Welcome! This series is an educational exercise to provide practice in
interpreting blood gas results in a clinical context. Most of the case scenarios will have
an emphasis on Anaesthetic or Intensive Care practice but interesting gas results on other
patients that I come across may be included. Most of the results are ones you could
typically come across in your daily practice so the emphasis is not on finding strange or
extreme results though some unusual results may be included. Perhaps you have come across
an 'interesting gas' - please feel free to send this along to me at firstname.lastname@example.org. The style of analysis
used here is that presented in my book "Clinical Acid-Base pHysiology".
Please consider the following clinical scenario and then
consider the question & discussions that follow.
Answer to the Problem posed in Gas of
the Week No 2.
 How would you analyse these results?
 What is the unifying feature in most cases with this disorder?
 What are the implications of the blood-gas results for further management?
Consultant (to Registrar)
What do you think of this set of results?
Based on the information given then:
At 7.1 there is a significant acidaemia present so there is an underlying acidosis
(of pCO2 & HCO3- results): The pCO2 & the HCO3-
are both elevated: This pattern occurs with a respiratory acidosis and also with a
In a respiratory acidosis, the rise in the pCO2 is the primary event and the rise in HCO3-
is the compensatory event.
In a metabolic alkalosis, the rise in bicarbonate is the primary event and the rise in
pCO2 is the compensatory event.
In this case then, the pH is 7.1 so it
is known that an acidosis is present so the choice is easy: there is a respiratory
Respiratory disorders do not provide specific biochemical clues. Such additional
biochemical evidence is really most useful with metabolic acid-base disorders. Such clues
should of course be checked for but their role in this situation is to alert you to the
presence of a metabolic acid-base disorder. For example, if the blood glucose was 35
mmol/l I would start thinking about diabetic ketoacidosis, or if the creatinine was 0.54
mmol/l, I would be alerted to consider a renal acidosis. In this case I have not been
given any biochemistry results so there are no alerting clues to consider.
So far then I have detected the presence of a respiratory acidosis, but this does not
exclude the possibility that a second acid-base disorder is present.
The general principle I use here is that
I use a simple bedside rule appropriate for the disorder diagnosed so far and estimate the
amount of compensation that should be present and compare this 'expected value' with the
measured 'actual value'. If they are close then there is no evidence for a second primary
disorder; if these values are quite different, then this is clear evidence for a second
primary acid-base disorder. By definition, the presence of two (or more) acid-base
disorders is referred to as a 'mixed disorder'.
Now to do it. For an acute respiratory
acidosis I use rule 1 (the 1 for 10 rule) which says:
"The [HCO3-] will increase by 1 mmol/l for every 10mmHg
elevation in pCO2 above 40mmHg."
Applying that in this case: the pCO2 is
70mmHg which is 30mmHg higher than the reference value of 40mmHg. Thirty is 3 lots of 10,
so the [HCO3-] should be elevated by 3 multiplied by 1 which is 3.
The reference value for HCO3- is 24mmol/l so 24 plus 3 is 27. Now to
do the comparison:
* 'Expected' [HCO3-] = 27 mmol/l
* 'Actual' [HCO3-] = 27 mmol/l.
These values are identical so I
can conclude that there is no evidence for the presence of a second acid-base
My acid-base diagnosis is an acute respiratory acidosis with no evidence for a metabolic
There is no specific biochemical check I can make for a respiratory acid-base disorder so
this is not relevant here.
If I may jump in at this point as I have a couple of queries. Why does the [HCO3-]
increase when pCO2 increases?
This is a simple physicochemical event because CO2 and HCO3-
are in equilibrium with one another:
CO2 + H2O <=>
H2CO3 <=> HCO3- + H+
So an increase in the concentration of one of the reactants on the left
hand side of this equilibrium (CO2
in this case) will drive the reaction to the right resulting
in an increase in HCO3 (in this case). This is the 'law of mass action'.
So is this just buffering then by the 'bicarbonate buffering system'?
