Home | Gas Archive

Welcome! This is the 4th in this series. Please consider the following clinical scenario and then consider the questions & discussion that follows.

History
A 63 year old man was evacuated to Australia by Air from a Pacific island. He had been admitted to hospital there and was transferred in a stable condition with a diagnosis of "left lower lobe pneumonia, exacerbation of chronic obstructive airways disease & probable pulmonary embolus". Treatment prior to arrival here was with antibiotics, salbutamol, frusemide, prednisone and warfarin.

Past history was of 'bullous emphysema' with several admissions in recent years to our hospital with acute infective exacerations of his chronic respiratory disease. His previous admission here 9 months previously was complicated by pulmonary embolism and he had been ventilated via a tracheostomy in the Intensive Care Unit. He had been on daily prednisone for the previous few years. 

On admission, he was febrile and in severe respiratory distress. A small amount of purulent sputum was noted. Crackles were heard throughout his lungfields. Heart sounds were normal. Initial coagulation testing showed an INR of >7.0. Mild abdominal pain was present.

He was initially managed in a General Medical Ward but developed severe abdominal pain 3 days after admission. An erect xray showed a large amount of gas under each hemidiaphragm. He was transferred to Intensive Care for assessment, insertion of lines and stabilisation prior to theatre for laparotomy. He received fresh frozen plasma (FFP) to rapidly correct his coagulopathy. 

Preop Investigations
Na 132, K 3.8, Cl 106, bicarbonate 25, urea 7.3 & creatinine 0.07 (all in mmol/l). Haemoglobin 135 g/l. INR 3.1 (prior to FFP).

On arrival in theatre
He was mildly febrile and tachypnoeic. He was able to talk and was not confused. Saturation was noted to be 85% (SpO₂) on room air and blood gases were collected from the arterial line in theatre prior to Anaesthetic induction.

Blood gases
 The results (while breathing room air) were:

Questions

[1] How would you analyse these results?
[2] What is the most serious abnormality present?
[3] What would you do about it?

Consultant (to Registrar)
What do you think of these gases? 

Registrar
I could discuss these from the acid-base point of view but the thing that sticks out here is the low arterial pO₂. This requires comment first.

Consultant
Why? What has the pO₂ value got to do with the acid-base situation? Why don't you just look at the ABG results and give me your analysis?

Registrar
The pO₂ is very low and warrants immediate attention particularly as the situation for the patient can be improved simply by administering an increased oxygen concentration. The 'blood-gases' are not just about the acid-base balance but provide other information as well. One way of thinking about this is to consider that the 3 main 'gas' results (pH, pCO₂, pO₂) in arterial blood gas report give information about three different, but interrelated processes:

pH -> acid-base state
pO₂ -> oxygenation
pCO₂ -> alveolar ventilation

This frame of reference allows one to dissect out these three components and make a quick overall assessment. So a glance at these results tells you immediately that he is alkalotic, hyperventilating and hypoxaemic. The most prominent problem is that this patient's ability to oxygenate his arterial blood is poor. This is no real surprise as it is consistent with the obvious significant respiratory dysfunction in this patient.

Consultant
Lets focus on oxygenation this week then. What arterial pO₂ would you have accepted as 'normal'?

Registrar
This patient is hypoxaemic because this is defined as an abnormally low arterial pO₂ in the arterial blood. Definitions vary a little but any person with a value less than 50mmHg is hypoxaemic by any definition.

The reference range for arterial pO₂ decreases with increasing age for adults over 20 years old. This can be expressed as:

pO₂  = 100 - 1/3 (age in years) +/- 10 mmHg

(The +/- 10mmHg gives the 95% reference range)

For this 63 year old man then:
Expected pO₂  = 100 - [(1/3)x(63)] +/- 10  =  79 mmHg (range 69-89mmHg)

Consultant
In general terms, what are the causes of hypoxaemia?

Registrar
A general systematic approach would be to consider the 5 possible causes of hypoxaemia:
1. Low inspired pO₂
2. Hypoventilation
3. Diffusion block 
4. Ventilation-perfusion inequality
5. Shunt

Proceeding through these systematically in this patient then:
1. This patient is breathing room air (F_IO₂ = 0.21) so cause 1 is not the problem.  I should state this more accurately because it is not really the fractional concentration that is important but the inspired pO₂. Afterall, on the top of Mount Everest, the F_IO₂ is still 0.21 but the barometric pressure is low so the inspired pO₂ is low.

