LineWidth',1.0);
hold on
pause
end
box off
xlabel('Time (min)','FontSize',12);
ax = gca;
ax.LineWidth = 1.5;
Font_size = 9;
text(22.5,1.0,'Nitrous oxide','Interpreter','Tex','Fontsize',Font_size)
text(22.5,0.92,'Desflurane','Interpreter','Tex','Fontsize',Font_size)
text(22.5,0.80,'Sevoflurane','Interpreter','Tex','Fontsize',Font_size)
text(22.5,0.63,'Isoflurane','Interpreter','Tex','Fontsize',Font_size)
text(22.5,0.50,'Halothane','Interpreter','Tex','Fontsize',Font_size)
function dP = Mapleson_ODE(~,P,Fi_FG,Fi_SG,Parameters)
% This function solves the first order differential equations
% The vectors VL_, VT_,BP_, and Q_ are shown but only Q_ is actually used
ci = 1; co = 2;
% Unpack the entries in Parameters
Pb = Parameters(1);
Ph2o = Parameters(2);
FG = Parameters(3);
SG = Parameters(4);
flow = Parameters(5);
%pause
% Lung volumes
VL_a = 2.75; % Volume of alveolar air (litre)
VL_p = 0.6; % Volume of parenchymal tissue (litre)
VL_b = 2; % Volume of blood in equilibrium with lung compartment (litre)
%VL_ = [VL_a VL_p VL_b]; % Volume vector
% Tissue volumes
VT_v = 6.2; % very rich group (litre)
VT_m = 39.2; % muscle group (litre)
VT_f = 12.2; % fat (litre)
%VT_ = [VT_v VT_m VT_f]; % Volume vector
if flow == co
Ve = 4.0; % Expired alveolar ventilation (l/min)
else
Vi = 4.0; % Inspired alveolar ventilation
end
Q = sum(Q_);% Pulmonary capillary blood flow (l/min)
% State variables
dP = [dPa_FG,dPvrg_FG,dPmus_FG,dPfat_FG,dPa_SG,dPvrg_SG,dPmus_SG,dPfat_SG]';
% dP contains the solutions to all 8 differential equations.
The computer programs used to produce Figure 5.1 and Figure 8.1 were
written in Microsoft Visual Basic. It was only after running into problems with
would be beneficial. The power and scope of this language soon became evident.
I will be forever grateful to Ranjan Dash for insisting that I learn MATLAB if I
wished to continue receiving his help in debugging faulty code. It is not practical
to include the code of every program written to produce the results used in this
thesis. Instead, a selection of these programs will be presented. Note that all
APPENDIX C.1: ROUTINES USED TO DRAW FIGURES 6.1, 6.2 AND 6.3
% Pp_solubility_dV_main
% This program furnishes the partial pressure-solubility diagram for
the second gas in the form of PG2/Pi2. Values are obtained for mixed
alveolar gas and mixed pulmonary end-capillary blood.
% A log normal distribution of ventilation and blood flow is
assumed, having a degree of mismatch given by sigma.
% WE STUDY THE EFFECT OF CHANGING (Vi - Va) I.E. WE DON'T FIX FiN2O
% The volume change dV = Vi - Va is kept constant as sigma is
varied.
% This means that FiN2O is changed to keep dV constant.
% The values of FiN2O are determined by the program
FiN2O_Fsolve_main.m found in the Fsolve_FiN2O directory.
An equivalent function is performed below by the function
Call_Fsolve_FiFG_25Mar19.m
global Pb PH2O
Pb = 760; % Barometric pressure of the air, mmHg
PH2O = 47; % Partial pressure of H2O in the
inspired air, mmHg (T=37oC)
dt = 0.01;
N = 2*(1/dt-1); %Partition used in applying trapezoidal rule.
% TABLE 1
% sigma 0 0.5 1 1.5 2 dV
FiN2O_(1,:) = [0.0334 0.0353 0.0413 0.0440 0.0712]; % 0.05
FiN2O_(2,:) = [0.0659 0.0696 0.0812 0.1025 0.1379]; % 0.10
FiN2O_(3,:) = [0.0977 0.1030 0.1198 0.1504 0.2004]; % 0.15
FiN2O_(4,:) = [0.1287 0.1356 0.1571 0.1961 0.2588]; % 0.20
FiN2O_(5,:) = [0.1589 0.1673 0.1932 0.2397 0.3134]; % 0.25
FiN2O_(6,:) = [0.1885 0.1982 0.2280 0.2814 0.3645]; % 0.30
FiN2O_(7,:) = [0.2174 0.2283 0.2618 0.3212 0.4123]; % 0.35
FiN2O_(8,:) = [0.2456 0.2576 0.2944 0.3592 0.4570]; % 0.40
FiN2O_(9,:) = [0.2732 0.2862 0.3260 0.3955 0.4987]; % 0.45
FiN2O_(10,:)= [0.3002 0.3141 0.3566 0.4302 0.5376]; % 0.50
FiN2O_(11,:)= [0.3266 0.3413 0.3862 0.4633 0.5740]; % 0.55
FiN2O_(12,:)= [0.3525 0.3679 0.4149 0.4949 0.6079]; % 0.60
FiN2O_(13,:)= [0.3777 0.3937 0.4426 0.5251 0.6395]; % 0.65
FiN2O_(14,:)= [0.4024 0.4190 0.4694 0.5539 0.6690]; % 0.70
FiN2O_(15,:)= [0.4267 0.4437 0.4954 0.5814 0.6964]; % 0.75
FiN2O_(16,:)= [0.4504 0.4678 0.5206 0.6076 0.7220]; % 0.80
PA2_Pi2_0_0 = PA2_Pi2;
Pa2_Pi2_0_0 = Pa2_Pi2;
% We could just calculate PA2_Pi2 and Pa2_Pi2 values for each value
of sigma and dV at the start and just use them when necessary in
drawing the respective graphs but the Get_PAPi routine is so fast
that I often just call it again and again.
% Next we obtain values for PA/Pi and Pa/Pi with dV = 0 and sigma =
0 and 1.5
% NB when sigma = 0, PA_Pi = Pa_Pi, as we have a homogeneous lung.
DV_choice = 0 / 1000;
DV = DV_choice;
sigma = 1.5;
[PA2_Pi2,Pa2_Pi2] = Get_PAPi_PaPi(DV,sigma,Other);
PA2_Pi2_15_0 = PA2_Pi2;
Pa2_Pi2_15_0 = Pa2_Pi2;
xlabel('log_{10}(\lambda)')
ylabel({'P_A/P_I ';' ';'or ';' ';'P_a/P_I
'})
hYLabel = get(gca,'YLabel');
set(hYLabel,'rotation',0,'VerticalAlignment','middle')
num_fig = num2str(num_graph);
set(gca,'Fontsize',16,'Linewidth',1.5,'box','on','FontName', 'Times
New Roman', 'FontWeight', 'bold')
% Next we obtain values for PA/Pi and Pa/Pi with sigma = 0 and dV =
800 mL/min
% NB when sigma = 0, PA_Pi = Pa_Pi, as we have a homogeneous lung.
DV = 600 / 1000;
sigma = 0;
[PA2_Pi2,~] = Get_PAPi_PaPi(DV,sigma,Other);
PA2_Pi2_0_600 = PA2_Pi2;
%*******************************************
% Next we obtain Fig. 6.1C by adding the results in Fig. 6.1A and
6.1B
num_graph = 3;
figure(num_graph);
set(figure(num_graph),'Units','inches','Position',[0 0.3 20 10]);
xlim([-4 4])
ylim([0 1.5])
xlabel('log_{10}(\lambda)')
ylabel({'P_A/P_I ';' ';'or ';' ';'P_a/P_I
'})
% Next we obtain values for PA/Pi and Pa/Pi with dV = 0 and sigma =
0 and 1.5
% NB when sigma = 0, PA_Pi = Pa_Pi, as we have a homogeneous lung.
PA2_PI2_15_0 = PA2_Pi2_15_0+PA2_Pi2_0_600-PA2_Pi2_0_0;
Pa2_PI2_15_0 = Pa2_Pi2_15_0+PA2_Pi2_0_600-PA2_Pi2_0_0;
%**************************************************************
% Graphs for Fig. 6.2A and 6.2B
num_graph = 4;
figure(num_graph);
set(figure(num_graph),'Units','inches','Position',[0 0.3 20 10]);
xlim([-4 4])
ylim([0 1.5])
xlabel('log_{10}(\lambda)')
ylabel({'P_A/P_I '})
hYLabel = get(gca,'YLabel');
set(hYLabel,'rotation',0,'VerticalAlignment','middle')
set(gca,'Fontsize',16,'Linewidth',1.5,'box','on','FontName', 'Times
New Roman', 'FontWeight', 'bold')
% Next we obtain values for PA/Pi and Pa/Pi with sigma = 0 and dV =
800 mL/min
% NB when sigma = 0, PA_Pi = Pa_Pi, as we have a homogeneous lung.
PA2_Pi2_0_600 = PA2_Pi2;
DV = 0;
sigma = 1.5;
[PA2_Pi2,~] = Get_PAPi_PaPi(DV,sigma,Other);
PA2_Pi2_15_0 = PA2_Pi2;
% Next we obtain PA/Pi and Pa/Pi with sigma = 1.5 and dV = 600
mL/min
DV = 600/1000;
[PA2_Pi2,~] = Get_PAPi_PaPi(DV,sigma,Other);
PA2_Pi2_15_800 = PA2_Pi2;
pause
xlabel('log_{10}(\lambda)')
ylabel({'P_a/P_I '})
hYLabel = get(gca,'YLabel');
set(hYLabel,'rotation',0,'VerticalAlignment','middle')
% Next we obtain values for PA/Pi and Pa/Pi with sigma = 0 and dV =
800 mL/min
% NB when sigma = 0, PA_Pi = Pa_Pi, as we have a homogeneous lung.
DV = 600 / 1000;
sigma = 0;
[~,Pa2_Pi2] = Get_PAPi_PaPi(DV,sigma,Other);
Pa2_Pi2_0_600 = Pa2_Pi2;
DV = 0;
sigma = 1.5;
[~,Pa2_Pi2] = Get_PAPi_PaPi(DV,sigma,Other);
Pa2_Pi2_15_0 = Pa2_Pi2;
% Next we obtain PA/Pi and Pa/Pi with sigma = 1.5 and dV = 600
mL/min
DV = 600/1000;
[~,Pa2_Pi2] = Get_PAPi_PaPi(DV,sigma,Other);
Pa2_Pi2_15_600 = Pa2_Pi2;
pause
%*************************************
% AUGMENTATION RATIOS
% We plot ARg and ARb for values of (Vi-Va) = 200 and 600 mL/min
% with sigma = 1.5.
% As we already have PA/Pi and Pa/Pi wnen sigma = 0 and sigma = 1.5
% for dV = 0 and dV = 600 mL/min, we only need to determine values
for sigma = 1.5 and dV = 200 mL/min
% Next we obtain PA/Pi and Pa/Pi with sigma = 1.5 and dV = 200
mL/min
DV = 200 / 1000;
sigma = 1.5;
[PA2_Pi2,Pa2_Pi2] = Get_PAPi_PaPi(DV,sigma,Other);
PA2_Pi2_15_200 = PA2_Pi2;
Pa2_Pi2_15_200 = Pa2_Pi2;
xlim([-4 4])
ylim([1 1.6])
xlabel('log10(\lambda)','FontName','Times New Roman','FontSize',40)
ylabel('Augmentation ratio','FontName','Times New
Roman','FontSize',40)
set(gca,'Fontsize',20,'Linewidth',1.5,'box','on')
y1 = get(gca,'ylim');
idx = log10(0.007);
plot([idx idx], y1,'g','LineStyle', ':','Linewidth',2)
hold on
idx = log10(13);
plot([idx idx], y1,'g','LineStyle', ':','Linewidth',2)
hold on
ARg = PA2_Pi2_15_600./PA2_Pi2_15_0;
ARb = Pa2_Pi2_15_600./Pa2_Pi2_15_0;
plot(log10(L),ARg,'b',log10(L),ARb,'r','LineStyle','-',
'Linewidth',3)
hold on
txt = 'A';
text(3.0,1.5,txt,'FontSize',40,'FontName','Times New Roman')
hold on
pause
%***************************************
% Now plot ARg and ARb for (Vi-Va) = 800 mL/min for various values
of sigma
% This is Fig. 6.3B
num_graph = 7;
figure(num_graph);
set(figure(num_graph),'Units','inches','Position',[0 0.3 20 10]);
% plot(log10(L),ARg,'b',log10(L),ARb,'r','LineStyle','-',
'Linewidth',3)
% hold on
xlim([-4 4])
sigma = 0.5;
DV = 0;
[PA2_Pi2,Pa2_Pi2] = Get_PAPi_PaPi(DV,sigma,Other);
PA2_Pi2_5_0 = PA2_Pi2;
Pa2_Pi2_5_0 = Pa2_Pi2;
DV = 800 / 1000;
[PA2_Pi2,Pa2_Pi2] = Get_PAPi_PaPi(DV,sigma,Other);
PA2_Pi2_5_800 = PA2_Pi2;
Pa2_Pi2_5_800 = Pa2_Pi2;
ARg = PA2_Pi2_5_800./PA2_Pi2_5_0;
ARb = Pa2_Pi2_5_800./Pa2_Pi2_5_0;
plot(log10(L),ARg,'b',log10(L),ARb,'r','LineStyle','-',
'Linewidth',3)
hold on
pause
sigma = 1;
DV = 0;
[PA2_Pi2,Pa2_Pi2] = Get_PAPi_PaPi(DV,sigma,Other);
PA2_Pi2_10_0 = PA2_Pi2;
Pa2_Pi2_10_0 = Pa2_Pi2;
DV = 800 / 1000;
[PA2_Pi2,Pa2_Pi2] = Get_PAPi_PaPi(DV,sigma,Other);
PA2_Pi2_10_800 = PA2_Pi2;
Pa2_Pi2_10_800 = Pa2_Pi2;
ARg = PA2_Pi2_10_800./PA2_Pi2_10_0;
ARb = Pa2_Pi2_10_800./Pa2_Pi2_10_0;
plot(log10(L),ARg,'b',log10(L),ARb,'r','LineStyle','-',
'Linewidth',3)
hold on
pause
sigma = 1.5;
DV = 0;
[PA2_Pi2,Pa2_Pi2] = Get_PAPi_PaPi(DV,sigma,Other);
PA2_Pi2_15_0 = PA2_Pi2;
Pa2_Pi2_15_0 = Pa2_Pi2;
DV = 800 / 1000;
[PA2_Pi2,Pa2_Pi2] = Get_PAPi_PaPi(DV,sigma,Other);
PA2_Pi2_15_800 = PA2_Pi2;
Pa2_Pi2_15_800 = Pa2_Pi2;
ARg = PA2_Pi2_15_800./PA2_Pi2_15_0;
ARb = Pa2_Pi2_15_800./Pa2_Pi2_15_0;
plot(log10(L),ARg,'b',log10(L),ARb,'r','LineStyle','-',
'Linewidth',3)
sigma = 2;
DV = 0;
[PA2_Pi2,Pa2_Pi2] = Get_PAPi_PaPi(DV,sigma,Other);
PA2_Pi2_20_0 = PA2_Pi2;
Pa2_Pi2_20_0 = Pa2_Pi2;
DV = 800 / 1000;
[PA2_Pi2,Pa2_Pi2] = Get_PAPi_PaPi(DV,sigma,Other);
PA2_Pi2_20_800 = PA2_Pi2;
Pa2_Pi2_20_800 = Pa2_Pi2;
ARg = PA2_Pi2_20_800./PA2_Pi2_20_0;
ARb = Pa2_Pi2_20_800./Pa2_Pi2_20_0;
plot(log10(L),ARg,'b',log10(L),ARb,'r','LineStyle','-',
'Linewidth',3)
hold on
pause
x = [0.22 0.5];
y = [0.33 0.7];
annotation('textarrow',x,y,'LineWidth',2)
hold on
x = [0.8 0.6];
y = [0.3 0.2];
annotation('textarrow',x,y,'LineWidth',2)
hold on
txt = 'B';
text(3.0,2.1,txt,'FontSize',40,'FontName','Times New Roman')
hold on
%***************************
return
%DV_pos = [ 0 1 2 3 4 5 6 7 8 9 10 11
12 13 14 15 16 ];
%DV_val = [0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.4 0.45 0.50
0.55 0.60 0.65 0.70 0.75 0.80];
%sig_pos = [ 0 1 2 3 4 ];
%sig_val = [0.0 0.5 1.0 1.5 2.0 ];
%Other = {L,L1,dt,V0,Q0,DV_val,sig_val,FiN2O_,Fv1}
L = Other{1}
L1 = Other{2};
dt = Other{3};
V0 = Other{4};
Q0 = Other{5};
FiN2O_ = Other{6};
Fv1 = Other{7};
num_sols = length(L);
N = 2*(1/dt-1);
sig_pos = 2*sigma + 1;
DV_pos = DV*20;
if DV == 0
Fi1 = 0;
else
Fi1 = FiN2O_(DV_pos,sig_pos);
end
PA2_Pi2 = zeros(1,num_sols);
Pa2_Pi2 = zeros(1,num_sols);
for j = 1:num_sols
L2 = L(j);
t = zeros(1,N+1);
PA_Pi = zeros(1,N+1);
Pa_Pi = zeros(1,N+1);
for i = 1:N+1
t(i) = -1 + i*dt;
ti = t(i);
x = ti/(1-ti*ti);
dx = dt*(1+ti^2)/(1-ti^2)^2;
f_e = (V0/Q0)*exp(-(2*sigma*x-sigma*sigma)/2);
f_x = ((f_e + L1*(1-Fv1))/(f_e + L1*(1-Fi1)))*f_e/(f_e +
L2);
PA_Pi(i) = (exp(-x*x/2)) * f_x * dx;
Pa_Pi(i) = (exp(-(x-sigma)*(x-sigma)/2)) * f_x * dx;
end