No, not at all. A buffer system cannot 'buffer' itself. This is like giving yourself money
from your own pocket- you can't increase your net wealth in this way. If you look at the
bicarbonate reaction, the increase in pCO2 shifts the equilibrium to the right
(as drawn above) and more HCO3- is produced . . .BUT . . for
every HCO3- produced there is a H+ produced as well. This
is certainly not buffering.
What is a 'buffer' then?
The definition I use is this:
A buffer is a solution containing
substances which have the ability to minimise changes in pH when an acid or base is added
A buffer typically consists of a
solution which contains a weak acid (call it HA) mixed with the salt of that acid with a
strong base. (call this salt NaA). The principle is that the salt provides a reservoir of
A- to replenish [A-] when A- is removed by reaction with
added H+. So if for example lactic acid was added to the extracellular
fluid (ECF). The pKa of lactic acid is low (3.8) so at pH in the physiological range
essentially all is in the form of lactate (the conjugate base) and almost no lactic acid
is present. The increase in H+ is buffered by HCO3- that is, the H+
combines with HCO3- to produce CO2 and H2O. This removes
the H+ and minimises the change in [H+] (and pH). This works exactly
the same with any other metabolic acid but cannot work with the H+ produced from CO2
because it cannot alter the equilibrium between CO2 and HCO3-
So what if anything provides the buffering for a respiratory acid-base disorder?
This buffering is predominantly intracellular. It has been estimated that 99% of this
buffering occurs intracellularly for a respiratory acidosis. The most important
intracellular buffers are the proteins and the organic phosphate compounds. There will be
some buffering by the non-bicarbonate extracellular buffers but this is limited by their
relatively low total amount. Haemoglobin accounts for about 80% of these non-bicarbonate
ECF buffers and it is the imidazole groups of the histidine residues that are responsible.
Another point is that the pKa for some of the imidazole groups in haemoglobin is altered
by the state of oxygenation of the haemoglobin. These 'oxy-labile' groups are what account
for the increased buffering ability of deoxyhaemoglobin (as compared to oxyhaemoglobin).
This is the Bohr effect.
The increase in HCO3-
that occurs in an acute respiratory acidosis is sometimes referred to as 'buffering'
but as just discussed is wrong. A better term would be titration.
Where does this '1 for 10' rule that you used come from?
This derives from experiments where the arterial pCO2 of the subjects (human
and animal) was altered and the acute effect on the blood gases was measured. The '1 for
10' rule is useful because it is easy to remember and apply at the bedside. The term
'bedside rule' is sometimes used.
Are there any limits on its use?
Yes. You are limited by two things:
[1.] Firstly, if breathing room air,
there is a limit on how high the pCO2 can increase before you become severely
hypoxaemic. As a 'rule of thumb' (not perhaps a true 'bedside rule'), an arterial pCO2
of greater than 90mmHg is not compatible with life if breathing room air. (If high FIO2 is
used, then much higher pCO2 values may occur - 'supercapnia'). Using the
alveolar gas equation (in its simplified form) for an arterial pCO2 of 90mmHg:
pAO2 = [0.21 x (760-47)] - ( 90 / 0.8) =
 Secondly, the experiments were not
extended to extremely high levels of arterial pCO2 so the '1 for 10 rule'
should really only be used up to perhaps 80 to 90mmHg arterial pCO2 values. Allowing for the
variation in results then, this sets a practical limit of about 31 to 32 mmol/l for [HCO3-]
in acute respiratory acidosis. A [HCO3-] value higher than this
basically means that there is EITHER some renal compensation (retention of HCO3-)
occurring (ie a chronic respiratory acidosis) OR that there is a metabolic acidosis
present as well.
What is the unifying feature of most cases of respiratory acidosis?
The vast majority of cases are due to decreased alveolar ventilation.
The relationship between arterial pCO2
(ie paCO2), alveolar ventilation (VA) and CO2
paCO2 = k .
VCO2 / VA (where k is the proportionality constant)
This also assumes inspired pCO2
This is important for Anaesthetists because we are occasionally in the situation where the
unusual causes other then hypoventilation may be important. These are:
hyperthermia. There is a sudden major increase in CO2 production which
causes acute respiratory acidosis. If the patient is on controlled ventilation then the
ventilation is not decreased - though it is certainly lower than what is required to
excrete the increased CO2 production. This is one of the few examples when
increased CO2 production is the cause of respiratory acidosis and this is
relevant to Anaesthesia.