2. Hypoventilation is also excluded because the arterial pCO₂ is not elevated. As mentioned above the arterial pCO₂ provides an accurate measure of the level of alveolar ventilation. A normal or low arterial pCO₂ means that hypoventilation cannot be present. This is interesting because it means that hypoventilation as a cause of hypoxaemia can be excluded immediately in most cases just be inspecting the arterial pCO₂ on the gas results. 

3. Diffusion block: This is easy to exclude because it is never a practical cause of hypoxaemia at sea level. This is generally only a theoretical concern.

(Except perhaps in people exercising at altitude where there is a lower gradient and a shorter time of contact between pulmonary capillary blood and alveolar gas. In this stressed situation, any diffusion abnormality may be revealed).

4 & 5.  These 2 causes can be differentiated by the the response of the patient's arterial pO₂ when the patient is given 100% oxygen to breathe. The hypoxaemia due to V/Q inequality can be overcome by 100% O₂. With shunt and 100% O₂, arterial pO₂ will increase but the increase is less. 

Consultant
So what values of pO₂ distinguishes between these two causes? 

Registrar
For V/Q inequality:
When the low arterial pO₂ is due to 'V/Q inequality', the administration of 100% oxygen will always result in the elevation of arterial pO₂ to very high levels. This happens no matter how much V/Q inequality is present. One catch is that the rise can sometimes take a while as it may take some time to flush all the nitrogen out of areas with low ventilation (ie low V/Q ratios). But give it time and the arterial pO2 will always rise to high levels. The key point here is that this always happens. 

For shunt:
When the hypoxaemia is due to 'shunt' (meaning right to left shunt), the administration of 100% oxygen will elevate the arterial pO₂ but the actual level reached depends on the amount of shunt (which we can express as % of cardiac output). If the shunt is more than about 10%, then the arterial pO₂ will never rise higher then 500mmHg.

Consultant
So a value of 500mmHg represents the 'cutoff point'?

Registrar
Well thats probably a reasonable working rule of thumb. It also takes account of the wide variability whereby studies have shown that healthy people can be found to have a gradient of as much as 100mmHg between alveolar and arterial pO₂ when breathing 100% oxygen.

When breathing 100% oxygen, each 1% of shunt adds about 15mmHg to the alveolar-arterial pO₂ gradient. For a person breathing 100% O₂, the alveolar pO₂ calculated from the alveolar gas equation (assuming R= 0.8 & p_aCO₂ = 40mmHg) is 660mmHg.The 'normal' 1 to 2% shunt in healthy people drops the arterial pO₂  to 30 to 45 mmHg lower than this. An 11% shunt (by the rule of thumb) will drop the arterial pO₂ by about 165mmHg which brings the paO₂ to this cutoff level of about 500mmHg.

Consultant
But surely the question really is: So what? Is the ability to differentiate between V/Q inequality and right to left shunt of therapeutic importance? How would you correct hypoxaemia?

Registrar
I take your point as the therapeutic options are limited. However their effectiveness is usually adequate in most cases. The immediate options are basically:
*  increasing F_IO₂ (easy to do & applicable to every situation where there are oxygen supplies)
*  intubation & ventilation

Ventilation allows the clinician to manipulate most ventilatory parameters including tidal volume, respiratory rate & pattern, I:E ratio & application of PEEP (in addition to controlling F_IO₂ of course). 

Consultant
So, thats the theory. What would you do here then?

Registrar
This patient is just about to have an Anaesthetic. The induction sequence for this patient with an 'acute abdomen' will include preoxygenation with 100% oxygen. In the meantime, the patient must be given an increased inspired oxygen concentration and continuously monitored with pulse oximetry. 

Consultant
What do you think of the acid-base situation here then?

Registrar
This patient has respiratory failure due to an infective exacerbation of chronic respiratory disease associated with a perforated abdominal viscus and abdominal sepsis. A respiratory acid-base disorder and/or a metabolic acidosis are the most likely possibilities.

The pH indicates the presence of an alkalosis. The pattern of a slightly elevated [HCO₃] and a slightly depressed pCO₂ suggests a mixed disorder. The arterial pCO₂ is slightly lowered so will cause an increase in pH but at a pCO₂ of 35.6mmHg any respiratory alkalosis is very mild. Is there any evidence of a metabolic acidosis? The anion gap is not elevated. The plasma chloride level is slightly elevated. No lactate level was reported. Lactic acidosis is associated more with poor perfusion rather than hypoxaemia. None of the other biochemistry results provides support for any other acid-base diagnosis. On the available evidence there is no evidence of a serious acid-base disorder. A mild respiratory acidosis is explained by the patient's respiratory distress. The slight elevation in bicarbonate could be due to the diuretic use. 