APPENDIX C.2 ROUTINES USED TO DRAW FIGURE 7.4 & FIGURE 7.5
% Main program of routine to draw Figs 8A and 8B
V0 = 4;
Q0 = 5;
V0_Q0 = V0/Q0;
sigma = 0:0.01:2;
num_sigmas = length(sigma);
L2_max_xu = zeros(1,num_sigmas);
Max_L2_value = zeros(1,num_sigmas);
for k = 1:num_sigmas
%% Set the value of s being investigated
s = sigma(k);
VarVec = [V0 Q0 s L1 Fv];
Guess = 0.5;
dx = 0.01;
x = -4+s/2:dx:4+s/2;
num_points = length(x);
% Now allocate memory for all variables changing with x for this
value of sigma
Va = zeros(1,num_points);
V1i = zeros(1,num_points);
V1u = zeros(1,num_points);
V1r = zeros(1,num_points);
Vi = zeros(1,num_points);
Qi = zeros(1,num_points);
V2i = zeros(1,num_points);
V2u_01 = zeros(1,num_points);
V2r_01 = zeros(1,num_points);
V2u_02 = zeros(1,num_points);
V2u_12 = zeros(1,num_points);
for i = 1:num_points
z = x(i);
% Work on Vi first for this value of sigma and V1
Va(i) = (V0/sqrt(2*pi))*exp((-(z-s/2)^2)/2);
V1u(i) = 1/((L1p*exp(((z-
s/2)^2)/2))+V0_Q0*exp(((z+s/2)^2)/2));
V1u(i) = (L1*F1i*V0/sqrt(2*pi))*V1u(i); % uptake of FG
Vi(i) = Va(i)+V1u(i);
V1i(i) = Vi(i)*F1i; % inflow of FG
V1r(i) = V1i(i) - V1u(i); % remainder of FG after uptake
Qi(i) = (Q0/sqrt(2*pi))*exp((-(z+s/2)^2)/2);
end
SGE = sum(V2u_12)-(V2u_12(1)+V2u_12(num_points))/2;
SGE = SGE * dx;
G2u_12(k,j) = SGE;
G2u_12;
[Max_L2xu, I] = max(G2u_12(k,:).');
Max_Uptake_soly = L2vec(I);
L2_max_xu(k) = Max_L2xu;
Max_L2_value(k) = Max_Uptake_soly;
Correct_AR;
end
num_fig = 1;
figure(num_fig); set(figure(num_fig),'Units','inches','Position',[0
0.5 20 10]);
ylim([0 .0025])
xlim([-4 3])
xlabel('Solubility of second gas in blood, log_{10}(\lambda_{SG)')
ylabel('Second gas uptake in step 2 (l/min x 10^3)')
Title = 'Second gas uptake in step 2 as a function of
log_{10}(\lambda_{SG})';
title(Title)
hold on
%Draw a vertical line at x = 0
y1 = get(gca,'ylim');
idx = 0;
plot([idx idx], y1,'k','LineStyle','-')
hold on;
plot(log10(L2vec),G2u_12(num_sigmas,:),'b','LineWidth',4)
hold on
pause
num_fig = 2;
figure(num_fig); set(figure(num_fig),'Units','inches','Position',[0
0.5 20 10]);
ylim([0 .0025])
xlim([-4 3])
xlabel('Solubility of second gas in blood, log_{10}(\lambda_{SG})')
ylabel('Second gas uptake in step 2 (l/min x 10^3)')
Title = 'Second gas uptake in step 2 as a function of
log_{10}(\lambda_{SG}) for various V/Q various V/Q values';
title(Title)
hold on
%Draw a vertical line at x = 0
y1 = get(gca,'ylim');
idx = 0;
plot([idx idx], y1,'k','LineStyle','-')
hold on;
k = 1;
plot(log10(L2vec),G2u_12(k,:),'Color',[1,0.2,0],'LineWidth',10)
hold on
for k = 2:num_sigmas-1
if k == 1
plot(log10(L2vec),G2u_12(k,:),'p','LineWidth',10)
elseif k == num_sigmas
plot(log10(L2vec),G2u_12(k,:),'Color',[0.9,1,0],'LineWidth',1)
end
hold on
end
k = num_sigmas;
plot(log10(L2vec),G2u_12(k,:),'b','LineWidth',4)
hold on
txt = 'ALIGNMENT';
text(-3.0,2.2*10^-3,txt,'FontSize',35)
hold on
% Add letter A
txt = 'A';
text(-0.6,1.9*10^-3,txt,'Fontsize',30)
hold on
txt = 'RETENTION';
text(1.1, 0.85*10^-3,txt,'FontSize',35)
hold on
% Add letter B
txt = 'B';
text(0.8,0.5*10^-3,txt,'Fontsize',30)
hold on
return
[Vars] = fsolve(@Get_FiFG_25Mar19,Guess,options,V1,VarVec);
[Eqs,Fi] = Get_FiFG_25Mar19(Vars,V1,VarVec);
end
% Fi = Fi1
% V1 = the uptake of FG in L/min
% L = L1, the solubility of FG in blood
% Va = total expired alveolar ventilation in L/min
% Q = total pulmonary blood flow in L/min
% sigma = degree of V/Q mismatch
% Fv = Fv1
% Revised on 23/12/2017 by rewriting the function for Vi to try and
prevent Fi1 exceeding 1
% Last revised 30/10/2018
Va = VarVec(1);
Q = VarVec(2);
sigma = VarVec(3);
L = VarVec(4);
Fv = VarVec(5);
Fi = Vars(1);
if (nargout == 1)
varargout = {Eqs};
else
varargout = {Eqs,Fi};
end
end
% Started 8Dec18
% Last Revised 15Sep19
% Coordinates of median
xG_50_V = zeros(1,num_mismatch);
yG_50_V = zeros(1,num_mismatch);
x_vector = zeros(num_mismatch,N+1);
for i = 1:num_mismatch
x_vector(i,:) = x;
end
y_vector = zeros(num_mismatch,N+1);
for i = 1:num_mismatch
sigma = mismatch(i);
mu = sigma/2;
y = 1./((L*exp(((x-mu).^2)/2)+((V0/Q0)*exp(((x+mu).^2)/2))));
y_vector(i,:) = y/I;
num_fig = 1;
figure(num_fig); set(figure(num_fig),'Units','inches','Position',[0
0.3 20 10]);
if i<num_mismatch
pause
end
end
return
[Vars] = fsolve(@Get_G_xy_50,Guess,options,Varvec,I);
[Eqs,x_50,y_50] = Get_G_xy_50(Vars,Varvec,I);
End
V0 = Varvec(1);
Q0 = Varvec(2);
mu = Varvec(3);
L = Varvec(4);
V0_Q0 = V0/Q0;
X = Vars(1);
Y = 1/(L*(exp(((X-mu)^2)/2))+V0_Q0*(exp(((X+mu)^2)/2)));
Y = Y/I;
if (nargout == 1)
varargout = {Eqs};
else
varargout = {Eqs,X,Y};
end
end
V0 = 4;
Q0 = 5;
V0_Q0 = V0/Q0;
L2ml = 1000;
L1 = 0.47; %0.47;
L1 = 0.47;
%% Set the value of sigma being investigated
s = 2;
VarVec = [V0 Q0 s L1 Fv];
% CONSTANT OUTFLOW
Guess = 0.5;
[Eqs,F1i] = Call_Fsolve_FiFG_25Mar19(Guess,dV,VarVec);
L1p = L1*(1-F1i);
F1v = Fv;
F2i = 0.01;
L2 = 0.42;
dx = 0.001;
x = -4+s/2:0.01:4+s/2;
num_points = length(x);
Va = zeros(1,num_points);
V1u = zeros(1,num_points);
Vi = zeros(1,num_points);
Qi = zeros(1,num_points);
for i = 1:num_points
z = x(i);
% Work on fixed Vi first
Va(i) = (V0/sqrt(2*pi))*exp((-(z-s/2)^2)/2);
V1u(i) = 1/((L1p*exp(((z-s/2)^2)/2))+V0_Q0*exp(((z+s/2)^2)/2));
V1u(i) = (L1*F1i*V0/sqrt(2*pi))*V1u(i); % uptake of FG
Vi(i) = Va(i)+V1u(i);
V1i(i) = Vi(i)*F1i; % inflow of FG
V1r(i) = V1i(i) - V1u(i); % remainder of FG after uptake
Qi(i) = (Q0/sqrt(2*pi))*exp((-(z+s/2)^2)/2);
V2i(i) = Vi(i)*F2i;
V2u(i) = L2*F2i*Vi(i)/(L2 + Vi(i)/Qi(i));
V2r(i) = V2i(i) - V2u(i);
y_ll = -0.025*L2ml;
y_ul = +0.025*L2ml;
num_fig = 1;
figure(num_fig); set(figure(num_fig),'Units','inches','Position',[0
0.5 20 10]);
box on
ylim([y_ll y_ul])
xlim([-3 5])
hold on
plot(x,V2i*L2ml,'b','LineStyle','-','LineWidth',4);
hold on
plot(x,-V2u*L2ml,'r','LineWidth',1);
hold on
plot(x,V2r*L2ml,'k','LineWidth',4);
hold on
% Shade scaled contraction
upperY = ones(1,num_points);
upperY = y_ul*upperY;
fill([x fliplr(x)], [upperY fliplr(upperY-(V2i-F2i*Va)*L2ml)], 'y');
% Shade remaining second gas
fill([x fliplr(x)], [V2r fliplr(V2r-V2r)],[0.94 0.99 0.94]);
% Shade absorbed gas
fill([x fliplr(x)], [-V2u fliplr(V2u-V2u)],'r');
%Draw a horizontal line at y = 0
Xl = x(1);
set(gca,'FontSize',20','LineWidth',3);
xticks([-3 -2 -1 0 1 2 3 4 5])
if L2 == 0.0076
print -dtiff -r600 'Fig8A_CO_L.png'
elseif L2 == 0.42
print -dtiff -r600 'Fig8B_CO_L.png'
else
print -dtiff -r600 'Fig8C_CO_L.png'
end
pause
if L2 == 0.0076
print -dtiff -r600 'Fig8A_CO_R.png'
elseif L2 == 0.42
print -dtiff -r600 'Fig8B_CO_R.png'
else
print -dtiff -r600 'Fig8C_CO_R.png'
end
pause
%% CONSTANT INFLOW
%% Note that Vi now equals V0+dV to exactly match the constant
outflow case.
V0 = V0+dV;
% Start by determining Fi for this value of V_fg = dV
Guess = F1i;
[Eqs, F1i] = Call_FSolve_FG_uptake_ci(Guess,V0,Q0,L1,s,dV);
mu = s/2;
eta = V0/(L1*Q0);
for i = 1:num_points
z = x(i);
% Work on fixed Vi first
Vi(i) = (V0/sqrt(2*pi))*exp((-(z-s/2)^2)/2);
Va(i) = Vi(i)-V1u(i);
Qi(i) = (Q0/sqrt(2*pi))*exp((-(z+mu)^2)/2);
V2i(i) = Vi(i)*F2i;
V2u(i) = L2*F2i*Vi(i)/(L2 + Vi(i)/Qi(i));
V2r(i) = V2i(i) - V2u(i);
y_ll = -0.025*L2ml;
y_ul = +0.025*L2ml;
num_fig = num_fig + 1; %Fig 3
figure(num_fig); set(figure(num_fig),'Units','inches','Position',[0
0.5 20 10]);
ylim([y_ll y_ul])
hold on
plot(x,V2i*L2ml,'b','LineStyle','-','LineWidth',4);
hold on
plot(x,-V2u*L2ml,'k','LineWidth',1);
hold on
plot(x,V2r*L2ml,'k','LineWidth',4);
hold on
% Shade scaled contraction
upperY = ones(1,num_points);
upperY = y_ul*upperY;
fill([x fliplr(x)], [upperY fliplr(upperY-(V2i-F2i*Va)*L2ml)], 'y');
% Shade remaining second gas
fill([x fliplr(x)], [V2r*L2ml fliplr(V2r-V2r)],[0.94 0.99 0.94]);
% Shade absorbed gas
fill([x fliplr(x)], [-V2u*L2ml fliplr(V2u-V2u)],'r');
%Draw a horizontal line at y = 0
Xl = x(1);
Xr = x(length(x));
hold on;
plot([Xl Xr], [0 0],'k','LineStyle','-')
hold on
%
xlabel('(1/\sigma) log(v/q)')
ylabel('Second gas flows')
%Title = 'First step of second gas uptake as a function of
(1/\sigma) log(v/q)';
%title(Title)
%Draw a vertical line at x = 0
hold on
if L2 == 0.0076
print -dtiff -r600 'Fig8A_CI_L.png'
elseif L2 == 0.42
print -dtiff -r600 'Fig8B_CI_L.png'
else
print -dtiff -r600 'Fig8C_CI_L.png'
end
pause
if L2 == 0.0076
print -dtiff -r600 'Fig8A_CI_R.png'
elseif L2 == 0.42
print -dtiff -r600 'Fig8B_CI_R.png'
else
print -dtiff -r600 'Fig8C_CI_R.png'
end
return
%------------------------------------------------------------------
% This is the FUNCTION which returns the Fi for the first gas which
produces a particular value for the uptake of first gas V1 (STPD).
% Started 20/08/2017
%------------------------------------------------------------------
[Vars] = fsolve(@Get_FiFG_25Mar19,Guess,options,V1,VarVec);
[Eqs,Fi] = Get_FiFG_25Mar19(Vars,V1,VarVec);
end
1. Anderson JC, Babb AL, and Hlastala MP. Modeling soluble gas exchange
in the airways and alveoli. Ann Biomed Eng 31: 1402-1422, 2003.
2. Anderson JC, and Hlastala MP. Breath tests and airway gas exchange. Pulm
concentration: ongoing relevance and clinical utility. Anaesthesia 68: 512-522, 2013.
2001, p. 458.
5. Biro GP. From the Atmosphere to the Mitochondrion: The Oxygen Cascade.
p. 27-53.
6. Bloom WR, and Boyce P. Alexander Proban Robertson. Aust Math Soc
7. Calvey TN, and Williams NE. Inhalational Anaesthetic Agents. In: Principle
9. Chinard FP, Enns T, and Nolan MF. Diffusion and solubility factors in
Bibliography 205
10. Colburn WE, Evans JW, and West JB. Analysis of effect of the solubility
11. Community EE. Options to reduce nitrous oxide emissions (Final Report)
12. Dantzker DR, Wagner PD, and West JB. Instability of lung units with low
VA/Q ratios during O2 breathing. Journal of applied physiology 38: 886-895, 1975.
13. Dash RK, and Bassingthwaighte JB. Erratum to: Blood HbO2 and HbCO2
dissociation curves at varied O2, CO2, pH, 2,3-DPG and temperature levels. Ann
physiological conditions – evaluation and comparison with other models. Eur J Appl
15. Divatia JV, Vaidya JS, Badwe RA, and Hawaldar RW. Omission of nitrous
oxide during anesthesia reduces the incidence of postoperative nausea and vomiting.
17. Ebert TJ, and Naze SA. Inhaled Anesthetics. In: Clinical Anesthesia, edited
by Barash PG, Calahan MK, Cullen BF, Holt NF, Ortega R, Sharar SR, Stock MC,
18. Eger EI. Applications of a mathematical model of gas uptake. In: Uptake and
Distribution of Anesthetic Agents, edited by Papper EM, and Kitz RJ. New York:
206 Bibliography
19. Eger EI. Effect of inspired anesthetic concentrations on rate of rise of alveolar
20. Eger EI. Inhaled anesthetics:uptake and distribution. In: Miller's Anesthesia,
21. Eger EI. A mathematical model of uptake and distribution. In: Uptake and
Distribution of Anesthetic Agents, edited by Papper EM, and Kitz RJ. New York:
22. Eger EI. Uptake, distribution, and elimination of inhaled anaesthetics. In:
23. Eger EI, Tang M, Liao M, Laster MJ, Solt K, Flood P, Jenkins A, Raines
D, Hendrickx JFA, Shafer SL, Yasumasa T, and Sonner JM. Inhaled anesthetics
24. Epstein RM, Rackow H, Salanitre E, and Wolf GL. Influence of the
Concentration Effect on the Uptake of Anesthetic Mixtures The Second Gas Effect.
Pulmonary Gas Exchange, edited by West JB. New York, New York: Academic Press,
26. Evans JW, Wagner PD, and West JB. Conditions for reduction of pulmonary
27. Farhi LE, and Olszowka AJ. Analysis of alveolar gas exchange in the
Bibliography 207
28. Flippo TS, and Holder Jr WD. Neurologic degeneration associated with
nitrous oxide anesthesia in patients with vitamin B12 deficiency. Archives of Surgery
edited by Longnecker DE, Mackey SC, Newman MF, Sandberg WS, and Zapol WM.
