The basics of arterial blood gas (ABG) analysis
What do we hope to achieve in this course? I hope to show you how to read an ABG properly. I also want to show you how to avoid the common pitfalls we make in ABG analysis. In my experience, mistakes are made for a limited number of reasons. In Case 1, you will learn how to analyse an ABG in the presence of oxygen therapy. This can help you avoid serious error. In Case 1 and Case 2, I show you why calculation of the A-a gradient is so important in ABG analysis. In Case 3, we will discuss the importance of recognising the transition from type I to type II respiratory failure (or pattern) in a sick patient. In Case 4, we show you the value of a VBG in clinical practice. In each of these case scenarios, we try and teach you important aspects of the underlying conditions encountered. We include a discussion article showing you how to fabricate an ABG result for teaching purposes-you can learn a lot from this exercise. The emphasis in this section (at the moment) is very much on assessment of arterial oxygenation.
back to course menu
General description of an ABG
pH, paO2 (mmHg or kPa), paCO2 (mmHg or kPa), HCO3- (mmol/l)
Four key variables are measured on the ABG. We measure the concentration of two gases dissolved in the arterial blood. The partial pressure of oxygen in the arterial blood (the paO2) and the partial pressure of CO2 in the arterial blood (the paCO2). Remember the partial pressure of a gas is just a way of measuring the concentration of that gas. The concentration of the metabolite, bicarbonate (HCO3- in mmol/l) and the pH of the arterial blood (an inverse log scale of H+ concentration) complete the four. The ABG allows us to examine aspects of the oxygenation of the arterial blood. It also allows us detect acid-base disturbances in the extracellular fluid (ECF). ABG findings are never diagnostic but give us clues as to how and where to look for pathology! They also guide treatment in several situations. You need to be able to read an ABG.
‘You get too much you get too high, not enough and you’re gonna die’ (The Sweet c1978)
The paO2, the partial pressure of oxygen in the arterial blood, is the concentration of oxygen dissolved in the arterial blood. The paO2 represents a tiny fraction of the total amount of oxygen in the arteries, yet, it is a key physiological variable. Why? It is a key variable because the paO2 determines the percentage saturation of haemoglobin (SaO2) and the percentage saturation of haemoglobin in turn determines the total amount of oxygen in the arterial blood. Haemoglobin is a massive oxygen store and if 90% of the oxygen binding sites on haemoglobin are occupied by oxygen, then the total content of oxygen in the arterial blood is more than sufficient to supply tissue requirements.
The relationship between the paO2 and the SaO2 is such that 90% or more of hemoglobin oxygen binding sites are occupied by oxygen if the amount of dissolved oxygen in the arterial blood (the paO2) is at or above 60 mmHg (8kPa). Due to its impact on haemoglobin saturation, a paO2 below this level will result in tissues being deprived of oxygen (true tissue hypoxia). The numbers here are important and are worth remembering: 60 mmHg (8 kPa) and 90%. We will see that the body holds the paO2 way above this minimum required level under normal circumstances.
We tend to think of O2 as king but CO2 is, in fact, king. paO2 plays second fiddle to paCO2. Groups of neurons in the brainstem control the rate and depth of respiration by controlling the action of the muscles of respiration. The rate and depth of respiration determines the volume of air per unit time entering the alveoli (the ‘alveolar ventilation’). The alveolar ventilation is set by the brainstem at a level sufficient to eliminate (in the exhaled air) all CO2 produced during metabolism.
While the alveolar ventilation is set at a level to ensure elimination of all CO2 produced during metabolism, it is crucial to realise that the alveolar ventilation is also the key determinant of oxygen delivery to the alveoli in inhaled air. The alveolar ventilation is the key determinant of the partial pressure of oxygen in the alveoli (the pAO2). As oxygen diffuses rapidly across the alveolar capillary membrane, the pAO2 is in turn the key determinant of the level of dissolved oxygen in the arterial blood (the pAO2 determines the paO2).
Within the alveolar capillary unit, oxygen diffuses across the alveolar capillary membrane equilibrating with the pAO2 in the alveolar lumen. This process is rapid occurring within the first one third of the journey the deoxygenated blood from the right side of the heart takes through the alveolar capillary unit. Crucially however, as we will see, it is not 100% efficient. The paO2 never quite reaches the level of the pAO2.
In a young person, lets say aged 20 years, breathing room air, the need to eliminate all CO2 produced during metabolism necessities an alveolar ventilation (generally around 4 litres/minute) which results in an alveolar pAO2 in the region of 100 mmHg. As oxygen diffuses rapidly into the alveolar capillaries, this in turn results in an arterial paO2 in the region of 90 mmHg. So, the need to eliminate all CO2 produced during metabolism results in a paO2 well above the 60 mmHg (8 kPa) required to saturate hemoglobin.
You will notice one striking thing. Even though we said that oxygen diffuses rapidly across the alveolar capillary membrane, the pAO2 quoted above as ‘typical’ is slightly higher than the paO2 quoted as ‘typical’. In fact, the paO2 never reaches the level of the pAO2. There is a ‘gradient’ between the concentration of oxygen in the alveolus and the concentration of oxygen in the arterial blood, This is called the A-a gradient. It reflects physiological imperfections in lung function (physiological V/Q mismatch, physiological shunting etc). For any individual, there is a normal ‘predicted’ A-a gradient. The A-a gradient rises with age and varies with oxygen therapy. We will see in the cases we are about to analyse that the predicted normal A-a gradient can be calculated very simply for any individual.
