Queensland Anaesthesia





HomeGas Archive

Welcome! This is the first in a series of discussions which 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 kbrandis@bigpond.net.au. The method of analysis used here is that presented in my Acid-Base text.

Please consider the following clinical scenario and then consider the question & discussions that follow.

Gas Analysis No. 1 

A 20 year old female is listed for a posterior fossa craniotomy in the prone position for biopsy and debulking of a large posterior fossa tumour. She had a 2 month history of headaches and dizziness progressing to poor balance and a gait disturbance. No significant past history and usually on no medication. She has been commenced on oral dexamethasone 4mg 6th hourly, ranitidine and phenytoin preoperatively.

Na 135, K 4.0, Cl 102, bicarbonate 23, urea 4.2 & creatinine 0.05 (all in mmol/l). Albumin 44 g/l, Hb 151 g/l. Coagulation profile was normal.

Following an intravenous induction with propofol & rocuronium, she was intubated and ventilated. Maintenance was with oxygen, nitous oxide and sevoflurane. An indwelling urinary catheter was inserted, Mayfield head tongs were applied and the patient was positioned prone and slightly headup. All monitoring was checked.

Blood gases
About 45 minutes into the procedure when ventilation, blood pressure and end-tidal pCO2 had been stable for 15 minutes, blood was collected anaerobically from the arterial line and analysed within 5 minutes. It was noted that FIO2 was 0.32 and endtidal pCO2 was 29mmHg. The results were:

pH 7.51  
pCO2 32 mmHg
HCO3 25 mmol/l
pO2 161 mmHg
BE +3.1  
Hb 139 g/l
Na 137 mmol/l
K 4.0 mmol/l
Ionised Ca 1.06 mmol/l


[1] How would you analyse these results?
[2] What is the value of the endtidal-arterial pCO2 gradient? (and what is this an index of?)
[3] What are the implications of the blood-gas results for further management?

Theatre Discussion

Consultant (to Registrar)
What do you think of this set of results? How would you assess the situation?

Well I would analyse the results using a systematic approach. The 6-step approach I would  use is:

1. pH
First look at the pH to see whether a net alkalaemia or acidaemia was present

2. Pattern
Then I would review the pattern of the arterial pCO2 and bicarbonate results - this would give an idea whether a simple acid-base disorder only was present or may suggest that a mixed acid-base disorder is present

3. Clues
I would then look for ‘clues’ in the other biochemistry results to see if there was any suggestive evidence of the presence of a particular acid-base disorder

4. Compensation
Next, I would assess the appropriateness of the compensatory response using a set of ‘bedside rules’

5. Formulation
I would then consider all this evidence and formulate an acid-base diagnosis

6. Confirmation
Finally, I would check for any specific biochemical evidence of a particular disorder to try to confirm the presence of a specific acid-base disorder

Well, that sounds fairly thorough about HOW you would do it. So, can you apply your technique to these particular results then and see what you find.

Going through the steps then:

1. pH 7.51 - A net alkalaemia is present so there must be at least an alkalosis present, either resiratory or metabolic. This does not rule out a mixed disorder.

2. Checking the pattern of the PCO2 & the bicarbonate: What I am looking at here is whether the results are:

both higher then normal
both lower then normal
one result moves in one direction and the other result moves in the opposite direction (or is in the normal range)

These are the 3 possibilities or ‘patterns’ that I am checking for. In this case: the pCO2 is low (32mmHg as compared to a normal value of 40mmHg) and the bicarbonate at 25mmol/l is close to the normal value of 24mmol/l. I would consider this then to match the third pattern. A low pCO2 with a normal bicarbonate: this means the alkalosis I detected in step 1 is a respiratory alkalosis and not a metabolic alkalosis. With a metabolic alkalosis, the bicarbonate would be increased but it is still in the normal range here.

3. ‘Clues’ : The other results reported on the blood gas machine are all within the reference range. The recent preoperative results are all normal and the anion gap is also normal at 10mmol/l. So there is nothing here that alerts a suspicion of anything else.

4. Compensation for an acute respiratory alkalosis is assessed using the ‘2 for 10 rule’ ie I would expect the plasma bicarbonate to decrease by 2mmol/l for every 10mmHg decrease in pCO2. So at 32mmHg - this is close to a 10mmHg decrease so the bicarbonate should decrease by almost 2 from the standard ‘normal’ value of 24mmol/l - that is, to 22mmol/l. Now at 25 mmol/l it is slightly higher then I would predict so it looks like there is indeed a mild metabolic alkalosis present. This is small but explains why the pH is a little higher than expected for a pCO2 of 32mmHg. I hadn't considered this much in step 2 as I was comparing the actual value against the 'normal' value of 24mmol/l but I should have been comparing the 'actual' value against the 'expected value' of 22mmol/l.