30. Forster RE. Diffusion factors in gases and liquids. In: Uptake and Distribution
of Anesthetic Agents, edited by Papper EM, and Kitz RJ. New York: McGraw-Hill
emesis associated with nitrous oxide than with alternative anesthetics. Anesthesia &
33. Heller ML, Watson TRJ, and Imredy DS. Effect of nitrous oxide uptake on
34. Hendrickx JF, Carette R, Lemmens HJM, and De Wolf AM. Large volume
N2O uptake alone does not explain the second gas effect of N2O on sevoflurane during
35. Hunter JM, and Hemmings HC. The many contributions of WW Mapleson
36. Karris S. Numerical Analysis Using MATLAB and Excel. USA: Orchard
37. Katoh T, Ikeda K, and Bito H. Does nitrous oxide antagonize sevoflurane-
208 Bibliography
38. Kelman GR. Digital computer procedure for the conversion of PCO2 into blood
39. Kelman GR. Digital computer subroutine for the conversion of oxygen tension
40. Kelman GR. A new lung model: an investigation with the aid of a digital
41. Kennedy RR. A Second Look at the Second Gas Effect. Anesthesiology 128:
1053-1054, 2018.
42. Kety SS. The theory and applications of the exchange of inert gas at the lungs
43. Kitihata LM, Taub A, and Conte AJ. The effect of nitrous oxide on alveolar
p. 166.
45. Korman B, Dash RK, and Peyton PJ. Can mathematical modeling explain
the measured magnitude of the second gas effect? Anesthesiology 128: 1075 - 1083,
2018.
46. Korman B, Dash RK, and Peyton PJ. Effect of net gas volume changes on
47. Korman B, Dash RK, and Peyton PJ. Elucidating the roles of solubility and
ventilation-perfusion mismatch in the second gas effect using a two-step model of gas
48. Korman B, and Mapleson WW. Concentration and second gas effects: can
Bibliography 209
49. Kretzschmar M, Kozian A, Baumgardner JE, Batista Borges J, and
50. Lee ASJ, and Sun XG. Response to Letter to the Editor. Anesthesia &
51. Leslie K, Myles PS, Chan MTV, Forbes A, Paech MJ, Peyton PJ, and
Silbert BS. Nitrous oxide and long-term morbidity and mortality in the ENIGMA trial.
anaesthesia and abdominal surgery. Acta Anaesthesiol Scand 27: 167-173, 1983.
reconstructive vascular surgery in the leg. Acta Anaesthesiol Scand 27: 26-34, 1983.
55. Magnusson L, and Spahn DR. New concepts of atelectasis during general
57. Mapleson WW. An electric analogue for uptake and exchange of inert gases
58. Mapleson WW. Inert gas-exchange theory using an electric analogue. J Appl
210 Bibliography
59. Mapleson WW, and Korman B. The Second Gas Effect Is a Valid Concept.
60. McShane EJ. Integration. Princeton, N.J.: Princeton Univ Press, 1944.
61. Merkel G, and Eger EI. A comparative study of halothane and halopropane
62. Myles PS, Leslie K, Chan MTV, Forbes A, Paech MJ, Peyton PJ, Silbert
BS, and Pascoe E. Avoidance of nitrous oxide for patients undergoing major surgery:
63. Myles PS, Leslie K, Chan MTV, Forbes A, Peyton PJ, Paech MJ, Beattie
WS, Sessler DI, Devereaux PJ, and Silbert BS. The safety of addition of nitrous
64. Neufeld GR, Williams JJ, Klineberg PL, and Marshall BE. Inert gas a-A
differences: a direct reflection of V/Q distribution. J Appl Physiol 44: 277-283, 1978.
H, and Goto F. Second gas effect of N2O on oxygen uptake. Can J Anesth 47: 506-
510, 2000.
pulmonary blood flow measurement from soluble gas exchange. In: MD thesis,
Melbourne, 2004.
of the second gas effect on arterial sevoflurane partial pressure. Anesthesiology 108:
381-387, 2008.
Bibliography 211
68. Peyton P, Robinson G, and Thompson B. Effect of ventilation-perfusion
Thompson BR. The rate of alveolar‐capillary uptake of sevoflurane and nitrous oxide
73. Power GP, Ritchie MA, Drok KJ, and Macleod ID. Ian Mackay Ritchie
Model of Gas Uptake. In: Uptake and Distribution of Anesthetic Agents, edited by
Papper EM, and Kitz RJ. New York: McGraw-Hill Book Company, Inc., 1963, p. 98-
103.
75. Rackow H, Salanitre E, and Frumin MJ. Dilution of alveolar gases during
76. Rahn H, and Fenn WO. A Graphical Analysis of the Respiratory Gas
212 Bibliography
77. Ravishankara AR, Daniel JS, and Portmann RW. Nitrous oxide (N2O): the
dominant ozone-depleting substance emitted in the 21st century. Science 326: 123-
125, 2009.
78. Rehder K, Knopp TJ, Sessler AD, and Didier EP. Ventilation-perfusion
79. Riley RL, and Cournand A. "Idea" alveolar air and the analysis of ventilation-
80. Robinson GJB, Peyton PJ, Vartuli GM, Burfoot RB, and Junor PA.
2003.
82. Severinghaus JW. Can large volume N2O uptake alone explain the second gas
83. Severinghaus JW. The rate of uptake of nitrous oxide in man. J Clin Invest
84. Stoelting RK, and Eger EI. An Additional Explanation for the Second Gas
85. Sun XG, Su F, Shi YQ, and Lee C. The" second gas effect" is not a valid
Bibliography 213
86. Taheri S, and Eger EI. A demonstration of the concentration and second gas
effects in humans anesthetized with nitrous oxide and desflurane. Anesth Analg 89:
774-780, 1999.
88. Wagner PD, Saltzman HA, and West JB. Measurement of continuous
90. West JB. Effect of slope and shape of dissociation curve on pulmonary gas
92. West JB, and Wagner PD. Pulmonary gas exchange. In: Bioengineering
aspects of the lung, edited by West JB. New York: Marcel Dekker Inc, 1977, p. 378-
382.
214 Bibliography
Can Mathematical Modeling Explain the Measured
Magnitude of the Second Gas Effect?
Ben Korman, M.D., Ranjan K. Dash, Ph.D., Philip J. Peyton, M.D.
ABSTRACT
Background: Recent clinical studies suggest that the magnitude of the second gas effect is considerably greater on arterial
blood partial pressures of volatile agents than on end-expired partial pressures, and a significant second gas effect on blood
partial pressures of oxygen and volatile agents occurs even at relatively low rates of nitrous oxide uptake. We set out to further
investigate the mechanism of this phenomenon with the help of mathematical modeling.
Methods: Log-normal distributions of ventilation and blood flow were generated representing the range of ventilation-perfusion
scatter seen in patients during general anesthesia. Mixtures of nominal delivered concentrations of volatile agents (desflurane,
isoflurane and diethyl ether) with and without 70% nitrous oxide were mathematically modeled using steady state mass-balance
principles, and the magnitude of the second gas effect calculated as an augmentation ratio for the volatile agent, defined as the
partial pressure in the presence to that in the absence of nitrous oxide.
Results: Increasing the degree of mismatch increased the second gas effect in blood. Simultaneously, the second gas effect
decreased in the gas phase. The increase in blood was greatest for the least soluble gas, desflurane, and least for the most soluble
gas, diethyl ether, while opposite results applied in the gas phase.
Conclusions: Modeling of ventilation-perfusion inhomogeneity confirms that the second gas effect is greater in blood than in
expired gas. Gas-based minimum alveolar concentration readings may therefore underestimate the depth of anesthesia during
nitrous oxide anesthesia with volatile agents. The effect on minimum alveolar concentration is likely to be most pronounced
for the less soluble volatile agents in current use. (ANESTHESIOLOGY 2018; 128:1075-83)
Copyright © 2018, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
<zdoi;10.1097/ALN.0000000000002131>
Ventilation-Perfusion Mismatch and the Second Gas Effect
suggestion that has been questioned,11 but is supported by constant. With no second gas effect, the augmentation ratio is
clinical studies.5,9 1. Once the second gas effect becomes evident, the augmenta-
The idea that a phenomenon traditionally believed to be tion ratio exceeds 1. The term “augmentation ratio” was first
caused by the large reductions in gas volume associated with used by Epstein et al.1 Although they only calculated augmen-
high rates of nitrous oxide uptake can persist at clinically sig- tation ratios in end-tidal gas, the ratio may also be calculated
nificant concentrations even when those volume changes fall for blood.
away is a significant departure from traditional teaching. We We have assumed a log-normal distribution of V! / Q!
set out to further explore the mechanism of this proposed phe- ratios in our investigations.14 Colburn et al.15 have shown
nomenon with computer modeling of theoretical distributions that if σV! is the log SD of ventilation per unit volume and
of V! / Q! scatter seen in patients during general anesthesia, σ Q! is the log SD of blood flow per unit volume, then σ,
across a range of nitrous oxide uptake rates, and the predicted the degree of mismatch between ventilation and blood flow,
effect on second gases with different solubilities in blood. depends only on σV! − σ Q! , the absolute value of the differ-
ence between the two standard deviations. This is illustrated
Materials and Methods in figure 1. In figure 1A, the ventilation as a fraction of the
The minimum number of components in the gas phase neces- total alveolar ventilation (!v) has been drawn in blue while the
sary for a theoretical study of the second gas effect is as follows: blood flow as a fraction of the total pulmonary blood flow
! has been drawn in red. The unit on the x-axis is not log
(q)
A first gas. Alveolar-capillary uptake of this gas generates the (V! / Q! ) but (1/σ)log( v! / q! ), where σ = σV! − σ Q! . Drawn in
volume changes responsible for the second gas effect.
It must be soluble in blood. The difference in partial this way, the graphs of ventilation and blood flow are exactly
pressures between the inspired gas mixture and mixed the same size and shape, and do not change size or shape as
venous blood must be sufficient to produce the desired σ is varied. What does change is the distance between the
volume changes. mean for ventilation and the mean for blood flow. These
A second gas. This is usually a volatile agent used in low con- means are situated exactly σ/2 units on opposite sides of the
centrations. This gas responds to the volume changes by y-axis. When σ = 0, the curves overlap completely so that
exhibiting the second gas effect. matching between ventilation and blood flow is optimal. As
The remaining gas. This acts purely as a “vehicle” for deliver- σ increases, they overlap less and less and the degree of V! / Q!
ing the first and second gases, and for the purposes of mismatch increases. This is shown further in figure 1B, in
the analysis is regarded as being completely insoluble which the blood flow as a fraction of the total pulmonary
in blood. blood flow (q)! has been rotated through 180 degrees about
the x-axis. In awake healthy adults, σ varies from 0.25 to
In clinical practice, the first gas is always nitrous oxide (Ost- 1.0. Worsening of V! / Q! matching is known to occur soon
wald solubility coefficient in blood at 37°C, λ = 0.47). A after induction of anesthesia and has been well documented
fractional concentration of 70% in dry gas is assumed in all experimentally, with typical values for σ of 0.75 to 1.5.16–20
calculations. This is usually regarded as the highest clinically Further information regarding this way of presenting a log-
effective concentration that can be used safely in the general normal distribution of ventilation and blood flow is pro-
population during anesthesia without producing hypoxia.12 vided in appendix 1.
In modern clinical anesthesia, the second gases used are A general equation for the augmentation ratio (AR) was
mainly desflurane, sevoflurane, and isoflurane. We will con- obtained having the following form in which the difference
sider diethyl ether as well because it is an example of a vola- between inspired alveolar ventilation, V!I , and expired alveo-
tile anesthetic with a high solubility in blood. We examine the lar ventilation, V!A, appears:
second gas effect as the uptake of the first gas, nitrous oxide,
decreases toward zero, as is seen with time in clinical use. This is (V!I − V!A )
achieved in our model by increasing the mixed venous pressure AR = 1 + k (1)
V!
A
of nitrous oxide in a stepwise fashion. The proportion of nitrous
oxide washin is expressed as a washin ratio, which mirrors the Derivation of equation 1 is given in appendix 2. In the absence
usual alveolar/inspired concentration ratio, i.e., a washin ratio of nitrous oxide, inspired alveolar ventilation and expired
of 0.9 is roughly equivalent to the situation where the alveolar/ alveolar ventilation are equal, the augmentation ratio equals
inspired concentration ratio equals 0.9. We estimate that our 1, and there is no augmentation. In the presence of nitrous
steady state equations are reasonably accurate within 10 to 15 oxide, inspired alveolar ventilation is greater than expired
min from the start of induction, when nitrous oxide washin is alveolar ventilation, the augmentation ratio is greater than
more than 90% complete and its uptake rate has plateaued.13 1, and the second gas effect appears. Its magnitude is then
The second gas effect is expressed as an augmentation ratio, (
given by the augmentation ratio. The difference V!I − V!A is )
defined as the partial pressure of second gas in the presence
of nitrous oxide, divided by that which would have existed in
(
equal to the rate of nitrous oxide uptake. The V!I − V!A / V!A )
the absence of the nitrous oxide, with all other variables kept term is the “concentrating effect” or “shrinkage” factor given
Copyright © 2018, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
PERIOPERATIVE MEDICINE
Results
Figure 2 shows the results for the blood and gas phases. The
arrow in each diagram of figure 2 indicates the direction in
which σ is increasing. The x-axis shows the proportion of
nitrous oxide washin, and the y-axis shows the augmentation
ratio. The most consistent feature of these diagrams is the
divergence of the lines from the point (1, 1), the point on
the x-axis representing complete washin of the nitrous oxide,
at which time expired alveolar ventilation becomes equal to
inspired alveolar ventilation, so that according to equation 1,
Fig. 1. Log-normal distribution of ventilation and blood flow the second gas effect disappears completely.
drawn so as to show the effect of σ on the degree of match- Each augmentation ratio plot appears to be a straight line.
ing between the two. σ is defined as the absolute value of the The line for σ = 0, which represents a homogenous lung, is
difference between the log SDs of ventilation and blood flow
the same for both gas phase and blood phase, as expected for
each expressed per unit volume (see discussion in Materi-
als and Methods). (A) The y-axis shows the fractional ven-
a “single compartment” lung with full equilibration between
tilation V! / V!A and fractional blood flow Q! / Q! T (represented blood and gas phases. In addition, it is the same for all three
by the symbols v! and q! , respectively); the x-axis is in units gases, i.e., it is independent of the solubility of the second
of (1/σ)log( v! / q! ). The fractional ventilation is shown in blue, gas in blood.
the fractional blood flow in red. For the purpose of this il- The effect of increasing σ and hence the degree of V! / Q!
lustration, σ has been set equal to 2. The figure shows how mismatch is seen to be opposite for the blood and gas phases.
the means of v! and q! are spaced σ/2 units on opposite
In blood, an increase in σ increases the augmentation ratio,
sides of the y-axis (green dashed lines). (B) Here the curve
for blood flow has been rotated by 180 degrees around while in the gas phase, the augmentation ratio decreases.
the x-axis so that −q! is plotted against (1/σ)log( v! / q! ). Moreover, in blood, this amplification is greatest for the least
The region available for gas exchange is located between soluble gas, desflurane, and least for the most soluble gas,
these two graphs. The vertical strip bordered by a dashed diethyl ether, while in the gas phase, the opposite is true.
line represents a single lung compartment in which ventilation
(blue stippled area above the x-axis) has been equilibrated
with blood flow (red stippled area below the x-axis). With Discussion
σ = 0, the graphs of v! and −q! are mirror images, and match- Our mathematical model allows contributions from solu-
ing between ventilation and blood flow is optimal. bility and V! / Q! mismatch to the second gas effect to be
assessed separately. The results show that an increase in
in classical descriptions of the second gas effect. The k term is V! / Q! mismatch augments the second gas effect in blood,
positive and includes contributions from the solubility of the but reduces the second gas effect in the gas phase. This result
is most pronounced for the insoluble gases used in clinical
second gas, the overall alveolar ventilation-perfusion ratio of
practice today. Although not shown here for brevity, results
the lung, and the degree of mismatch between ventilation
obtainable for sevoflurane ( λ = 0.67) are similar to those
and blood flow.
shown for desflurane.