A comment on the information provided: There are no details of cardiovascular status (eg BP, pulse, urine output, CVP, assessment of peripheral perfusion) and these along with a lactate level provide very useful information. Knowing such relevant clinical details is necessary.

Consultant
I agree with you. But from what you have said so far, the disorders here are the hypoxaemia due to the respiratory disease and a mixed alkalosis.

What about the possibility of a respiratory acidosis (due to the respiratory disorder) with an elevation of bicarbonate occurring as a renal compensatory response and then the acute abdomen and associated pain causing hyperventilation and a superimposed respiratory alkalosis. So you would have a respiratory acidosis (with renal compensation) complicated by an acute respiratory alkalosis. Wouldn't that explain the findings?

Registrar
Sounds almost plausible sir, but you are making several errors. In particular:

[1] A respiratory acidosis and a respiratory alkalosis cannot co-exist. You can have one or the other present at any one time but NEVER both. They could follow each other but they cannot both be present at the same time. Actual arterial pCO₂ cannot be both higher and lower than the 'expected' pCO₂ at the same time. (The 'expected' pCO₂ is used as the reference value instead of 40mmHg so that respiratory compensation is not mistaken for a primary respiratory acid-base disorder.)

[2] Your comment about the slightly elevated bicarbonate being possibly due to renal compensation for a recent respiratory acidosis is plausible but speculative. There are often several acid base possibilities and I think you would be unwise to put too much emphasis on this one possibility. You should be cautious and review what evidence you have. For example, if an arterial result from a day or two previous showed an elevated pCO₂ and the bicarbonate had risen since, then this would be supportive evidence.

[3] I am also a little concerned about your use of the term 'hypoxaemia'. By this you mean an abnormally low arterial oxygen tension (paO₂) but the term is used in different ways by different people. For example, some use the term to mean not just a low tension but also a low oxygen content even if the tension is normal. An anaemic patient with a [Hb] of 6g/dl and an arterial pO₂ of 100mmHg on room air would be considered to be 'hypoxaemic' by some but not by your definition. Others seem to use the term 'arterial hypoxia' for a low arterial oxygen tension and/or content. We should be careful that the people we are talking to have the same understanding of these terms that we do.

Consultant
Very good. I will try to be more cautious in future. The following ABG results were obtained about half an hour after intubation. He was fully ventilated on 100% oxygen at this time:
pH 7.47, pCO₂ 40mmHg, pO₂ 131mmHg, [HCO₃ ] 28.7mmol/l.
Care to comment?

Registrar
This confirms a large shunt fraction. As a general working rule, I just plug the values into Nunn's iso-shunt diagram (eg see Nunn 5th ed p186) rather than attempt any fancy calculations myself. Just estimating from a visual inspection of the diagram, this corresponds to about a 25% shunt. This may not be completely accurate as cardiac output will have altered with Anaesthesia and the assumption of a constant 'normal' arterial-mixed venous oxygen content difference may not be correct. Consider it a ballpark figure. 

A final comment here is that here is a patient with an arterial pO₂ of over 100mmHg and yet this is the most severe and significant abnormality on this set of gas results.  

Consultant
Yes, that is one to be well aware of. It is also a situation that gets asked in exam vivas. You would be shown a set of results with the pH, pCO₂ & [HCO₃] results only slightly abnormal and sitting in there is a pO₂ result of say 80 to 100mmHg and the incidental information about breathing 100% oxygen (maybe gases collected during preoxygenation with 100%O₂). You would then be asked: 'Which of these results indicates the most severe abnormality?' My experience with asking this question is that few registrars spot the huge gas exchange problem and instead try to construct an acid-base problem from meagre details.

While we have been talking about oxygenation of arterial blood today we really have not considered the important aspects of oxygen content and oxygen delivery to the tissues and have instead concentrated on the oxygen tension. You alluded to this issue when you mentioned cardiac output and the effect of anaemia. We will discuss this further in the future, but let me leave you with this question (which relates to this topic):

What is Hufner's number? (& what is its physiological significance?)

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