 CO2 Rebreathing. If an
Anaesthetised patient is rebreathing CO2
from the circuit then this can cause an acute respiratory
acidosis. The acidosis will occur if the ventilation is controlled at a 'normal'
level. If the patient is breathing spontaneously then the respiratory depression due to
the Anaesthetic agent or opioids used adds a component of hypoventilation as a cause.
The above 2 situations need to be considered, but otherwise,
hypoventilation is the unifying factor in cases of respiratory acidosis. In other
circumstances of increased CO2
production (eg exercise), the respiratory stimulation that
occurs increases ventilation and respiratory acidosis does not occur. Increased CO2 is of
course a very powerful respiratory stimulant.
Are you in agreement with my acid-base diagnosis in this case? An acute respiratory
acidosis most likely due to alveolar hypoventilation. This is all that can be said based
on the information given.
Actually I am in great disagreement with your diagnosis as you have conducted your
analysis without regard to the history. This has resulted in a wrong result.
But I was not given any history or even any biochemistry results to work with.
You have performed your analysis in a situation where the information given to you was
insufficient and you should not have proceeded. You could and should have said something
like: 'You'll have to tell me more than that -What is the clinical history? What are
the important biochemistry results? I can't really do much of an assessment of these
results until you provide more information as otherwise this could result in an incorrect
OK, I see your point and I agree with it. I have deviated from the 3 step approach you
have outlined previously and I am going to be much more careful in the future. The
principle then is something like 'No blood-gas analysis without the history'.
Let me try again. What is the history in
Consider if I had told you this: The gases were collected intraoperatively from a healthy
27 year old man who was having an elective inguinal hernia repair under a nitrous oxide,
oxygen, isoflurane, rocuronium anaesthetic. There was no significant past medical history,
he was on no medication and preoperative urea and electrolytes were all in the reference
range. He had become tachycardic and sweaty under the anaesthetic and the Anaesthetist
collected arterial blood for a gas analysis.
What do you think now?
Well this would fit very well with my previous analysis. There is no information to
suspect any pre-existing abnormality and any acid-base disorder is acute as it has
developed while anaesthetised. An acute respiratory acidosis due to hypoventilation is my
first conclusion. Alternatively, and even though it is quite rare, I always consider
malignant hyperthermia. He is receiving a triggering agent - isoflurane - and he may have
received suxamethonium so I would be interested to know this. In my practice, this would
not be routine but I know in some places sux is used much more liberally. I would like to
know his temperature. You say he is sweaty but this is consistent with the elevated pCO2.
The gas results really are well against
malignant hyperthermia (MH). The respiratory acidosis must be acute so the analysis on
this basis shows no metabolic component whatsoever. A metabolic acidosis is excluded and
if his temperature was normal I would say this would make malignant hyperthermia very
He is receiving rocuronium so he must be
on controlled ventilation. So this leaves us with a hypoventilated patient despite
controlled ventilation- so something has gone wrong in the administration of the
anaesthetic. The circuit should be checked for leaks as a first priority. Also, as you
mentioned earlier, there may be significant rebreathing occurring so are all the circuit
components functioning correctly, for example are the one-way valves on the circle system
malfunctioning? (If not a circle system, then probably a Mapleson D system such as a
Bain system is in use: such circuits are designed for partial CO2 rebreathing.
The circuit should be checked for appropriate fresh gas flows or disconnections, for
I would also comment that there are some
unusual features here. Why was the blood gas collected? Maybe this was to do with the
possibility of malignant hyperthermia. So then there should also be a set of electrolytes,
particularly checking for the K+. The gases make MH very unlikely so far so
this seems to be a circuit problem which has persisted to the extent that pCO2
has had time to rise to 70mmHg. Possibly this is an inexperienced Anaesthetist managing
I again see you point. I am pretty
confident now in my acid-base diagnosis but I need more information.
Well done. With the history and other results you can proceed much more confidently and
then ask new questions to refine the analysis. This puts you in the situation to manage
the patient to treat and correct the problem.