5. So the diagnosis I would make is a respiratory alkalosis and a mild metabolic alkalosis as well.

6. Confirmation: For this particular acid-base situation there is really no reason to look for confirmation in other biochemistry results

Very impressive. But you don’t seem to have considered the pO2 result. Why not?

Well, because I don’t need need to. The acid-base diagnosis can be made by looking at the pH, pCO2 & the HCO3 results only.

The pO2 doesn’t tell me anything about the acid-base status, only about the amount of oxygen in the arterial blood.

Of course, other results provide useful supportive information.  As you know, I have for example used the electrolyte and other biochemical results to check the acid-base results. This is useful because certain acid-base disorders cause changes in these and can help me make the diagnosis. For example, if a metabolic acidosis was present, then I would use the anion gap to differentiate the types of metabolic acidosis.

Hmm. I don’t think you’re correct in that. Haven’t you made some assumptions here which you haven’t explained?

Well possibly but I’m pretty confident the diagnosis is correct. What is it that worries you about my comprehensive acid-base analysis?

Well, let me ask you this then. In your step 4, you used the ‘2 for 10’ rule which is the ‘bedside rule’ for assessing an acute respiratory alkalosis.

But there is also the ‘5 for 10 rule’ for a chronic respiratory alkalosis.

Why didn’t you use this other rule instead? If you had then you would have predicted a bicarbonate result of (24 minus 5) or 19mmol/l - this then is quite different from the ‘actual’ bicarbonate result of 25mmol/l and would give good evidence that a more marked metabolic alkalosis was present.

Yes, I see your point then sir. But it was obvious that this young lady was in good health preoperatively and did not have any clinical evidence of any other acid-base disorder. So it was obviously appropriate to use the rule for an acute disorder rather than a chronic disorder.

Obviously! That is getting closer to the point I was attempting to make. You made the decision about which rule was appropriate based on your ‘clinical knowledge’ of this patient. This is what I think you need to state explicitly if you are to understand the process of making the diagnosis in this patient. An acid-base diagnosis by itself is basically useless.

I don’t follow what you are saying. Do you disagree with my assessment? I have followed a systemic approach and I am confident about my diagnosis. This is what it is all about.

I do tend to agree with you but I think your approach though good is not complete. What I think you need is a structured approach consisting of 3 aspects (which also includes what you have done). Consider this:

The 3 components of this analysis are:

 Firstly: Initial Clinical Assessment
An Initial Clinical Assessment is an essential first step

i. From the history, examination & initial investigations, you can make a clinical decision about what is the most likely acid-base disorder(s)
ii. This is very important but you must remain aware that in some situations, the history may be inadequate , misleading or the range of possibilities large
iii. Mixed disorders are always difficult and the history & examination alone are insufficient

Secondly: Make an Acid-Base Diagnosis
Perform a systematic acid-base evaluation of the blood-gas and other results and make an acid-base diagnosis

Finally: Make a Clinical Diagnosis & act upon it.
Synthesise all the information and make an overall clinical diagnosis and respond to it.

The 3 components are:
1. Clinical assessment
2. Acid-base diagnosis
3. Clinical diagnosis

What you did was incomplete as you only carried out the middle step of making an acid-base diagnosis.

Yes I think I see what you mean. What I did was basically correct but I was implicitly partly using the clinical information to select one rule over another but I wasn’t clearly understanding or explaining what I was doing. Then I stopped just with the acid-base diagnosis of a respiratory alkalosis and really didn’t do anything with it. In a sense, like ordering a test and not doing anything about it. Bad medicine I suppose.

Yes, I agree with you. You need the clinical information to guide you to the correct acid-base assessment then you use these to formulate a plan of what to do.

So would you like to try again using this more complete approach?

Well, as I’m new to this way of working let me do this slowly.

Firstly: The Clinical Assessment.
This is a young healthy patient with a posterior fossa tumour. In a healthy patient, my expectation is that there will be no initial acid-base disorder present. However, such a patient could have raised intracranial pressure and have been vomiting which could cause a metabolic alkalosis and this could be associated with hypokalaemia. Also applying my clinical knowledge, I know that there is a difference in the mechanism of raised intracranial pressure with a supratenorial tumour and with a posterior fossa tumour.