Input variables for each gas were its inspired concentration, The choice of a log-normal distribution is mathemati-
mixed venous partial pressure, and solubility in blood. For sim- cally convenient and has been used previously in simulations
plicity, we have assumed trace concentrations of second gas in of gas exchange in the lungs.6–8,14,21 However, the method
the inspired gas mixture with mixed venous concentrations of used to display the distributions of ventilation and blood
zero. An expired alveolar ventilation of 4 l/min was assumed. flow in figure 1 is new, and illustrates clearly how σ controls
The pulmonary blood flow was set at 5 l/min. Equation 1 was the degree of V! / Q! mismatch. A further advantage of this
first solved for each second gas in a homogenous lung (i.e., method is that it facilitates equilibration of gas and blood
σ = 0), then in a multicompartment lung for a log-normal over the whole range of V! / Q! values, particularly at the
Copyright © 2018, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ventilation-Perfusion Mismatch and the Second Gas Effect
Fig. 2. Second gas effect for desflurane, isoflurane, and diethyl ether as a function of the proportion of washin of nitrous oxide
for different values of σ. The second gas effect is expressed as an augmentation ratio (AR), defined as the partial pressure of the
second gas in the presence of nitrous oxide (N2O), divided by that which would have existed in the absence of the nitrous oxide,
with all other factors kept constant. Upper three diagrams, second gas effect in the gas phase; lower three diagrams, second gas
effect in blood. In each case, the augmentation ratio is plotted on the y-axis as a function of the washin ratio. The long arrow indi-
cates the direction of increasing σ values from 0 to 2 in increments of 0.25 units. The line for σ = 0 is the same for all second gases.
extremities of each log-normal distribution. Log-normal dis- in k, indicating that “shrinkage” need not be the only factor
tributions are not the only distributions seen clinically,16–20 involved in producing the second gas effect. We have shown
but we believe the choice is not particularly relevant to the that k is affected by the degree of V! / Q! mismatch and that
conclusions reached here, which depend primarily on the the effect of V! / Q! mismatch on k differs in the gas phase
application of the Fick principle. and in blood. With no V! / Q! mismatch, σ = 0 and k = 1
Equation 1 provides an explanation of how the second in both phases. This is the situation that corresponds to the
gas effect can continue, even after nitrous oxide uptake classical volume “shrinkage” explanation of the second gas
decreases to levels seen in the “maintenance phase.” The effect. As V! / Q! mismatch increases, σ increases. In blood,
( )
term k V!I − V!A / V!A is responsible for the second gas as σ increases, k also increases and is greater than 1. In the
gas phase, as σ increases, k decreases and is less than 1. The
( )
effect. Here the difference V!I − V!A is produced by nitrous term k may be thought of as a scaling factor, controlling the
oxide. A decrease in this difference as nitrous oxide uptake
decreases can be compensated for by a simultaneous increase
( )
degree to which the volume “shrinkage,” V!I − V!A / V!A , is
expressed.
Copyright © 2018, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
PERIOPERATIVE MEDICINE
Our findings are consistent with the experimental clini- exist. This is consistent with most previous clinical and
cal findings of Peyton et al.,10 that the magnitude of the theoretical studies, including those of Peyton et al., who
second gas effect on arterial sevoflurane partial pressures measured the second gas effect on arterial sevoflurane and
was significantly greater than that measured simultaneously oxygen partial pressures in anesthetized patients.5,10 The aug-
on end-tidal partial pressures in patients immediately after mentation to blood desflurane partial pressures at 20 to 30
induction of anesthesia and for 30 min afterward. However, min predicted by the model is in the order of 10 to 15%
these authors did not attribute the finding to a concomitant with moderate degrees of V! / Q! scatter. This compares to a
diminution in second gas effect in the gas phase. Our model mean 12% increase in arterial sevoflurane partial pressures
provides a more precise explanation for their findings. While at this time point in ventilated patients measured by Peyton
effects in the gas phase are most easily detected using end- et al.10 The lesser augmentation in gas phase desflurane con-
tidal gas concentration monitoring by the infrared analyzers centrations predicted here is also broadly consistent with the
available in most operating theaters today, augmentation in findings of Taheri and Eger, and those of Hendrickx et al.
the blood phase is not measured in normal clinical practice for end-expired sevoflurane concentrations, in anesthetized
but is important because it most directly affects anesthetic patients.4,5
partial pressures in blood and therefore depth of anesthe- Modeling and predicting the effects on respiratory gases
sia. An obvious clinical implication of these findings is is more complex and was not part of the scope of this study.
that minimum alveolar concentration (MAC) calculations In their previous studies using multicompartment model-
based on end-tidal concentration measurements may well ing of lung gas exchange, Peyton et al. modeled oxygen as
underestimate the depth of anesthesia when nitrous oxide a second gas,6–8 and demonstrated significant elevation of
is supplemented with a volatile agent. This represents a new predicted partial pressures of oxygen in arterial blood with
factor that should be taken into consideration in addition moderate degrees of V! / Q! scatter, at 45 min or more into
to other relevant factors22–24 when interpreting an end-tidal inhalational anesthesia with 70% nitrous oxide as the first
MAC reading during anesthesia. gas. This prediction was confirmed in a subsequent clinical
Figure 2 illustrates the predicted effects of typical study.9 In attempting to mimic oxygen uptake as closely as
increases in the degree of V! / Q! scatter on the augmenta- possible, these authors introduced the nonlinear binding of
tion ratio. We can treat the line for the homogenous lung, oxygen to hemoglobin, shunt formation, hypoxic vasocon-
i.e., the line for which σ = 0, as the first line in the series. This striction, absorption atelectasis, carbon dioxide elimination
line is identical in the upper (gas phase) and lower (blood and carriage, and acid-base status. Each of these factors is a
phase) diagrams in figure 2, as expected for a homogeneous potential confounder in interpretation of their findings of
“single compartment” lung. As steady state conditions are the relationship of V! / Q! scatter to the second gas effect. In
approached for the first gas, augmentation in a homogenous contrast, we have chosen to eliminate all but the essential
lung is independent of the second gas solubility in blood. components in this study of the second gas effect, permitting
It depends only on the ratio of the uptake of nitrous oxide explicit mathematical expression of the underlying features,
( )
to the alveolar ventilation, which is equal to V!I − V!A / V!A . once a particular mathematical distribution of V! / Q! ratios
Once ventilation and perfusion are no longer perfectly has been assumed. It remains to be seen to what degree these
matched, solubility becomes the important factor. other factors affect the predictions made here and should be
It is evident from figure 2 that the second gas effect in the subject of further work in the field.
blood and its augmentation by increasing V! / Q! scatter is In summary, modeling of ventilation-perfusion inhomoge-
greater for the least soluble second gas, desflurane. This neity confirms that the second gas effect is greater in blood
finding is remarkable, as lung modeling in the absence of than in expired gas. As the degree of mismatch increases, the
nitrous oxide predicts that increasing V! / Q! scatter, with magnitude of the second gas effect increases in blood but
a reduction in effective alveolar ventilation, impairs alve- decreases in expired gas. The change is most pronounced in
olar-capillary gas exchange more severely for less soluble blood for the less soluble volatile agents in current use. Clini-
gases,15 a prediction recently confirmed in an animal study cally, this means that MAC readings may well underestimate
of desflurane or isoflurane anesthesia using methacholine the depth of anesthesia when nitrous oxide is supplemented
inhalation to induce increased V! / Q! inhomogeneity.25 with a volatile agent.
This would suggest that inclusion of nitrous oxide might
maintain the efficiency of the lung in exchanging less sol- Research Support
uble modern volatile agents in the face of V! / Q! scatter Supported by an Australian Government Research Training
induced by anesthesia itself. Program Fees Offset (to Dr. Korman) and by grant Nos. U01-
Our mathematical modeling suggests, moreover, that the HL122199 and P01-GM066730 from the National Institutes
second gas effect in blood for less soluble modern agents is of Health (Bethesda, Maryland; to Dr. Dash).
expected to remain significant, although progressively reduc-
ing, into the “maintenance phase” of inhalational anesthesia Competing Interests
with nitrous oxide, where typical degrees of V! / Q! scatter The authors declare no competing interests.
Copyright © 2018, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ventilation-Perfusion Mismatch and the Second Gas Effect
Correspondence ⎛ σ⎞
2
1 − ⎝⎜X + 2 ⎠⎟ 2
Address correspondence to Dr. Korman: Department of An- q! = e (1.9)
aesthesia and Pain Medicine, Royal Perth Hospital, Perth, 2π
Western Australia 6000, Australia. ben@korman.com.au.
This article may be accessed for personal use at no charge Equations 1.8 and 1.9 are of the form of a standard normal
through the Journal Web site, www.anesthesiology.org. distribution and have been used to plot figure 1.
These forms of the density function for the log-normal
Appendix 1: The Log-normal Distribution distributions are very useful in equilibrating gas and blood
Colburn et al.15 derive the following probability density compartment by compartment, in order to obtain flow-
functions for ventilation and blood flow from first weighted outputs in gas and blood. The x-axis is divided
principles: into N compartments of equal width. This involves N+1
points. At least 99.7% of cases26 will always be covered if
V! − x 2 the subdivision of the x-axis begins at –σ/2 – 3 and ends at
V! = A e
2
(1.1) σ/2 + 3. The width of each interval is then (σ + 6)/N. When
2π
calculating the contribution of each point to the output, it
is necessary to apply the trapezoidal rule so that the first and
Q!
Q! = T e ( )
2
− x −σ / 2
(1.2) last points carry half the weight of the other points.27 If this
2π is not done, an error will be introduced that becomes more
significant as the number of compartments is reduced. In
Here V!A is the alveolar ventilation and Q!T the total pul- determining the weight attached to the expired ventilation
monary blood flow. Letting v! = V! V!A and q! = Q! Q!T , the from each compartment, it is necessary to multiply v! by
fractional or normalized ventilation and blood flow respec- the expired alveolar ventilation, V!A . Similarly, q! must be
tively, we have multiplied by Q!T .
1 − x2 2
Appendix 2: Derivation of Equation 1
v! = e (1.3)
2π
Washin Ratio
1 − ( x −σ )2 / 2 The proportion of nitrous oxide washin completed is expressed
q! =
2π
e (1.4)
(
as a washin ratio (WR), where WR = 1 − V!N O V!IFIN O .
2 2 )
This ratio is used as an independent variable on the x-axis.
Dividing equation 1.3 by equation 1.4: Note that when V!N2 O = 0, WR = 1 and nitrous oxide uptake is
complete; when V!N2 O = V!IFIN 2O , its maximum possible value,
v!/q! = e ⎣
{ }{ }
⎡ ( x −σ )2 / 2 − x 2 2 ⎤
⎦ (1.5) WR = 0, a situation that can only apply at the commence-
ment of anesthesia (see table A2.1). Since, for simplicity, the
i.e., equations we use are only applicable in the steady state, we
restrict our attention to values of V!N2 O for 0.9 ≤ WR ≤ 1.0,
i.e., when washin of nitrous oxide is 90 to 100% complete
v!/q! = e
(σ 2
)
− 2σ x 2 (1.6)
and steady state equations represent a satisfactory approxima-
tion. Applying Severinghaus’s equation for the rate of uptake
of nitrous oxide,13 this should occur within 10 to 15 min
Taking natural logarithms of both sides and solving for x,
from the start of induction, at which time the rate of change
we obtain
in the uptake of nitrous oxide has declined to less than 5% of
its current rate of uptake.
σ 1
x= − log ( v!/q! ) (1.7)
2 σ
Basic Equations
1 As the steady state equations governing the exchange of the
Letting X =
σ
( )
log v! q! and substituting in equations 1.3 first gas in a single lung unit,* we take
and 1.4, we obtain V!IFI − V!AF = Q! λ( F − Fv ) (2.1)
2
⎛ σ⎞
1 − ⎝⎜X − 2 ⎠⎟ 2 and
v! = e (1.8)
2π (1 − F )V!A = (1 − FI )V!I (2.2)
Copyright © 2018, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
PERIOPERATIVE MEDICINE
Start of washin 0 V̇
IFIN2O 0 Substituting for V!I in equation 2.1, we may solve this equa-
Washin complete 1 0 1 tion for F.
FA/FI = alveolar/inspired concentration ratio.
where V!I is incoming alveolar ventilation per unit time (the V!AFI + λQ! (1 − FI )Fv
F= (2.6)
“inflow”), V!A is outgoing alveolar ventilation per unit time V!A + λQ! (1 − FI )
(the “outflow”), Q! is blood flow per unit time, FI is concen- Second Gas Effect in Blood Phase. We denote the uptake of
tration of the first gas in the dry fraction of the inflow, Fv is the first gas, nitrous oxide, by the symbol VN! 2O , the uptake
concentration of the first gas in the dry fraction of a sample of the second gas in the absence of the first gas, by V!SG ,
of gas equilibrated with the incoming mixed venous blood, and the uptake of the second gas in the presence of first gas,
F is concentration of the first gas in the dry fraction of the by V!SG / FG . VN! 2O is given by the left-hand side of equation
outflow, which is also in equilibrium with the outgoing arte-
2.1; V!SG is given by the left-hand side of equation 2.3 when
rial blood, and λ is the Ostwald solubility of the first gas in
inspired alveolar ventilation equals V!A ; and V!SG / FG is given
blood at 37°C.
by the left-hand side of equation 2.3 with inspired alveolar
To this system we now add a second soluble gas, denoted
ventilation determined from equation 2.2. Using the super-
by the subscript SG to allow it to be distinguished from the
script hl for a homogenous lung,
first gas. The second gas also obeys equation 2.1:
Copyright © 2018, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ventilation-Perfusion Mismatch and the Second Gas Effect
Copyright © 2018, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
PERIOPERATIVE MEDICINE
Here the second gas is contained in a smaller volume whose Physiological modeling of gas exchange. J Appl Physiol
(1985) 2001; 91:17–25
size depends on the rate of nitrous oxide uptake. The equa-
9. Peyton PJ, Horriat M, Robinson GJ, Pierce R, Thompson BR:
tions describing this phase are therefore more complicated. Magnitude of the second gas effect on arterial sevoflurane
Only the final result is given here. partial pressure. ANESTHESIOLOGY 2008; 108:381–7
10. Peyton PJ, Stuart-Andrews C, Deo K, Strahan F, Robinson
⎛ 1 ⎞ ⎛V!NnhlO ⎞ GJ, Thompson BR, Pierce R: Persisting concentrating and
AR(gas) = 1 + ⎜ 2 second gas effects on oxygenation during N2O anaesthesia.
⎟⎜ ! ⎟
λ(
⎝ SG − λ (1 − FI ) I ⎠ ⎝) V 0 ⎠
(2.21) 11.
Anaesthesia 2006; 61:322–9
Severinghaus JW, Hendrickx JF, Carette R, Lemmens HJ, De
⎛ λ SG I SG − λ (1 − FI ) I ⎞ Wolf AM: Can large volume N2O uptake alone explain the
⎜ 1 − λ SG I SG
⎟ second gas effect? Br J Anaesth 2006; 97:262; author reply
⎝ ⎠ 262–3
12. Magnusson L, Spahn DR: New concepts of atelectasis during
AR(gas) is the augmentation ratio in the gas phase. Note that general anaesthesia. Br J Anaesth 2003; 91:61–72
equations 2.20 and 2.21 may both be simplified to the follow- 13. Severinghaus JW: The rate of uptake of nitrous oxide in man.
J Clin Invest 1954; 33:1183–9
ing approximate form:
14. West JB, Wagner P: Pulmonary gas exchange, Bioengineering
AR = 1 + k(1 − WR ) (2.22) aspects of the lung. Edited by West JB. New York, Marcel
Dekker Inc., Publishers, 1977, pp 378–82
where k is a constant that differs for gas and blood and
15. Colburn WE, Evans JW, West JB: Analysis of effect of the
incorporates the effect of σ. When the augmentation ratio solubility on gas exchange in nonhomogenous lungs. J Appl
equals 1, there is no second gas effect, so when the second Physiol 1974; 37:547–51
gas effect occurs, we can write second gas effect, SGE = 16. Rehder K, Knopp TJ, Sessler AD, Didier EP: Ventilation-
perfusion relationship in young healthy awake and anesthe-
k(1 – WR). This indicates how the second gas effect can tized-paralyzed man. J Appl Physiol Respir Environ Exerc
be both a function of the degree of nitrous oxide uptake Physiol 1979; 47:745–53
(1 – WR) and the degree of ventilation-perfusion mismatch 17. Dueck R, Young I, Clausen J, Wagner PD: Altered distribution
(which appears in the term k along with the solubility of of pulmonary ventilation and blood flow following induction
of inhalation anesthesia. ANESTHESIOLOGY 1980; 52:113–25
the first and second gases and the inspired concentration
18. Lundh R, Hedenstierna G: Ventilation-perfusion relationships
of the first gas). during anaesthesia and abdominal surgery. Acta Anaesthesiol
Perhaps the most useful form of equation 2.21 involves Scand 1983; 27:167–73
the substitution of V!I − V!A for V!Nnhl
2O
. This gives rise to the 19. Lundh R, Johansson H: Alveolar stability during anaes-
thesia for reconstructive vascular surgery in the leg. Acta
following equation: Anaesthesiol Scand 1983; 27:26–34
AR = 1 + k
(V!I − V!A ) (2.23)
20. Lundh R, Hedenstierna G: Ventilation-perfusion relation-
ships during halothane anaesthesia and mechanical ventila-
V!A tion. Effects of varying inspired oxygen concentration. Acta
Anaesthesiol Scand 1984; 28:191–8
References 21. Dantzker D, Wagner P, West JB: Instability of lung units with
low VA/Q ratios during O2 breathing. J Appl Physiol 1975;
1. Epstein RM, Rackow H, Salanitre E, Wolf GL: Influence of the 38:886–95
concentration effect on the uptake of anesthetic mixtures: 22. Aranake A, Mashour GA, Avidan MS: Minimum alveolar con-
The second gas effect. ANESTHESIOLOGY 1964; 25:364–71 centration: Ongoing relevance and clinical utility. Anaesthesia
2. Stoelting RK, Eger EI 2nd: An additional explanation for the 2013; 68:512–22
second gas effect: A concentrating effect. ANESTHESIOLOGY 23. Eger EI 2nd, Tang M, Liao M, Laster MJ, Solt K, Flood P,
1969; 30:273–7 Jenkins A, Raines D, Hendrickx JF, Shafer SL, Yasumasa T,
3. Korman B, Mapleson WW: Concentration and second gas Sonner JM: Inhaled anesthetics do not combine to produce
effects: Can the accepted explanation be improved? Br J synergistic effects regarding minimum alveolar anesthetic
Anaesth 1997; 78:618–25 concentration in rats. Anesth Analg 2008; 107:479–85
4. Taheri S, Eger EI 2nd: A demonstration of the concentration 24. Katoh T, Ikeda K, Bito H: Does nitrous oxide antago-
and second gas effects in humans anesthetized with nitrous nize sevoflurane-induced hypnosis? Br J Anaesth 1997;
oxide and desflurane. Anesth Analg 1999; 89:774–80 79:465–8
5. Hendrickx JF, Carette R, Lemmens HJ, De Wolf AM: Large 25. Kretzschmar M, Kozian A, Baumgardner JE, Borges JB,
volume N2O uptake alone does not explain the second gas Hedenstierna G, Larsson A, Hachenberg T, Schilling T: Effect
effect of N2O on sevoflurane during constant inspired venti- of bronchoconstriction-induced ventilation-perfusion mis-
lation. Br J Anaesth 2006; 96:391–5 match on uptake and elimination of isoflurane and desflu-
6. Peyton PJ, Robinson GJ, Thompson B: Ventilation-perfusion rane. ANESTHESIOLOGY 2017; 127:800–12
inhomogeneity increases gas uptake: Theoretical modeling 26. Spiegel MR: Schaum's Outline of Theory and Problems of
of gas exchange. J Appl Physiol (1985) 2001; 91:3–9 Statistics in SI units, 1st edition. New York, McGraw Hill,
7. Peyton PJ, Robinson GJ, Thompson B: Ventilation-perfusion 1972, p 123
inhomogeneity increases gas uptake in anesthesia: 27. Mathews JH, Fink KD: Numerical Methods using Matlab, 3rd
Computer modeling of gas exchange. J Appl Physiol (1985) edition. New Jersey, Prentice-Hall, 1999, p 354
2001; 91:10–6 28. Evans J, Wagner P, West J: Conditions for reduction of pulmo-
8. Peyton PJ, Robinson GJ, Thompson B: Effect of ventila- nary gas transfer by ventilation-perfusion inequality. J Appl
tion-perfusion inhomogeneity and N(2)O on oxygenation: Physiol 1974; 36:533–7
Copyright © 2018, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
J Appl Physiol 126: 558–568, 2019.