Yes but my initial acid-base diagnosis was not altered.
Well consider this alternative scenario: These blood gas results were collected from a 75
year man who had a long history of severe chronic obstructive airways disease. He
presented to hospital with fever, confusion and significant respiratory distress. He lived
in an apartment alone but the neighbour who brought him to hospital said he has been
unwell for about a week but had deteriorated significantly over the last 4 days. He still
smoked heavily. Gases were collected but biochemistry results are still being
Well the difference here is that I would anticipate that this would be a chronic
respiratory acidosis as he has been in sufficient respiratory difficulty for long enough
to have had renal compensation (retention of bicarbonate) to occur. It would be relevant
to check previous admissions to see what happened then and also whether bicarbonate is
usually elevated on his electrolyte results. So this then is Step One: the initial
Step 2: Acid-base diagnosis
1. pH: Severe acidaemia therefore a severe acidosis is present
2. Pattern: consistent with a respiratory acidosis as before
3. Clues - Biochemistry not yet to hand: will check when available
4. Compensation - As the clinical assessment points strongly to a chronic respiratory
acidosis, then compensation would be assessed using Rule 2: 'For a chronic respiratory
acidosis, the bicarbonate will increase by 4mmol/l for every 10mmHg rise in pCO2
above 40mmHg.' This accounts for the renal compensatory response: the retention of
bicarbonate. The 'expected bicarbonate' is therefore 24 plus (3 x 4) which is 36mmol/l.
The actual value is 27mmol/l which is 7mmol/l lower. This indicates the presence of a
significant metabolic acidosis as a second acid-base disoder
5. Formulation: A mixed acid-base disorder with a chronic respiratory acidosis and a
6. Confirmation: Need to check the electrolyte results as soon as available, particularly
the [K+] and the anion gap. Also I would ask for a lactate level
Step 3: Clinical diagnosis: The
chronic respiratory acidosis is easily explained on the clinical evidence of an acute
infective exacerbation of his chronic respiratory disease. The metabolic acidosis is most
likely a lactic acidosis explained on the basis of poor peripheral perfusion. The pO2
value of 75mmHg is higher than I would have expected but is most likely increased from his
room air value because of oxygen administration in the ambulance and in the Emergency
Department. Oxygen administration is usual in ill patients and should never be withheld
until the gases are done. A saturation value from the pulse oximeter on arrival may be
Therapy for a respiratory acidosis is
directed towards doing those things necessary to increase alveolar ventilation. This will
be much easier in the healthy intraoperative patient than with this elderly patient with
severe pulmonary disease.
So what have you learned?
The diagnosis I made was very dependent on the associated history as you have
illustrated with these two cases who shared the same blood gas result. The management is
different and indeed the detection of the significant metabolic acidosis in the second
patient is very important and the state of his circulation may not have been fully
appreciated without the blood gas information. The history in each case allowed me to
choose the correct rule for assessing the degree of compensation (ie rule one or rule
Another point also is that you need not
just a good knowledge of the technique of blood gas assessment but you also need good
clinical knowledge of the conditions as well.
I had planned to discuss some aspects of oxygen transport this week but this interesting
example has distracted us. But we have managed to discuss aspects of CO2 and
respiratory acidosis and also discuss what is required for blood-gas analysis.
I was reading an article recently which
was discussing the end-tidal to arterial pCO2 gradient and I noticed an error
in fig 1. The article is:
Wahba RW & Tessler MJ Misleading end-tidal CO2 tensions.
Canadian Journal of Anaesthesia. 1996; 43(8):862-866.
What is this error in figure 1? This is interesting to me as it is an
error commonly made by registrars during practice vivas.
If you are of a competitive nature, then email me
if you think you have the correct answer (& explanation).
The clue this week is: 'How can the error be
'explained away' by the authors?'
The above is a purely hypothetical dialogue which is presented for
the Physiology Puzzle from 'Gas of the Week No 2'
The basic question to consider here was:
Which of the 4 given points does NOT lie on the standard oxygen dissociation curve for
Answer: A clear understanding of
the oxygen dissociation curve is required for Anaesthetic or Intensive Care practice.