A supratentorial tumour grows within the fixed volume cranial cavity but ICP does not initially rise because of compensatory processes, particularly ‘CSF translocation’ (ie the tumour grows and CSF moves out into the spinal subarachnoid space and the ventricles become smaller). Oedema surrounding the tumour also contributes.

A posterior fossa tumour can obstruct CSF drainage in the narrower confines of the posterior fossa and cause raised ICP. This can occur earlier (as the space is smaller) and as the problem is obstruction then the ventricles are large, rather than small.

In this patient I know that there is little clinical evidence of raised pressure (eg no vomiting or papilloedema, but there is a headache) and the MRI & CT scans do not show obstruction as the ventricles are normal in size. Additionally the patient is on dexamethasone which reduces oedema surrounding a cerebral tumour. The patient is also on ranitidine so even if vomiting were present then there is little chance of a significant metabolic acidosis occurring. The preoperative electrolyte profile shows a normal ‘bicarbonate’ which is against the presence of any significant metabolic disorder.

She is talking normally and is not mentally obtunded. There is no clinical evidence of respiratory depression so no suspicion of a respiratory acid-base disorder.

Overall, there is no preoperative evidence of a pre-existing acid-base disorder.

At the time of collection of the gases, the patient is anaesthetised, paralysed and ventilated with a pCO2 of 29mmHg and stable haemodynamics. As the patient was in good cardiorespiratory health I would expect the arterial pCO2 to be higher then this, perhaps by 2 to 8mmHg. I usually work on an expectation of 5mmHg based on an old study I think by Nunn. So possibly an arterial pCO2 of about 34mmHg or so. This would lead me to expect that the arterial blood gas would show a respiratory alkalosis, and it would be acute and mild.

Good. So putting together all the clinical evidence you have you have come up with a very reasonable prediction of the acid-base status. All this so far without using any information from the blood gas results. Continue.

Well the second step is to make the Acid-Base Diagnosis. I would follow the 6-step systematic procedure I did earlier but the reason why I select the ‘2 for 10 rule’ is now obvious. My final assessment then is as before: a mild respiratory alkalosis with probably a slight metabolic alkalosis as well.

Now: the putting it all together phase: Making a Clinical Diagnosis.
Well the cause of the respiratory alkalosis is hyperventilation. This is not a particularly difficult conclusion to make as virtually the only process that causes a lowered pCO2 is an increase in alveolar ventilation.  The actual relationship is:

paCO2 = k . VCO2 / VA

where  VCO2 is body CO2 production per minute, VA  is alveolar ventilation & k represents the proportionality constant. So theoretically a decreased CO2 production could also be a cause of a low arterial pCO2 but in practice this is virtually never the case.

So what is the hyperventilation all about?

The patient is on controlled ventilation so the hyperventilation is a primary process. What we aim to achieve clinically is a slightly lowered arterial pCO2 to cause some cerebral vasoconstriction to decrease intracranial pressure to facilitate surgery and patient survival.

Excessive vasoconstriction is avoided as this could compromise cerebral oxygenation and could lead to postoperative increased cerebral blood flow. This is because compensation would occur with a lowered brain interstitial fluid bicarbonate which would start returning the brain ISF pH towards normal. At the end of the operation then, a ‘normal’ arterial pCO2  would result in an an increase in cerebral blood flow (and in intracranial pressure). The decreased brain ISF pH due to the ‘normal’ arterial pCO2 will now stimulate the respiratory centre and this will counteract the effect. However, postoperatively, respiration may be depressed by the residual effects of the anaesthetic agents and by opioids so this  increased ventilation response may be obtunded.

Now tell me about why the arterial is higher then the end-tidal pCO2 in this patient?
- a
nd -

What is the significance of this (art-ET)pCO2 gradient?

Well as I understand it the situation is that CO2 diffuses very readily across the alveolar walls, thus in the lungs, the pCO2 of alveolar and end-capillary blood is virtually identical. The blood in all the pulmonary capillaries joins together to become the arterial blood so it has a pCO2 that is the same as the ‘average’ alveolar pCO2. (Indeed this is so well recognised that the measured value of the arterial pCO2 is substiituted in the Bohr equation for the ‘ideal alveolar’ pCO2)

Now the situation with expired gas is that it can be considered to be a mixture of this alveolar gas from these perfused alveoli COMBINED with the gas that comes from any unperfused alveoli. Such unperfused alveoli are of course known as the ‘alveolar dead space’ and the pCO2 composition of this gas is virtually zero. So the mixing of gas from perfused alveoli (pCO2 = paCO2) with gas from the alveolar dead space (pCO2 = 0) results in a lowering of the end-tidal pCO2. (It is at the end of expiration that the gas from the different alveoli mix, so an end-tidal sample contains the mixture of these gases).