First published December 6, 2018; doi:10.1152/japplphysiol.00689.2018.
RESEARCH ARTICLE
Effect of net gas volume changes on alveolar and arterial gas partial pressures
in the presence of ventilation-perfusion mismatch
Ben Korman,1 Ranjan K. Dash,2 and Philip J. Peyton3
1
Department of Anaesthesia and Pain Medicine, Royal Perth Hospital, Perth, Western Australia, Australia; 2Departments of
Biomedical Engineering and Physiology, Medical College of Wisconsin, Milwaukee, Wisconsin; and 3Anaesthesia,
Perioperative, and Pain Medicine Unit, Melbourne Medical School, University of Melbourne, Department of Anaesthesia,
Austin Health, Heidelberg, Victoria, Australia
Submitted 9 August 2018; accepted in final form 3 December 2018
Korman B, Dash RK, Peyton PJ. Effect of net gas volume pulmonary uptake of any gas with a linear dissociation curve
changes on alveolar and arterial gas partial pressures in the presence of (8). However, clinical investigations of the second gas effect
ventilation-perfusion mismatch. J Appl Physiol 126: 558 –568, 2019. (SGE) performed over the last decade have brought this prop-
First published December 6, 2018; doi:10.1152/japplphysiol.00689.
osition into question (9, 17). The SGE occurs when a soluble
2018.—The second gas effect (SGE) occurs when nitrous oxide enhances
the uptake of volatile anesthetics administered simultaneously. Recent first gas such as nitrous oxide (N2O) is delivered in high
work shows that the SGE is greater in blood than in the gas phase, that inspired concentrations during anesthesia (7, 25). The high
this is due to ventilation-perfusion mismatch, that as mismatch increases, inspired concentrations are associated with the transfer of large
the SGE increases in blood but is diminished in the gas phase, and that volumes of gas, of the order of 0.6 to 1 l/min from lungs to
these effects persist well into the period of nitrous oxide maintenance blood during the early stages of anesthesia (12, 22). Thereafter,
anesthesia. These modifications of the SGE are most pronounced with the the rate of N2O uptake falls progressively toward zero. Alve-
low soluble agents in current use. We investigate further the effect of net olar-capillary uptake of the first gas increases the alveolar
gas volume loss during nitrous oxide uptake on low concentrations of
other gases present using partial pressure-solubility diagrams. The steady-
concentrations of other gases present, accelerating their uptake.
state equations of gas exchange were solved assuming a log-normal These other gases include the commonly used volatile anes-
distribution of ventilation-perfusion ratios using Lebesgue-Stieltjes inte- thetic agents and are referred to individually and collectively as
gration. It was shown that under these conditions the classical partial second gases.
pressure-solubility diagram must be modified, that for currently used Until 2000, the most widely accepted explanation of the
volatile anesthetic agents the alveolar-arterial partial pressure difference SGE appealed to the contraction in volume associated with
is less than that predicted in the past, and that the alveolar-arterial partial N2O uptake as the factor responsible (12). The SGE was
pressure difference may even be reversed during uptake in the case of measured in expired gas and generally assumed to be of similar
highly insoluble gases such as sulfur hexafluoride. Comparing this with
the situation described previously for nitrogen in steady-state air breath-
magnitude in blood. More recently, it has been shown that the
ing, we show that for nitrogen, the direction of the alveolar-arterial effect is not only greater in blood but also continues to be
gradient is opposite to the direction of net gas volume movement. significant when N2O uptake falls to relatively low steady-state
Although gas uptake with ventilation-perfusion inequality exceeding that levels (9, 21). Based on computer simulations assuming a
when matching is optimal is shown to be possible, it is less likely than log-normal distribution of ventilation and blood flow, it was
alveolar-arterial partial pressure reversal. suggested that this newly recognized phenomenon is due to the
NEW & NOTEWORTHY Net uptake of gases administered with mismatch between ventilation and blood flow (V̇/Q̇ mismatch)
nitrous oxide may proceed against an alveolar-arterial partial pressure that occurs routinely following the induction of anesthesia
gradient. The alveolar-arterial gradient for nitrogen in the steady-state (18 –20).
breathing air depends not only on the existence of a distribution of In an earlier study, West et al. (26) obtained partial pressure-
ventilation-perfusion ratios in the lung but also on the presence of a solubility diagrams in the case of a single soluble inert gas
net change in gas volume and is opposite in direction to the direction present in the inspired gas mixture together with an insoluble
of net gas volume uptake.
gas that acts solely as a vehicle for delivering the soluble inert
alveolar-arterial partial pressure gradient; anesthetic uptake; mathe- gas. Their study focused on the elimination from the lung of
matical modeling; second gas effect; ventilation-perfusion mismatch a gas supplied at constant partial pressure in mixed venous
blood. Colburn et al. (3) explored further the exchange of inert
gases in such a system. They concluded that mismatching of
INTRODUCTION ventilation and blood flow must reduce the pulmonary uptake
of any gas with a linear dissociation curve, a conclusion stated
A long-held tenet of respiratory physiology is the belief that explicitly by Evans et al. (8), in a separate publication. How-
mismatching of ventilation and blood flow must reduce the ever, these authors did not consider possible effects associated
with net gas volume changes during gas exchange. In this
Address for reprint requests and other correspondence: B. Korman, Dept. of
paper, we use N2O uptake as a tool to investigate the effect of
Anaesthesia and Pain Medicine, Royal Perth Hospital, Perth, WA 6000, net gas volume changes on the partial pressure-solubility dia-
Australia (e-mail: ben@korman.com.au). gram and its effect on gas uptake in nonhomogenous lungs.
558 8750-7587/19 Copyright © 2019 the American Physiological Society http://www.jappl.org
Downloaded from www.physiology.org/journal/jappl at Medical Col of Wisconsin (141.106.122.102) on March 8, 2019.
VENTILATION-PERFUSION MISMATCH AND THE SECOND GAS EFFECT 559
METHODS ⌬V̇/V̇A. Note that (V̇I ⫺ V̇A)/V̇A is equal to (V̇I/V̇A) ⫺ 1 and is
therefore also the relative volume change associated with N2O uptake.
Consider gases with linear dissociation curves, commonly referred For convenience, we set Fv in Eq. 6 equal to 0. Alveolar ventilation
to as inert gases, being exchanged under steady-state conditions. If V̇A is set at 4 l/min, and pulmonary blood flow Q̇ is set at 5 l/min.
such a gas is delivered in the inspired gas mixture at a rate of V̇IFI and Partial pressure-solubility relationship for soluble second gas.
removed at a rate of V̇AFA in a homogenous lung of given V̇A/Q̇, then Consider now a second soluble gas present in such low concentrations
the uptake of the gas is V̇IFI ⫺ V̇AFA. This must equal the uptake by that it has no impact on Eq. 2. This will also obey Eq. 1 in each gas
blood, which is given by Q̇(Fa ⫺ Fv ), where Fa and Fv are equal to exchanging unit. Assuming a mixed venous partial pressure of zero
Pa ⁄ 共PB ⫺ PH2O兲 and Pv / 共PB ⫺ PH2O兲 respectively, so that for this agent, expressions may be obtained for the mixed alveolar and
mixed pulmonary end-capillary partial pressures as follows:
V̇IPI ⫺ V̇APA ⫽ Q̇(Pa ⫺ PV ). (1)
PA N 共V̇ ⁄ Q̇ j 兲
For the diluent gas, which is treated as if it were completely ⫽ 兺
Ij
共V̇ ⁄ V̇A 兲 (7)
共V̇ 兲
Aj
insoluble, we have PI j⫽1 ⁄ Q̇ j ⫹
Aj
共V̇A ⁄ Q̇兲PI ⫹ 共1 ⫺ FI兲PV Here, is the solubility of the second gas, the subscript j refers to
Pa ⫽ PA ⫽ . (3) the jth compartment, and the symbol Q̇t refers to the total pulmonary
共V̇A ⁄ Q̇兲 ⫹ 共1 ⫺ FI兲 blood flow. If a log-normal distribution of ventilation-perfusion ratios
Given that V̇I/Q̇ ⫽ (V̇I/V̇A) ⫻ (V̇A/Q̇), Eq. 2 provides the following is assumed, Eqs. 7 and 8 may also be converted to their equivalent
expression for V̇I/Q̇: Lebesgue-Stieltjes forms and then solved for PA /PI and Pa /PI respec-
tively. Graphs may now be drawn for each of these parameters as a
共V̇A ⁄ Q̇兲 ⫹ 共1 ⫺ FV 兲 function of for different values of and ⌬V̇.
共V̇I ⁄ Q̇兲 ⫽ 共V̇A ⁄ Q̇兲 . (4) The augmentation-solubility relationship. A useful measure of the
共V̇A ⁄ Q̇兲 ⫹ 共1 ⫺ FI兲 magnitude of the SGE is the augmentation ratio (AR), which is
defined as the partial pressure of the second gas in the relevant phase
Inspired alveolar ventilation with a log-normal distribution. In a (gas or blood) in the presence of N2O divided by the partial pressure
lung model consisting of N parallel compartments, the inspired the second gas would have in that phase in the absence of N2O, with
ventilation is the sum of the individual values for each compartment: all other factors being held constant (7, 11). When AR ⫽ 1, there is no
N SGE, so when the SGE does occur, its magnitude is given by the
V̇I ⫽ 兺
j⫽1
V̇I .
j
(5) amount by which AR exceeds unity. We have previously shown that
when the distribution of ventilation-perfusion ratios in the lung is log
When dealing with a log-normal distribution of expired alveolar normal, the following relationship applies:
ventilation-perfusion ratios, Colburn et al. (3) have shown that if V̇
is the log standard deviation of ventilation and Q̇ is the log standard ⌬V̇
AR ⫽ 1 ⫹ k . (9)
deviation of blood flow, then ⫽ |V̇ ⫺ Q̇| specifies the degree of
V̇A
mismatch between ventilation and blood flow. In awake healthy
adults, varies from 0.25 to 1.0; worsening of V̇/Q̇ matching is seen Here, k is a variable that depends primarily on the solubility of the
in disease states and is also known to occur soon after induction of second gas in blood () and the degree of mismatch between venti-
anesthesia, with typical values for of 0.75–1.5 in healthy adults (4, lation and blood flow () (10, 11). Equation 9 explains how it is
13–15, 24). With a log-normal distribution of V̇/Q̇ ratios, Eq. 5 may possible for the degree of ventilation-perfusion mismatch to influence
then be expressed as the equivalent Lebesgue-Stieljes integral and the magnitude of the SGE provided that ⌬V̇/V̇A, the relative volume
becomes change associated with N2O uptake, remains greater than zero. Aug-
再
mentation ratios may be obtained for the gas phase and for blood by
V̇A
兰 dividing PA /PI and Pa /PI at any given value of ⌬V̇ with their equiv-
⬁
V̇I ⫽ exp(⫺x2 ⁄ 2)
⫺⬁ alent values when ⌬V̇ ⫽ 0. In each case, all other variables must be
兹 2
冎
held constant. AR(gas) and AR(blood) may then be graphed as
共FI ⫺ FV 兲 functions of .
⫹ dx. (6)
共V̇A ⁄ Q̇兲exp(x ⫺ )2 ⁄ 2 ⫹ 共1 ⫺ FI兲exp(x2 ⁄ 2) Alveolar-arterial partial pressure differences. An estimate of the
A-a difference may be obtained by multiplying the difference
The left-hand term in the integrand represents the log-normal {PA /PI ⫺ Pa /PI} by the partial pressure of the second gas in the dry
distribution of expired ventilation-perfusion ratios. When the first gas inspired gas mixture. For example, when 1% sevoflurane is present in
is nitrous oxide, ⫽ 0.47 and the right-hand term gives the uptake of the dry inspired mixture, it is necessary to multiply the difference by
nitrous oxide. Once V̇A, Q̇, , and Fv are specified, Eq. 6 provides a 7.13 to obtain an estimate of the A-a difference in mmHg.
one-to-one relationship between V̇I and FI and hence, allows the The uptake-solubility relationship. As an index of gas exchange,
calculation of the inspired concentration of nitrous oxide associated Colburn et al. (3) studied R(), the ratio of the gas exchange rate of
with the desired total inspired alveolar ventilation. Letting ⌬V̇ ⫽ a compartmental model over the exchange rate of the homogenous
(V̇I ⫺ V̇A), we may now determine the inspired concentration of N2O model with the same total expired alveolar ventilation and blood flow.
necessary to produce the required values of ⌬V̇ or ⌬V̇/V̇A, either of In their model, R() is always ⬍1, indicating that gas exchange in a
which may be chosen as an independent variable for investigation, nonhomogenous lung is always less than in the equivalent homoge-
along with , the degree of mismatch between ventilation and blood nous lung. We use their expression for R() and apply it to an inert gas
flow. We use the terms “volume shrinkage,” “volume contraction,” present in the inspired gas mixture in low concentrations in the
and “net gas volume uptake” interchangeably when referring to presence of net volume changes generated by N2O uptake.