Registrars are often asked to draw this curve in the vivas in the ANZCA Primary exam. Four
important points to draw on the curve are the:
1. The origin: SO2 = 0 when
pO2 = 0 - Oddly enough, in the exam situation the curve is sometimes
drawn not passing through the origin so remembering this point will avoid this mistake.
2. The P50 point
3. The mixed venous point: SO2
75% at pO2 40mmHg
4. The arterial point: SO2 of
97.5% at pO2 100mmHg.
These are the best 4 points to use as
they have some physiological relevance and are probably the minimum points to use. To
quickly draw the curve, draw a suitable sigmoid shaped curve and place these 4 points on
The problem is that two of these points
do NOT actually lie on the 'standard' oxygen dissociation curve.
The conditions defined for determining
the standard curve are pH of 7.40, pCO2 of 40mmHg & a temperature of
Firstly: The mixed venous point does not lie on the curve because mixed
venous blood has a pCO2 of 45 to 46mmHg and this will cause a slight right shift of the curve (as
compared to the standard curve). This is the Bohr effect. So to be accurate 2 curves need
to be drawn to illustrate what happens with oxygen loading in the lungs (and also with
peripheral O2 unloading). Now this change is small and presumably this is the reason why most
texts ignore it. This seems unlikely though for Nunn's book because he has purposely drawn
the mixed venous point on the standard curve in the diagram which illustrates the Bohr
effect!! You will probably recognise this 'curve jumping' as similar to what happens with
the CO2 dissociation curve and the Haldane effect but with CO2 the physiological
significance is much important.
This curve jumping for oxygen due to the Bohr effect is noted in a
classic text ('The Lung' by Comroe, Forster et al 2nd ed 1962) where the following comment
is made: "A small change in blood pH occurs regularly in the body; eg., when mixed
venous blood passes through the pulmonary capillaries, pCO2 decreases from 46
to 40mmHg and pH rises from 7.37 to 7.40. During this time, blood changes from a pH 7.37
dissociation curve to a pH 7.40 curve." (on p143).
The second point that does not lie on the curve is the P50. Nunn draws
this point as point 'A' on the curve in Fig 11.9 (5th ed, p266, fig 11.9) but this is not
strictly correct. However, I did not intend to make this comment as I thought I was being
picky enough with the mixed venous point but as noted below, one registrar did email me
with this correct response. The p50
is the pO2 value at which the oxygen carrying protein (eg haemoglobin) is 50%
saturated. It is a pO2 value and should correctly be placed as a point on the
x-axis rather than as the point (SO2 50%, pO2 ~27mmHg) on the
dissociation curve. Most texts ignore this but not all. For example, the excellent
'Lecture Notes on Human Physiology' 4th ed (Blackwell, 1999) by Bray et al gets it
completely right (fig 16.28 on p446 & fig 16.29 on p447) as does Hlastala & Berger
(in 'Physiology of Respiration' Oxford, 1996) in fig 6.1 on p96. Bray also places the
mixed venous point on a slightly right shifted curve. Registrars sitting the ANZCA primary
should not be too taken by this fine print stuff as my experience is that a clear
understanding of the normal dissociation curve and its physiological relevance and the
factors that affect it is far more important knowledge to master.
prize this week goes to 2 registrars who responded with
different answers both of which were
One prize to Dale Gardiner who replied:
"The mixed venous point won't lie on the O2-Haemoglobin curve because it is at
different CO2 and the Bohr effect will shift it to another curve."
Another prize to Daniel Tsui who
replied: " My answer: The P50 point is the partial pressure of oxygen at which Hb
is 50% saturated. The normal value is ~ 27mmHg. This point does not lie on the curve but
rather on the x-axis."
As regards last week's clue: This
referred to Christian Bohr (1855-1911) who was Professor of Physiology at the University
of Copenhagen and the father of Niels Bohr (Nobel Prize for Physics 1922). Niels' son Aage
Bohr also won the Nobel Prize for Physics in 1975. The clue suggested that the Bohr effect
should be considered in solving the problem.
Kerry Brandis, 2001