The net result is that end-tidal pCO2 is lower than arterial pCO2. In healthy young people (who have little alveolar dead space) the (a-ET)pCO2 gradient is small, typically 2 to 5 mmHg. Under general anaesthesia the gradient may increase and may be as much as 5 to 10mmHg. In any case, there is variability and many of our patients are not healthy so it could be higher then this. If there is hypotension then the amount of non-perfused areas of the lung (or areas of high V/Q ratio) may increase.

So, the practical implication of all this is that we monitor endtidal pCO2 continuously but ‘calibrate’ and check against intermittent arterial pCO2 sampling because it is the arterial pCO2 that we are aiming to reduce (to cause cerebral vasoconstriction) in certain neurosurgical cases.

The size of the (a-ET)pCO2 gradient is an index of the amount of alveolar dead space. The gradient increases (eg to >15mmHg) in situations where the amount of alveolar dead space is increased. The most particularly important clinical situations are with pulmonary embolism (acute situation) and ventilation of pulmonary bullae (chronic situation). Indeed with a large pulmonary embolism, the amount of alveolar dead space can be very large indeed!

Very good, I see that you have a pretty good understanding of the physiology and of the clinical requirements.

 But I have to pull you up on one thing: you say that the pCO2 of gas from unperfused alveoli (alveolar dead space) is zero - Are you sure about that? Really sure?

Well its obvious, though you seem to be saying I’m wrong!

CO2 is delivered to the lungs in the pulmonary arterial blood and is evolved into only those alveoli that are perfused. However, if you mean that there is some perfusion of this ‘alveolar dead space’ then it is really a situation of ‘high V/Q ratio’ rather then ‘alveolar dead space’ (V/Q of infinity). If this is what you are implying then it really isn’t true ‘alveolar dead space’ so I would say you are mistaken.

Alternatively, there will of course be some metabolism in the pneumocytes lining the unperfused alveoli and in their walls as well so there will be a very small local CO2 production. I think this would be so small that it would not account for any significant pCO2 level in the alveolar dead space.

So my answer is I’m sure: if they are NOT perfused then they absolutely cannot take part in gas exchange so there cannot be any CO2 evolved into them, and local CO2 production is too small to be considered.

Also, I must admit I had heard that you asked one of the other registrars this question last week so I took the precaution of looking it up as an extra check before I came along today. For example, in “Anesthesia Review” by Michelle Bowman-Howard (Lippincott Williams & Wilkins, published 2000) on p115 she says “The pETCO2 of dead space is zero”. And on p249 of “Physiology” 3rd edition by Bullock, Boyle & Wang (Williams & Wilkins, 1995) it has a diagram (fig 18-12) of an alveolus labelled ‘Alveolar dead space’ which has the actual gas values in the alveolus indicated: pO2 150mmHg (the same as in humidified inspired gas) and pCO2 of 0 mmHg just as I said.

I am impressed by your application of the basic respiratory physiology and your preparation for today's list. You certainly have a good understanding of this.

BUT, you are indeed quite wrong. However, I think you can work it out yourself, because this requires an understanding at a very simple level and not memorising without understanding, or complex analysis or looking something up in a book. I’ll give you a clue, though a somewhat cryptic one: Many many years ago when I was a child we used to sing a little ditty, which as I recall included words something like this:

“Blood goes round and round, air goes in and out” . . . etc 

The words may not be completely correct (it has been many years) but the principle is clear.

(after some thought): . . . .Now I understand. Its so obvious!

Excellent. We seem to have strayed a little from just talking about acid-base physiology but it has certainly been interesting. One point from the earlier discussion though: When I asked you about the use of the pO2 value from the blood gas results  your comments were not quite correct and I was concerned you did not quite understand the principles of oxygen transport. So have a think about what you said and we’ll perhaps discuss it further next week.

Have you worked out the obvious answer to the little mystery posed above which has confused even the authors of prestigious texts?
 I will post the answer with the next Gas 

The above is a hypothetical dialogue which is presented for educational purposes.

[Back to Top]

Kerry Brandis, 2001

Last updated Sunday, 04 August 2002 12:53 PM EST

Hit Counter