All programs were written in MATLAB and solved using the fsolve noted by Stoelting and Eger (25). Until the work of Peyton et
nonlinear algebraic equations solver. The integral function was used al. (18 –20), no particular effect was ascribed to the addition of
to integrate improper integrals such as Eq. 6. The equations used to V̇/Q̇ mismatch. If the effects were merely additive, they would
generate the figures in the RESULTS and DISCUSSION sections are dis-
cussed further in APPENDIX 1 and APPENDIX 2. For individual figures,
produce the result shown in Fig. 1C.
values of ⌬V̇ were chosen to demonstrate differences when these Partial pressure-solubility diagrams in the presence of net
occur, while also lying within the range believed to occur clinically gas-volume uptake. Figure 2 shows what actually happens in
during N2O anesthesia. our model when a net gas volume reduction is combined with
V̇/Q̇ mismatch. It shows the effect of increasing ⌬V̇ from 0 to
RESULTS
600 ml/min on the alveolar and arterial partial pressures in a
We begin by drawing a partial pressure-solubility diagram of homogenous lung (i.e., ⫽ 0) and a lung with significant V̇/Q̇
the form described by West et al. (26). This is achieved in our mismatch ( ⫽ 1.5).
model by setting the inspired nitrous oxide concentration to With no volume change (⌬V̇ ⫽ 0), the graph of PA /PI is
zero, so that with the very low inspired concentrations of the displaced to the right by V̇/Q̇ scatter, as in Fig. 1A. A similar
other soluble gas present, there is effectively no change in net displacement occurs when ⌬V̇ ⫽ 600. Now it is the increase in
gas volume. The resulting curves are shown in Fig. 1A. PA /PI of the continuous blue line relative to the continuous
When there is no V̇/Q̇ mismatch, i.e., ⫽ 0, alveolar and black line that reflects the impaired uptake in the presence of an
arterial partial pressures are equal. This is indicated by the increased degree of mismatch. For arterial blood, when ⌬V̇ ⫽
black dash-dotted line in Fig. 1A. Moving along this line 0, the line for Pa /PI is displaced to the left by V̇/Q̇ scatter, as in
toward the left end of the x-axis, we are dealing with less Fig. 1A. With ⌬V̇ ⫽ 600, the graph of Pa /PI with ⫽ 1.5
soluble gases; very little is transferred to blood so that the (continuous red line) greatly exceeds that for ⫽ 0 (continu-
alveolar (and hence arterial) partial pressure approaches the ous black line) for much of its course, especially at low
partial pressure in the inspired gas mixture, and both PA /PI and solubilities ( ⬍ 0.05). This reflects the increased uptake that
Pa /PI approach 1. As we move along the black dash-dotted line occurs with significant gas volume shrinkage. The effect of the
in Fig. 1A toward the right end of the x-axis, we are dealing contraction in net gas volume on PA /PI and Pa /PI at each value
with more and more soluble gases; more and more of the of is obtained by comparing the dash-dotted line with the
alveolar gas is transferred to blood so that the alveolar (and solid line for each color. The greater magnitude of the SGE in
hence, the arterial) partial pressure approaches zero, and so do blood than in the gas phase is indicated by the larger area
the respective ratios PA /PI and Pa /PI. between the two red lines in Fig. 2B compared with the area
The presence of V̇/Q̇ mismatch acts as an impediment to gas between the two blue lines in Fig. 2A.
exchange; less gas is removed from the alveolar gas mixture,
Augmentation-solubility diagrams in the presence of net gas
producing a higher partial pressure in the gas phase and a lower
volume uptake. The augmentation or fractional increase in
partial pressure in blood. Thus, with ⫽ 1.5, the lines in Fig.
partial pressure of a gas due to gas volume shrinkage may be
1A for alveolar gas (blue dash-dotted line) and arterial blood
(red dash-dotted line) separate, producing an alveolar-arterial derived in Fig. 2 by dividing the value of PA /PI or Pa /PI when
partial pressure difference. At each end of the solubility scale, ⌬V̇ ⫽ 600 by its value when ⌬V̇ ⫽ 0 for the value of
PA /PI and Pa /PI behave as before and approach 1 and 0, associated with that gas. In Fig. 3A, we have shown the
respectively. This is the classical description given by West et augmentation ratio with ⫽ 1.5 for gas (blue) and blood (red)
al. (26). when ⌬V̇ ⫽ 200 ml/min (dash-dotted lines) and 800 ml/min
Figure 1B depicts partial pressure-solubility diagrams in the (solid lines); augmentation, (and hence, the SGE) is greater at
absence of V̇/Q̇ mismatch, i.e., with ⫽ 0. The black dash- the higher value of ⌬V̇, an effect most noticeable as the
dotted line in Fig. 1B is the same as in Fig. 1A and represents solubility of the gas is decreased. In Fig. 3B, the augmentation
the situation where there is no net gas volume change (⌬V̇ ⫽ ratio is shown for gas and blood for a range of values with
0); the solid black line represents the situation with a net gas ⌬V̇ set at 100 ml/min. In the absence of V̇/Q̇ mismatch, the
volume loss of 600 ml/min (⌬V̇ ⫽ 600). The difference augmentation ratio is 1.025 (horizontal black line). Increasing
between the two lines reflects the magnitude of the second gas the degree of mismatch has opposite effects on the gas phase
effect. The effect is greater with gases of low solubility, a fact (blue lines) and blood (red lines). As increases, the SGE
Fig. 1. Alveolar and arterial partial pressures as a fraction of the inspired partial pressure vs. solubility expressed as log10(). A: PA /PI (blue dash-dotted line)
and Pa /PI (red dash-dotted line) with no net volume change i.e., ⌬V̇ ⫽ 0 and ⫽ 1.5 The black dash-dotted line represents the situation with ⫽ 0, in which
case PA /PI and Pa /PI are equal. B: no ventilation-perfusion mismatch, i.e., ⫽ 0; PA /PI and Pa /PI are equal, lower curve ⌬V̇ ⫽ 0 (black dash-dotted line) and
upper curve ⌬V̇ ⫽ 600 ml/min (black solid line). C: curves in A and B are summed for gas (blue) and blood (red).
Fig. 2. Alveolar and arterial partial pressures as a fraction of the inspired partial pressure vs. solubility expressed as log10(). A: PA /PI with ⫽ 0, ⌬V̇ ⫽ 0 (black
dash-dotted line); ⫽ 0, ⌬V̇ ⫽ 600 ml/min (solid black line); ⫽ 1.5, ⌬V̇ ⫽ 0 (blue dash-dotted line); ⫽ 1.5, ⌬V̇ ⫽ 600 ml/min (solid blue line). B: Pa /PI
with ⫽ 0, ⌬V̇ ⫽ 0 (black dash-dotted line); ⫽ 0, ⌬V̇ ⫽ 600 ml/min (solid black line); ⫽ 1.5, ⌬V̇ ⫽ 0 (red dash-dotted line); ⫽ 1.5, ⌬V̇ ⫽ 600 ml/min
(solid red line). Dotted green vertical line on the left indicates the position of sulfur hexafluoride, dotted green vertical line on the right indicates the position
of methoxyflurane.
increases in blood but decreases in the gas phase (black The effect of increasing ⌬V̇ on the A-a partial pressure
arrows). This is in line with our previous findings (11). difference is shown in Fig. 5 for several gases in order of
Alveolar-arterial pressure differences. These may be in- decreasing solubility: diethyl ether, sevoflurane, ethylene, and
ferred from the partial pressure-solubility diagram. For exam- sulfur hexafluoride ( ⫽ 12, 0.59, 0.14, and 0.0076, respec-
ple, in Fig. 4, A–D, PA /PI and Pa /PI have been drawn with tively). It is assumed that the inspired concentration of each gas
⫽ 0 (black dash-dotted line) and ⫽ 1.5 (solid lines; is 1%, which represents an inspired partial pressure of 7.13
blue ⫽ gas, red ⫽ blood) for increasing values of ⌬V̇ (from 0 mmHg. The degree of V̇/Q̇ mismatch, , is shown on the
to 800 ml/min). x-axis. Each curve begins at ⫽ 0, where there is no V̇/Q̇
When ⫽ 0, there is no mismatching between ventilation mismatch and the A-a partial pressure difference is zero.
and blood flow; PA /PI and Pa /PI are equal. This is represented Alveolar and arterial partial pressures become equal again
by the single black dash-dotted lines in Fig. 4, A–D regardless when a curve crosses the x-axis, i.e., y ⫽ 0. Parts of a curve
of the value of ⌬V̇. Notice that in each case, the blue line lying below the x-axis indicate a reversal of the A-a partial
representing PA /PI always lies above or coincides with the pressure gradient; i.e., the partial pressure in arterial blood is
black line. However, whereas the red line representing Pa /PI greater than the partial pressure in mixed alveolar gas.
lies below the blue line for PA /PI on the right of the solubility The red arrow in each diagram shows the direction of the
axis, it crosses the blue line as one moves left along the increase in ⌬V̇, which is varied from 0 to 1 l/min in increments
solubility axis. This represents the situation where the partial of 200 ml/min. Ether exhibits a worsening in A-a gradient as
pressure in mixed pulmonary end-capillary blood exceeds that ⌬V̇ is increased, but this is almost imperceptible. In the case of
in mixed alveolar gas. The point of intersection is affected by sevoflurane and ethylene, the A-a partial pressure gradient is
the shrinkage factor and moves to the right as ⌬V̇ is increased. reduced and moves toward zero. With sulfur hexafluoride, the
From this, we can conclude that the solubility at the point of A-a partial pressure difference is less than zero when ⌬V̇ is 400
intersection depends on the magnitude of the net gas volume ml/min or greater, with the partial pressure in pulmonary
uptake. end-capillary blood predicted to be 1 mmHg higher than that in
Fig. 3. A: augmentation-solubility diagram with ⫽ 1.5. Gas phase, ⌬V̇ ⫽ 200 ml/min (blue dash-dotted line); ⌬V̇ ⫽ 800 ml/min (solid blue line). Blood,
⌬V̇ ⫽ 200 ml/min (red dash-dotted line); ⌬V̇ ⫽ 800 ml/min (solid red line). B: effect of increasing with ⌬V̇ constant and set at 100 ml/min. Arrows indicate
the direction of increase of ventilation-perfusion mismatch ( ⫽ 0, 0.5, 1.0, 1.5, and 2.0).
Fig. 4. Alveolar (blue) and arterial (red) partial pressures as a fraction of the inspired partial pressure (on the y-axis) vs. log10() (on the x-axis) for a range of
values with ⫽ 1.5, A: ⌬V̇ ⫽ 0, B: ⌬V̇ ⫽ 200, C: ⌬V̇ ⫽ 400, D: ⌬V̇ ⫽ 800 ml/min. Black dash-dotted line in each diagram is associated with no mismatch
of ventilation and perfusion, i.e., ⫽ 0.
mixed alveolar gas, when ⫽ 1.2 and ⌬V̇ ⫽ 600 ml/min. It shifted to the right; each curve begins above the R() ⫽ 1 line
can be seen that for any specified value of ⌬V̇, the occurrence and ends below it. At the left end of each curve in Fig. 6B, R()
of A-a partial pressure reversal depends on the solubility of the increases as increases, whereas at the right end, R() de-
agent and the degree of V̇/Q̇ mismatch. creases as increases. At the left end of each curve in Fig. 6C,
Uptake-solubility diagrams. Many of the above results are R() increases as ⌬V̇ increases, whereas at the right end, R()
consistent with gas uptake in the presence of ventilation- decreases as ⌬V̇ increases.
perfusion mismatch exceeding that in the absence of ventila- Where R() is greater than 1, uptake of a low inspired
tion-perfusion mismatch, a possibility that is the subject of our concentration of an inert gas is predicted to be greater in a
next investigation. In Fig. 6, we compare the situation that nonhomogenous lung than in the equivalent homogenous lung.
applies when V̇/Q̇ mismatch occurs in the absence of net gas However, only sulfur hexafluoride appears to exhibit this
volume changes with two scenarios in the presence of such property (see blue arrow in Fig. 6B), and only then it is under
changes. extreme conditions, which may not be attainable clinically
Colburn et al. (3) define R() as the ratio of the gas exchange (⌬V̇ ⫽ 1,000 ml/min; ⫽ 4).
rate of a lung with V̇/Q̇ mismatch over the exchange rate of the
equivalent homogenous lung and describe the shape of the DISCUSSION
graph of R() against log10(). Figure 6A shows a series of
such graphs for values of ranging from 0 to 4 in increments Historically, the properties of the SGE have first been
of 0.5. With very low inspired concentrations of inert gas there predicted from theory and then demonstrated experimentally.
is no net volume change during gas exchange, and the family Thus, the existence of the SGE was first predicted from Eger’s
of curves is symmetric about the vertical line x ⫽ log10(V̇A/Q̇). mathematical model of anesthetic uptake (5, 6). Similarly, the
As is increased, R() falls as predicted by Colburn et al. (3) disproportionate effect of V̇/Q̇ mismatch on the partial pressure
and is never greater than 1. of gases in blood was first predicted from theory (18 –20). Both
This is the classical result predicted by Colburn et al. (3) propositions were then proven experimentally (7, 9, 17, 22). As
with which we now compare our two scenarios. Figure 6B has been pointed out by Kennedy (10), the theoretical approach
shows the effect of increasing the value of from zero, with has many advantages. With respect to the present study, a key
the net volume uptake fixed at 1,000 ml/min [body temperature advantage is the ability to control V̇/Q̇ mismatch as an inde-
and pressure dry (BTPD)], whereas Fig. 6C shows the effect of pendent variable, something that is not currently possible in
increasing the value of ⌬V̇ from zero with the degree of experimental situations. Moreover, this is the first time that a
mismatch fixed at ⫽ 2. In each case, once ⌬V̇ is greater than formal mathematical analysis of the type referred to by West
zero, the curves are no longer symmetric about the vertical line and Wagner (27) has been attempted for the second gas effect.
x ⫽ log10(V̇A/Q̇). The minimum for each curve has been In addition, we have used the powerful technique of Lebesgue-
Fig. 5. Graphs of the A-a partial pressure difference (mmHg) for diethyl ether (A), sevoflurane (B), ethylene (C), and sulfur hexafluoride (D) as a function of
(the degree of V̇/Q̇ mismatch) for ⌬V̇ ⫽ 0, 200, 400, 600, 800, and 1,000 ml/min. In each case, the direction of increase in ⌬V̇ is shown by the red arrow.
Stieltjes integration described by Colburn et al. (3) to perform therefore, it must always be positive. However, this rule clearly
our calculations. Why is this important? The alternative is to cannot apply at the lower end of the solubility scale in Fig. 4,
use a compartmental approach. But how many compartments B–D. Our findings are shown in Fig. 5. With V̇A ⫽ V̇I, the
should be used: 10, 20, 50, 100? The more compartments, the A-a partial pressure difference always increases as the degree
greater the accuracy, but this requires more numerical calcu- of V̇/Q̇ mismatch increases, but less so as the solubility of the
lations and hence, more computer time. By transforming the gas is decreased. As the difference between V̇A and V̇I is
problem into one with an algebraic solution, it can be dealt increased with the less soluble gases shown in Fig. 5 B–D, the
with expeditiously using the inbuilt functions of a modern A-a difference decreases. In the case of sulfur hexafluoride, the
problem-solving program like MATLAB. least soluble of the gases, the pressure gradient is reversed,
Our study provides several insights into the SGE. Perhaps with the partial pressure in blood predicted to be higher than
the most interesting finding is the effect on the predicted A-a that in the gas phase, although the movement of the sulfur
partial pressure difference. According to West et al. (26), hexafluoride is still from gas phase to blood. One would
during gas uptake, the alveolar component of the A-a differ- normally expect the partial pressure of a gas to be lower at all
ence is equal to over V̇A/Q̇ times the arterial component, and points downstream during gas uptake, for example, as is
Fig. 6. Graph of R() vs log10() for a range of values. A: in the absence of net volume change. Red arrows indicate the direction of increase of from 0 to
4 in increments of 0.5. B: with a net volume change of 1,000 ml/min [body temperature and pressure dry (BTPD)], so that ⌬V̇/V̇A ⫽ 0.25. Red arrows are same
as in A. Blue arrow indicates the point where the uptake ratio of sulfur hexafluoride exceeds 1. C: with the degree of V̇/Q̇ mismatch fixed at ⫽ 2, the net volume
change has been increased in the direction of the red arrows from 0 to 1,000 ml/min (BTPD) in increments of 200 ml/min. In each diagram, the solid vertical
green line is drawn at x ⫽ log10(0.8). The position of sulfur hexafluoride is shown by the dotted green vertical line on the left, and that of methoxyflurane is
shown by the dotted green vertical line on the right.
commonly described in the transfer of oxygen from inspired air “unsafe” FIO2. In the region above the dark blue surface and
to the mitochondrion, the so-called “oxygen cascade” (1). below the yellow FIN O ⫽ 0.7 plane (green arrow), it is possible
As is apparent from the comments in RESULTS in relation to 2
for nitrous oxide uptake to generate a reversal of the alveolar-
Figs. 4 and 5, whether this phenomenon occurs in a clinical arterial partial pressure gradient for the second gas.
setting will depend on the value of , , and ⌬V̇/V̇A, the factors
During anesthesia, generally lies between 0.75 and 1.5.
controlling augmentation in Eq. 9. Although an increase in the
With ⫽ 0.75, the highest solubility for which A-a reversal
degree of V̇/Q̇ mismatch favors augmentation through its effect
can occur is ⬃0.35 so that none of the volatile agents in
on the term k in Eq. 9, it impedes nitrous oxide uptake, and it
common use can be expected to demonstrate A-a partial
can therefore reduce augmentation through its effect on the
pressure reversal during uptake. However at least six other
“shrinkage” term ⌬V̇/V̇A. Increasing the inspired nitrous oxide
gases may demonstrate the effect: xenon ( ⫽ 0.2), ethylene
concentration can compensate for this but is limited by the
( ⫽ 0.14), ethane ( ⫽ 0.092), methane ( ⫽ 0.038), nitrogen
maximum concentration compatible with safe anesthesia.
( ⫽ 0.014), and sulfur hexafluoride ( ⫽ 0.0076). With
These competing effects are shown in Fig. 7. Each point on the
⫽ 1.5, the highest solubility for which A-a reversal can occur
blue colored curved surface in Fig. 7 represents a point at
which the alveolar and arterial partial pressures are equal. This is ⬃0.15 so that only ethane, methane, nitrogen, and sulfur
surface is shown as a function of the gas solubility (x-axis) and hexafluoride can be expected to demonstrate the phenomenon.
the degree of V̇/Q̇ mismatch (y-axis). Shown on the z-axis is So far, we have only considered a second gas with a mixed
venous partial pressure of zero undergoing net uptake in the
the FIN O corresponding to each point on the surface. The
2 presence of volume “shrinkage” produced by nitrous oxide
surface slopes upward and backward from the bottom right to uptake. A slightly different scenario was investigated by Can-
the top left of Fig. 7. A yellow horizontal plane is drawn for field and Rahn (2), who studied alveolar-arterial nitrogen
FIN O ⫽ 0.7, the highest inspired concentration that can be used partial pressure differences in anesthetized dogs being venti-
2
safely in most healthy subjects without producing hypoxia lated with air. Here, the difference between inspired and
(16). In the region beneath the blue surface (red arrow), the expired alveolar ventilation is smaller than during nitrous oxide
inspired nitrous oxide concentration is too low to produce the anesthesia. It is produced by the discrepancy between oxygen
required volume change. The region above the yellow plane on uptake and carbon dioxide output, but there is no net exchange
which FIN O ⫽ 0.7 (Fig. 7, black arrow) is associated with an of nitrogen. They showed that the mean arterial pressure of
2
Fig. 7. Three-dimensional surface showing all points for which the alveolar and arterial partial pressures of the 2nd gas are equal, shown as a function of the
inspired nitrous oxide concentration necessary to produce the required volume change for the specified values of and . The surface is colored blue and slopes
upward and backward from the bottom right (dark blue) to the top left (light blue). On the yellow horizontal plane, FIN O ⫽ 0.7. This plane intersects the blue
2
surface as shown. Red arrow points to the region below the blue surface in which the FIN O is too low to produce the required volume change. Green arrow points
2
to the region between the dark blue curved surface and the yellow horizontal plane compatible with the phenomenon of A-a partial pressure reversal. Black arrow
points to the region above the yellow plane associated with an “unsafe” FIO . When ⫽ 0, PA ⫽ Pa for all values of FIN O, so the graph has been offset slightly
2 2
from ⫽ 0 to facilitate discussion.
Fig. 8. A: alveolar-arterial partial pressure difference for nitrogen as a function of , the degree of ventilation-perfusion mismatch, for ⌬V̇ ⫽ 0, 25, 50, 75, and
100 ml/min. The direction of increase in ⌬V̇ is shown by the red arrow. The dotted horizontal green line indicates an arterial-alveolar difference of 16 mmHg.
B: A-a partial pressure difference for nitrogen as a function of , the degree of ventilation-perfusion mismatch, for ⌬V̇ ⫽ 0, ⫺25, ⫺50, ⫺75, and ⫺100 ml/min.
The direction of decrease in ⌬V̇ is shown by the blue arrow.
nitrogen was greater than that of mean alveolar gas and Uptake of the second gas may occur against an overall alveo-
ascribed it to the variation in V̇/Q̇ ratios throughout the lung. lar-arterial partial pressure gradient if the degree of V̇/Q̇
The situation may be mimicked in our model by setting Pv mismatch and the rate of nitrous oxide uptake are sufficiently
equal to Pa for the inert gas involved (see APPENDIX 2). The large and can occur with highly insoluble gases such as sulfur
result is shown in Fig. 8A for values of ⌬V̇ ⫽ 0, 25, 50, 75, and hexafluoride but is less likely with agents such as desflurane
100 ml/min. and sevoflurane in common use today. We have also shown
This may be compared with the graphs in Fig. 5. The red that it is possible for V̇/Q̇ mismatch to increase second gas
arrow shows the direction of increase in ⌬V̇. This is associated uptake compared with the equivalent homogenous lung.
with an increase in the size of the a-A partial pressure gradient.
Canfield and Rahn (2) obtained an average gradient of 16 APPENDIX 1
mmHg, which is indicated by the horizontal green dotted line in
Fig. 8A. The curve for ⌬V̇ ⫽ 50 ml/min intersects this line with Derivation of Lebesgue-Stieltjes integrals for second gas partial
approximately equal to 1.8. This is close to the upper limit of pressure-solubility diagrams. In a lung model consisting of N parallel
values known to occur in humans during anesthesia. The compartments, each having a different ventilation/perfusion ratio, the
partial pressure of the mixed expired alveolar gas PA is the ventila-
authors’ statement, “Whenever differences in ventilation-per-
tion-weighted mean of the individual compartmental partial pressures:
fusion ratio exist among alveoli of the lung . . . the mean
N
arterial inert gas pressure is greater than that of the mean
alveolar gas” is borne out in Fig. 8A with one important 兺
j⫽1
V̇A PA j j
PA ⫽ . (10)
proviso, namely that the ventilation-perfusion differences must N
partial pressure-solubility diagram of West et al. (26) must be Here, the symbols Q̇j and V̇j refer to the blood flow and expired
modified. At the lower end of the solubility range, the partial alveolar ventilation in the jth compartment, and FI to the solubility
pressure-solubility curve of the second gas in arterial blood is and inspired concentration of nitrous oxide, and SG to the solubility
displaced upward, resulting in a disproportionate increase in of the second gas in blood. Setting = ⫽ (1 ⫺ FI), and expressing Eq.
the second gas effect in blood compared with alveolar gas. 12 using partial fractions,
再 冉 冊 冎
FISG ⬘ ⫺ PISG j⫽1 PISG j⫽1
SG
PA 共V̇A ⁄ Q̇t兲 V̇j ⁄ Q̇j ⫹ V̇j ⁄ Q̇j SG共P V ⁄ PI兲SG
⫽ 兵共 ⫺ SG兲I1 ⫺ 共 ⫺ ⬘兲I2其 , (14) ⫹ .
PISG 共⬘ ⫺ SG兲兹2 SG ⫹ V̇j ⁄ Q̇j 共1 ⫺ FI兲 ⫹ V̇j ⁄ Q̇j SG ⫹ V̇j ⁄ Q̇j
PaSG 共V̇A ⁄ Q̇t兲 (24)
⫽⫺ 兵共 ⫺ SG兲I3 ⫺ 共 ⫺ ⬘兲I4其 , (15) Defining I5 as follows:
PISG 共⬘ ⫺ SG兲兹2
⬁
兰
⬁ dx
兰
dx I5 ⫽ , (25)
I1 ⫽
⫺⬁ SGe
共x 2⫹2x⫺2 ⁄ 2
共
兲 ⫹ V̇A ⁄ Q̇t ex2⁄ 2 兲
, (16)
⫺⬁ SG e 共 x ⫺ 兲 ⁄ 2 ⫹
2
共V̇A ⁄ Q̇t兲e共x ⫺4x⫹2 兲 ⁄ 2
2 2
⬁ P A SG 共V̇A ⁄ Q̇t兲
⫽ 兵共 ⫺ SG兲I1 ⫺ 共 ⫺ ⬘兲I2其
兰
dx
I3 ⫽ , (18) PISG 共⬘ ⫺ SG兲兹2
⫺⬁ SGe
x2⁄ 2
共
⫹ V̇A ⁄ Q̇t e 兲 共 x ⫺ 兲 2 ⁄ 2
SG共P V ⁄ PI兲SG
and ⫹ I3 . (26)
⬁
兹2
兰 ⬘e
dx
I4 ⫽ . (19) and
⫺⬁
x2⁄ 2
共
⫹ V̇A ⁄ Q̇t e共x ⫺ 兲兲 2⁄ 2
兰 e
dx
Iv ⫽ AUTHOR CONTRIBUTIONS
x2⁄ 2 2⁄ 2
⫺⬁ ⫹ (V̇A ⁄ Q̇t)e(x ⫺ ) B.K. conception and design of research; B.K. performed experiments; B.K.,
R.K.D., and P.J.P. analyzed data; B.K., R.K.D., and P.J.P. interpreted results
FV N O ⫽ ⫺
兹2⌬V̇ , (33)
of experiments; B.K. prepared figures; B.K. drafted manuscript; B.K., R.K.D.,
and P.J.P. edited and revised manuscript; B.K., R.K.D., and P.J.P. approved
2
V̇AIv final version of manuscript.
SG
PA (V̇A ⁄ Q̇t)
⫽ 兵共* ⫺ 兲H1 ⫺ 共* ⫺ SG兲H2其 REFERENCES
PISG 共SG ⫺ 兲兹2 1. Biro GP. From the atmosphere to the mitochondrion: the oxygen cascade.
SG共P V
⁄ PI兲SG In: Hemoglobin-Based Oxygen Carriers as Red Cell Substitutes and
⫹ H3 , (34) Oxygen Therapeutics (edited by Kim H and Greenburg A). Berlin, Heidel-
兹2 berg: Springer, 2013, p. 27–53.
2. Canfield RE, Rahn H. Arterial-alveolar N2 gas pressure differences due
PaSG (V̇A ⁄ Q̇t)
⫽ 兵共* ⫺ 兲H4 ⫺ 共* ⫺ SG兲H5其 to ventilation-perfusion variations. J Appl Physiol 10: 165–172, 1957.
PISG 共 SG 兲兹
⫺ 2 doi:10.1152/jappl.1957.10.2.165.
3. Colburn WE Jr, Evans JW, West JB. Analysis of effect of the solubility
SG共PV
⁄ PI兲SG on gas exchange in nonhomogenous lungs. J Appl Physiol 37: 547–551,
⫹ H6 , (35)
兹2 1974. doi:10.1152/jappl.1974.37.4.547.
4. Dueck R, Young I, Clausen J, Wagner PD. Altered distribution of
⬁
pulmonary ventilation and blood flow following induction of inhalation
兰 e
dx
H1 ⫽ , (36) anesthesia. Anesthesiology 52: 126 –130, 1980. doi:10.1097/00000542-
(x2⫹2x⫺2) ⁄ 2 2⁄ 2
⫺⬁ ⫹ (V̇A ⁄ Q̇t)ex 198002000-00004.
⬁ 5. Eger EI. Applications of a mathematical model of gas uptake. In: Uptake
兰
dx and Distribution of Anesthetic Agents (edited by Papper EM and Kitz RJ).
H2 ⫽ , (37) New York: McGraw-Hill, 1963, p. 88 –103.
(x2⫹2x⫺2) ⁄ 2 2⁄ 2
⫺⬁ SGe ⫹ (V̇A ⁄ Q̇t)ex 6. Eger EI. A mathematical model of uptake and distribution. In: Uptake and
⬁ Distribution of Anesthetic Agents (edited by Papper EM and Kitz RJ). New
兰
dx
H3 ⫽ , (38) York: McGraw-Hill, 1963, p. 72–87.
x2⁄ 2 2⁄ 2
⫺⬁ SGe ⫹ (V̇A ⁄ Q̇t)e(x ⫺ ) 7. Epstein RM, Rackow H, Salanitre E, Wolf GL. Influence of the concen-
tration effect on the uptake of anesthetic mixtures: the second gas effect.
⬁
兰 e
dx Anesthesiology 25: 364–371, 1964. doi:10.1097/00000542-196405000-00015.
H4 ⫽ , (39) 8. Evans JW, Wagner PD, West JB. Conditions for reduction of pulmonary
x2⁄ 2 2⁄ 2
⫺⬁ ⫹ (V̇A ⁄ Q̇t)e(x ⫺ ) gas transfer by ventilation-perfusion inequality. J Appl Physiol 36: 533–
⬁ 537, 1974. doi:10.1152/jappl.1974.36.5.533.
兰
dx 9. Hendrickx JF, Carette R, Lemmens HJM, De Wolf AM. Large volume
H5 ⫽ , (40)
x2⁄ 2 2⁄ 2
⫺⬁ SGe ⫹ (V̇A ⁄ Q̇t)e(x ⫺ ) N2O uptake alone does not explain the second gas effect of N2O on
sevoflurane during constant inspired ventilation. Br J Anaesth 96: 391–
and 395, 2006. doi:10.1093/bja/ael008.
10. Kennedy RR. A second look at the second gas effect. Anesthesiology 128:
⬁
兰
dx 1053–1054, 2018. doi:10.1097/ALN.0000000000002206.
H6 ⫽ . (41) 11. Korman B, Dash RK, Peyton PJ. Can mathematical modeling explain
(x ⫺ )2⁄ 2 2⫺4x⫹22) ⁄ 2
⫺⬁ SGe ⫹ (V̇A ⁄ Q̇t)e(x the measured magnitude of the second gas effect? Anesthesiology 128:
1075–1083, 2018. doi:10.1097/ALN.0000000000002131.
Because PVSG ⫽ PaSG, we can proceed as before and eliminate the 12. Korman B, Mapleson WW. Concentration and second gas effects: can
term containing Pv from Eq. 35: the accepted explanation be improved? Br J Anaesth 78: 618 –625, 1997.
doi:10.1093/bja/78.5.618.
PaSG 共V̇A ⁄ Q̇t兲 13. Lundh R, Hedenstierna G. Ventilation-perfusion relationships during
⫽ 兵共* ⫺ 兲H4 ⫺ 共* ⫺ SG兲H5其.
共SG ⫺ 兲共兹2 ⫺ SGH6兲
PISG halothane anaesthesia and mechanical ventilation. Effects of varying
inspired oxygen concentration. Acta Anaesthesiol Scand 28: 191–198,
(42) 1984. doi:10.1111/j.1399-6576.1984.tb02039.x.
Although we have multiplied Eqs. 26, 27, 34, and 35 by 563, the 14. Lundh R, Hedenstierna G. Ventilation-perfusion relationships during
anaesthesia and abdominal surgery. Acta Anaesthesiol Scand 27: 167–173,
partial pressure of nitrogen in dry air, to produce the curves in Fig. 8,
1983. doi:10.1111/j.1399-6576.1983.tb01929.x.
the quantitative results are only approximate since our analysis is 15. Hedenstierna G, Lundh R, Johansson H. Alveolar stability during
based on a second gas present in infinitesimal concentrations. This is anaesthesia for reconstructive vascular surgery in the leg. Acta Anaesthe-
equivalent to the simplifying assumption that V̇AIFIN ⫽ V̇AFAN in siol Scand 27: 26 –34, 1983. doi:10.1111/j.1399-6576.1983.tb01900.x.
2 2
each gas-exchanging compartment, an assumption made initially by 16. Magnusson L, Spahn DR. New concepts of atelectasis during general
Rahn and Fenn (23) to enable V̇/Q̇ lines to be drawn on the oxygen- anaesthesia. Br J Anaesth 91: 61–72, 2003. doi:10.1093/bja/aeg085.
17. Peyton PJ, Horriat M, Robinson GJ, Pierce R, Thompson BR. Magnitude
carbon dioxide diagram. Once this no longer applies, Eq. 2 must be of the second gas effect on arterial sevoflurane partial pressure. Anesthesiol-
modified to include the second gas, giving rise to two equations that ogy 108: 381–387, 2008. doi:10.1097/ALN.0b013e318164caf3.
must be solved simultaneously to yield the mean alveolar and mixed 18. Peyton PJ, Robinson GJ, Thompson B. Effect of ventilation-perfusion
pulmonary end-capillary partial pressures of each gas. These compli- inhomogeneity and N2O on oxygenation: physiological modeling of gas ex-
cated equations must be broken down into a series of expressions for change. J Appl Physiol (1985) 91: 17–25, 2001. doi:10.1152/jappl.2001.91.1.17.
19. Peyton PJ, Robinson GJ, Thompson B. Ventilation-perfusion inhomoge- 23. Rahn H, Fenn WO. A Graphical Analysis of the Respiratory Gas Exchange.
neity increases gas uptake in anesthesia: computer modeling of gas exchange. Washington, DC: American Physiological Society, 1955, p. 39.
J Appl Physiol (1985) 91: 10 –16, 2001. doi:10.1152/jappl.2001.91.1.10. 24. Rehder K, Knopp TJ, Sessler AD, Didier EP. Ventilation-perfusion
20. Peyton PJ, Robinson GJ, Thompson B. Ventilation-perfusion inhomo- relationship in young healthy awake and anesthetized-paralyzed man. J
geneity increases gas uptake: theoretical modeling of gas exchange. J Appl Appl Physiol 47: 745–753, 1979. doi:10.1152/jappl.1979.47.4.745.
Physiol (1985) 91: 3–9, 2001. doi:10.1152/jappl.2001.91.1.3. 25. Stoelting RK, Eger EI II. An additional explanation for the second gas
21. Peyton PJ, Stuart-Andrews C, Deo K, Strahan F, Robinson GJ, effect: a concentrating effect. Anesthesiology 30: 273–277, 1969. doi:10.
Thompson BR, Pierce R. Persisting concentrating and second gas effects 1097/00000542-196903000-00007.
on oxygenation during N2O anaesthesia. Anaesthesia 61: 322–329, 2006. 26. West JB, Wagner PD, Derks CM. Gas exchange in distributions of V̇A/Q̇
doi:10.1111/j.1365-2044.2006.04579.x. ratios: partial pressure-solubility diagram. J Appl Physiol 37: 533–540,
22. Peyton PJ, Fortuin M, Robinson GJB, Stuart-Andrews C, Pierce R, 1974. doi:10.1152/jappl.1974.37.4.533.
Thompson BR. The rate of alveolar-capillary uptake of sevoflurane and 27. West JB, Wagner PD. Pulmonary gas exchange. In: Bioengineering
nitrous oxide following anaesthetic induction. Anaesthesia 63: 358 –363, Aspects of the Lung (edited by West JB). New York: Marcel Dekker, 1977,
2008. doi:10.1111/j.1365-2044.2007.05355.x. p. 378 –382.
RESEARCH ARTICLE
Korman B, Dash RK, Peyton PJ. Elucidating the roles of solu- of gas, of the order of 0.5 to 1 L/min, from lungs to blood
bility and ventilation-perfusion mismatch in the second gas effect during the early stages of anesthesia (12, 23). Thereafter, the
using a two-step model of gas exchange. J Appl Physiol 128: 1587– rate of N2O uptake falls progressively toward zero. Alveolar-
1593, 2020. First published March 19, 2020; doi:10.1152/jappl- capillary uptake of the FG increases the alveolar concentrations
physiol.00049.2020.—The second gas effect occurs when high in-
spired concentrations of a first gas, usually nitrous oxide, enhance the
of other gases present, accelerating their uptake. These other
uptake of other gases administered simultaneously. The second gas gases include the commonly used volatile anesthetic agents
effect is greater in blood than in the gas phase, persists well into the desflurane and sevoflurane ( ⫽ 0.42 and 0.65) and are re-
period of nitrous oxide maintenance anesthesia, increases as the ferred to individually and collectively as second gases (SGs).
degree of ventilation-perfusion mismatch increases, and is most pro- They are commonly added because, used alone, N2O is not
nounced with the low soluble agents in current use. Yet, how low gas sufficiently potent to produce and maintain anesthesia. We use
solubility and increased ventilation-perfusion mismatch can combine the symbol for solubility in this paper. By solubility, we
to improve gas transfer remains unclear, which is the focus of the mean the blood/gas partition coefficient at 37°C.
present study. Specifically, we have used a two-step model of gas For decades, the most-widely accepted explanation of the
exchange to separate the effect of gas volume contraction, which SGE appealed to the contraction in gas volume associated with
accompanies the first gas uptake, from other factors. Step 1 involves
the uptake of the second gas at constant volume. Contraction of gas N2O uptake as the factor responsible (12). The SGE was
volume takes place in step 2 and is most effective in transferring measured in expired gas and generally assumed to be of similar
further amounts of gas to blood if the volume of second gas exposed magnitude in blood. Since 2001, it has been shown that the
to the contraction is maximized, i.e., if the loss of second gas in step effect is not only greater in blood but also continues to be
1 is minimized. Minimization depends on having a gas with a low significant for sevoflurane and desflurane when N2O uptake
solubility in blood and increases as the degree of ventilation-perfusion falls to relatively low steady-state levels and that this is due to
mismatch increases. The effectiveness of the contraction also requires the mismatch between ventilation and blood flow (V̇/Q̇ mis-
a favorable alignment with the retained second gas. Alignment de- match) that occurs routinely following the induction of anes-
pends on the solubility of both gases and the degree of ventilation- thesia (8, 10, 17–21, 23). Other manifestations of increased SG
perfusion mismatch. The model is fully consistent with classical
uptake have been described due to increased V̇/Q̇ mismatch but
concepts of gas exchange.
also only for gases with a low solubility in blood (10, 11).
NEW & NOTEWORTHY Gas exchange in the lung can always be Exactly how a low solubility combines with a high degree of
represented as the sum of two components: gas exchange at constant V̇/Q̇ mismatch to increase SG uptake remains unexplained.
volume followed by gas exchange on volume correction. Using this According to basic concepts of respiratory physiology, the
sequence to study the second gas effect, low gas solubility and
effect on gas transfer should be the opposite (16, 27, 28).
increased ventilation-perfusion mismatch are shown to act together to
enhance second gas uptake. While appearing to contravene classical These counterintuitive findings have prompted us to inves-
concepts of gas exchange, a detailed theoretical analysis shows it is tigate more closely the roles of the first and second gases and
fully consistent with these concepts. how these gases with different solubilities interact in the
presence of V̇/Q̇ mismatch to produce the SGE.
anesthetic uptake; mathematical modeling; second gas effect; venti-
lation-perfusion mismatch
METHODS
efficient , being exchanged under steady-state conditions. If such a Eq. 3 but the FI term may be ignored. In addition, the second term of
gas is delivered to the alveoli in the inspired gas mixture at a rate V̇IFI Eq. 5 may also be applied to determine the uptake of SG. These
and removed at a rate V̇AFA in a homogenous lung of given V̇A/Q̇t, relationships are discussed further in APPENDIX A.
where V̇I, V̇A, and Q̇t are the total inspired alveolar ventilation, total Description of gas uptake as a two-step process in a single lung
expired alveolar ventilation, and total pulmonary blood flow, respec- unit. The uptake of the FG in a single gas-exchanging compartment is
tively, then the uptake of the gas is V̇IFI ⫺ V̇AFA, where FI and FA are given by either side of Eq. 1. More generally, the steady-state
the fractional concentrations of FG in the dry portion of the inspired exchange V̇G, of any gas G, is given by the left side of Eq. 1. We may
alveolar gas mixture and expired alveolar gas. This must be equal to therefore write:
the gas uptake by blood, which is given by Q̇t共Fa ⫺ Fv兲, where Fa V̇G ⫽ V̇IFI ⫺ V̇AFA (6)
and Fv are given by Pa ⁄ 共PB ⫺ PH2O兲and Pv ⁄ 共PB ⫺ PH2O兲, respectively,
with Pa and Pv equal to the partial pressures of the gas in arterial and Consider now this process as taking place in two steps, the first step
mixed venous blood and PB and PH2O the barometric pressure and at constant volume V̇I and the second step during the volume correc-
saturated vapor pressure of water at 37°C, respectively, so that: tion from V̇I to V̇A. This assumption gives rise to the following pair
of equations:
V̇IFI ⫺ V̇AFA ⫽ Q̇t(Fa ⫺ Fv) (1)
V̇G(01) ⫽ V̇IFI ⫺ V̇IF⬘ (7)
For the diluent gas, which is treated as if it were completely insoluble,
we have: V̇G(12) ⫽ V̇IF⬘ ⫺ V̇AFA (8)
V̇I共1 ⫺ FI兲 ⫽ V̇A共1 ⫺ FA兲 (2) In Eq. 7, FI and F= are the fractional concentrations in dry gas at the
Assuming complete equilibration between gas and blood phases so start and end of step 1, and V̇G共01兲 is the uptake (or output) of the gas
in this step. Similarly, in Eq. 8, F= and FA are the fractional concen-
that Pa ⫽ PA, V̇I may be eliminated from Eq. 1, which gives rise to the
following general equation for the alveolar and arterial partial pres- trations in dry gas at the start and end of step 2, and V̇G共12兲 is the
sures of the FG: uptake (or output) of the gas in this step. It follows from the addition
of Eqs. 7 and 8 by comparison with Eq. 6 that V̇G ⫽ V̇G共01兲 ⫹ V̇G
共V̇A ⁄ Q̇t兲PI ⫹ 共1 ⫺ FI兲Pv 共12兲 and may therefore also be referred to as V̇G共02兲.
Pa ⫽ PA ⫽ (3)
共V̇A ⁄ Q̇t兲 ⫹ 共1 ⫺ FI兲 Note that the output of Eq. 7 becomes the input to Eq. 8. Moreover,
this treatment makes no assumptions about the nature of the relation-
In practice, the FG is usually N2O. It is also possible to obtain the ship governing the interaction between the gas G and blood and is
following expression for V̇I: therefore always applicable under steady-state conditions. Further-
共FI ⫺ Fv兲 more, defining VA as the alveolar air volume, we can add the term
V̇I ⫽ V̇A ⫹ V̇A (4) d(FAVA)/dt to the right-hand side of Eq. 1, a term which represents the
共
共1 ⫺ FI兲 ⫹ V̇A ⁄ Q̇t 兲 change in the amount of gas in the lungs with time, to extend the
principle enunciated in Eqs. 7 and 8 to the non steady-state, indicating
Extension to a nonhomogenous lung with a continuous distribution that the principle is universally applicable to gas exchange in the lung.
of V̇/Q̇ ratios. With a log normal distribution of V̇/Q̇ ratios, Eq. 4 takes When the uptake of SG occurs in the presence of a soluble FG, we
the following form: therefore use the following sequence which mimics Eqs. 7 and 8: 0 ⫽
V̇I(x) ⫽ 再 V̇A
兹2
e⫺ 冉 冊 ⁄2
x⫺
2
2 冎 both gases are introduced via the inspired alveolar ventilation to the
gas-exchanging unit; 1 ⫽ SG is then equilibrated with the blood flow
to the unit; and 2 ⫽ gas exchange is then completed by allowing the
⫹
再 V̇A
the left-hand panel of each pair, the second step in the right-
hand panel. In the left-hand panel, exchange of 1% SG is
allowed to occur before the uptake of 70% N2O. The vertical
hatched region in each graph shows the distribution of the
inflowing SG. The portion of this gas transferred to blood is
shown in yellow, that remaining after uptake is shown in cyan.
The cyan-colored area in step 1 becomes the inflow of SG to
step 2.
N2O uptake is now permitted to take place in step 2 and is
accompanied by a contraction in volume proportional in mag-
nitude to the size of the red area. Because of the large
discrepancy between the concentrations of the N2O and SG
(70% vs. 1%), the red area is scaled to approximately (1/70)th
of the N2O uptake per minute. This area shows the distribution Fig. 3. Second gas (SG) uptake profile in step 2 of schematic mechanism
of the N2O uptake. shown in Fig. 1, as a function of log10(SG) with varied from 0 to 2 in
The contraction in volume produces an increase in the partial increments of 0.01. The N2O uptake is kept constant at 575 mL/min by
pressure of SG in the gas phase, leading to a further uptake of changing FI. The blue curve represents a value of ⫽ 2 (as in Fig. 2), and the
red curve represents a value of ⫽ 0. The direction from red to blue of a
SG by blood shown in yellow. The sequence is similar to that vertical line drawn at any given solubility indicates whether SG uptake in step
shown in step 1; however, the cyan-colored area in step 1 2 increases or decreases as is increased. Red and blue curves intersect at
becomes the starting point for SG uptake in step 2 and, points X and Y. To the right of Z, gas exchange occurs predominantly in the
therefore, forms the vertical hatched region in each right-hand airways. In this simulation, the points X, Y, and Z correspond to values of
⫽ 0.5, 10, and 100, respectively.
panel.
It immediately becomes obvious that, for each SG, the
yellow area in the right-hand panel depicts the SGE, which has
On the other hand, if the transfer of SG to blood in step 2 is
now been isolated from any other contributions to the alveolar
to be maximized, SG should be as high as possible. These
uptake of SG. The largest transfer of SG occurs in Fig. 1B with
conflicting requirements for maximization suggest that, for any
desflurane. With sulfur hexafluoride and diethyl-ether, the
uptake in step 2 is virtually imperceptible (see arrows). In the given set of the other input variables (FG, V̇A, Q̇t, , FI, and
case of acetone, there is negligible further uptake in step 2. Fv), there is a value of SG for which the SGE is maximized. In
the sequence of gases shown in Fig. 1, this occurs with
DISCUSSION desflurane. In Fig. 2, we have drawn a graph of SG uptake
in step 2 vs. log10(SG), using the conditions that apply in
Having separated the uptake of SG into two components, the
Fig. 1. It can be seen that the maximum occurs for
first, which takes place at constant volume, the second associ-
log10(SG) ⫽ ⫺0.41, which corresponds to a value of 0.4 for
ated with the contraction of volume due to N2O uptake, it is
SG, consistent with the position of desflurane ( ⫽ 0.42) in
now possible to analyze the contributions of gas solubility and
the sequence shown in Fig. 1.
V̇/Q̇ mismatch to the SGE.
Role of ventilation-perfusion mismatch in relation to the
Role of the solubility of the second gas. For the SGE to be
second gas. It is now evident that V̇/Q̇ mismatch plays a
maximized, the amount of SG exposed to the contraction in net
similar role to that of SG, needing to be as high as possible in
gas volume in step 2 must be as great as possible. This gas is
step 1 so as to conserve as much of the SG as possible in the
shown in cyan in each left-hand panel of Fig. 1. Maximization
first step but as low as possible in step 2 so as to facilitate
of the cyan-colored area will occur if the loss of SG to blood
maximum transfer of the SG to blood during the second step.
during the first step is minimized, i.e., if the solubility of SG in
We can therefore expect that, for each value of SG, there is a
blood, SG, is as low as possible.
value of that maximizes the transfer of SG in step 2.
Alignment vs. retention. In Fig. 3 we have graphed the SG
uptake in step 2 as a function of log10(SG) for values of
ranging from 0 to 2. The contraction volume has been kept
constant and equal to that in Fig. 1 by varying the inspired
concentration of N2O. The red curve corresponds to the
situation when ⫽ 0; the blue curve corresponds to the
situation when ⫽ 2. The two curves cross at the points
labeled X and Y. To the left of X, an increase in V̇/Q̇
mismatch is associated with an increase in the amount of SG
transferred in step 2. Because there is already a significant
degree of SG retention in this solubility range (see Fig. 1A),
the likeliest cause of the increased SGE is the improved
alignment between the N2O uptake and the SG retained at
the end of step 1. To the right of point Y, an increase in V̇/Q̇
Fig. 2. Second gas (SG) uptake profile in step 2 of schematic mechanism mismatch is associated with a similar but smaller increase in
shown in Fig. 1, as a function of log10(SG) with FI ⫽ 0.7 for N2O and the amount of SG transferred in step 2. There can be little
⫽ 2. improvement in alignment here because of the wide sepa-
Table 1. Factors affecting gas uptake in each step of a mechanistic model of gas exchange
Factor Under Consideration Step 1 Step 2
Low effective solubility Favors retention in gas phase Impedes gas transfer to blood
High effective solubility Favors loss of gas to blood Favors gas transfer to blood
Low-degree V̇/Q̇ mismatch Favors loss of gas to blood Favors gas transfer to blood
High-degree V̇/Q̇ mismatch Favors retention in gas phase Impedes gas transfer to blood
兹2
e⫺(xk ⫺ )
2⁄2
冎
再 冎
x⫽ ⫺ (A4)
2 V̇A (FI ⫺ Fv)
⫹
兹2 关(1 ⫺ FI)e共x 兴
2⁄2
k ⫺ 兲 ⁄ 2 ⫹ (V̇A ⁄ Q̇t)e(xk ⫹ )
2
Standard graphs of ventilation and blood flow as functions of log(V̇/ (B1)
Q̇) change size and shape as is increased; the graphs for ventilation
and blood flow move away from each other in proportion to the square 关
V̇I ⫽ V̇I1 . . . V̇In⫹1 兴
of , the spread or dispersion of the graphs increases in proportion to The expired alveolar gas flow:
, and the height of each graph decreases in proportion to . An
alternative, suggested by Eq. A4, is to use (1/)log(v/q) as the V̇A 2⁄2
independent variable (10). V̇Ak ⫽ e⫺(xk ⫺ )
Using this substitution in Eqs. A1 and A2, and letting ⫽ ⁄ 2, 兹2 (B2)
we obtain the following alternative expressions for V̇ and Q̇:
关
V̇A ⫽ V̇A1. . .V̇An⫹1 兴
V̇A 2⁄2 The pulmonary capillary blood flow:
V̇ ⫽ e⫺(x ⫺ ) (A5)
兹2 Q̇t 2⁄2
Q̇k ⫽ e⫺(xk ⫹ )
Q̇ ⫽
Q̇t
e⫺(x ⫹ )
2⁄2
(A6)
兹2 (B3)
兹2 关
Q̇ ⫽ Q̇1 . . . Q̇n⫹1 兴
Drawn in this way, the graphs for ventilation and blood flow are Uptake of gas G in step 1:
always exactly the same size and shape with amplitude 1 ⁄ 兹2 and
SD ⫽ 1. When ⫽ 0, there is no V̇/Q̇ mismatch, and the two curves V̇Ik(FI ⫺ Fv)
V̇G(01)k ⫽
overlap completely. As increases, the curves move apart, with the ⫹ (V̇Ik ⁄ Q̇tk) (B4)
means for ventilation and blood flow located exactly /2 units to
the right and left of the y-axis (x ⫽ 0 line), respectively. Using 关
V̇G(01) ⫽ V̇G(01)1 . . . V̇G(01)n⫹1 兴
(1/)log(v/q) as the variable on the x-axis therefore provides a stable
platform for studying the uptake of the first and second gases when Combined uptake of gas G in steps 1 and 2:
is allowed to vary. This alternative unit has been used on the x-axis in
V̇A (FI ⫺ Fv)
Fig. 4. However, because is kept constant and equal to 2 in Fig. 1, V̇G(02) ⫽
兹2 兵(1 ⫺ FI)e(x 其
2 2⁄2
it is just as convenient to plot gas uptake in this figure as a function k ⫺ ) ⁄ 2 ⫹ (V̇A ⁄ Q̇t)e(xk ⫹ )
共 兲
of log10 V̇ ⁄ Q̇ .
关
V̇G(02) ⫽ V̇G(02)1 . . . V̇G(02)n⫹1 兴
Using the chain rule, it is possible to show that changes in the
distributions as a function of log10(V̇/Q̇) are mirrored by changes of (B5)
similar sign when the distributions are expressed as a function of
(1/)log(v/q). Hence, the direction of the arrows in Fig. 4 are the same The inflow of gas G in step 1 is obtained by multiplying the vector V̇I
as those that would apply if we were plotting against log10(V̇/Q̇). by the fractional concentration of G in inspired gas i.e., V̇IG共01兲 ⫽