You will also realise that what I have described here implies that there is a tight relationship between the paCO2 on a blood gas and the pAO2 in the patients alveoli (and by extension, a tight relationship between the paO2 and paCO2). This is the case and the relationship is described by the alveolar gas equation.
pAO2: alveolar partial pressure of oxygen, FiO2: percentage inhaled oxygen (as a decimal fraction), Patm: atmospheric pressure, PH2O; a correction for vapour pressure in the airways (43 mmHg), paCO2: arterial partial pressure of oxygen, RQ: respiratory quotient (usually 0.8). Note all in US units (mmHg).
So, it is possible to calculate the pAO2 using the results on the ABG and, therefore, we can calculate the A-a gradient (pAO2 – paO2). One of our tasks in this course is to show you how important this calculation is.
We hope to show you in the cases that
1. Disease may cause the paO2 to fall below the ‘normal level’ (what many refer to as ‘hypoxia’ in clinical practice). There are two broad categories of disease which can cause this to happen. I) a group of diseases associated with a rise in the A-a gradient. These include the classical pathologies we encounter day to day on the wards. They affect the ability of a region or regions of the lung to oxygenate blood in the alveolar capillaries (for example pulmonary embolism, pneumonia, fibrosis etc). II) The second group of conditions capable of causing hypoxia are caused by diseases which result in a global reduction in alveolar ventilation (the patient is not shifting enough air into the lungs). This manifests as a rise in paCO2. This is termed hypoventilation. In this situation the pAO2 is reduced resulting in a fall in paO2 in the presence of a normal A-a gradient.
These two categories of disease have very different differentials and very different treatment modalities. Both may be associated with a fall in the paO2 and in I) above this is associated with a normal or low paCO2 (a ‘type I pattern’ on the ABG), while in II) above, hypoventilation manifests as and is defined by a raised paCO2 on the blood gas (a ‘type II pattern’)
2. Disease affecting lung function may exist in the presence of an apparently ‘normal’ level of paO2. The body can compensate for disease in the lung by raising the pAO2 and disguise a real underlying problem. Or a low paO2 may be disguised by oxygen therapy. We only become aware of an underlying problem in these situations if we calculate the A-a gradient. We hope in one of the cases to show you how vital this can be. When analysing an ABG, in a sense we should not only look for a ‘normal’ or abnormal paO2, we should also ask, is the paO2 appropriate to the level of ventilation and the inhaled oxygen concentration. We do this by calculating the A-a gradient.
3. The transition from a type I pattern to a type II pattern on the ABG is a life-threatening event which should not be missed if possible.
*We say type I respiratory failure is present when the paO2 falls below the critical 60 mmHg (8 kPa) level required to saturate hemoglobin in association with a low or normal paCO2 on the ABG. Below this level hemoglobin saturation will fall precipitously and the total amount of oxygen in the arterial blood will fall significantly. We refer to a situation as type II respiratory failure when the paCO2 rises above 55 mmHg (7 kPa). Remember, CO2 is king. If the CO2 is elevated, it is a type II picture, no matter what the paO2 is doing. However, as we’ll see in Case 3, the underlying problem may be mixed in aetiology!
Acid Base Status
(Remember an acid lowers the pH of a solution. A base raises the pH of a solution)
The ABG also gives us information about acid base status in the body. This information is never diagnostic and it could be argued that most of the things we will talk about could be diagnosed and managed without an ABG!
For historical reasons acid-base disturbances are analysed by their ‘footprint’ on the arterial blood gas, that is the pattern of changes produced in the concentration of components of the bicarbonate buffer system.
The patten of changes in these components tells us something of the mechanism causing a pH change in the ECF. This can help us identify why the patient is ill and also help us decide just how ill they are!
Einstein said ‘makes things simple but not too simple’. Unfortunately, this is a luxury we cannot afford here. We have to simplify, some would say oversimplify things to get anywhere in this area. So, here goes.
The body tissues generate vast quantities of the gas CO2 during the normal process of metabolism. As shown below, CO2 in solution is a weak acid, it dissolves in water to yield carbonic acid which in turn yields nanomolar quantities of H+ ions (but remember when we talk about pH we are talking about nano molar quantities of H+ ions) lowering pH. CO2 is eliminated by the lungs and if lung function is normal, the rate of elimination of this gas can be adjusted.
(a word to the bewildered) note that in addition to hydrogen ions, this reaction also generates nanomoles of HCO3- (the CO2 antagonist in our story!). When you start to understand this stuff, it will greatly help you to realise that this particular equilibrium generates only minuscule amounts (nanomoles) of bicarbonate. Most of the milimolar amounts of bicarbonate present in the ECF are generated by other processes outside of this equilibrium. nano molar amounts of H+ ions count for a lot in terms of pH. nano molar quantities of bicarbonate influence very little.
The normal processes of metabolism in certain tissues also produce large quantities of bicarbonate which appears in the ECF. Bicarbonate is a base. It can react with H+ in solution reducing their numbers and elevating pH. Bicarbonate is eliminated by the kidney and, if kidney function is normal, its rate of elimination can be adjusted but this takes a little time.
So, think of the control of pH (nanomolar quantities of H+) in the ECF as a battle between a weak respiratory acid, CO2, and a weak metabolic base, (HCO3-, bicarbonate).
A change in ECF pH due to a rise or fall in respiratory acid (CO2 levels) is termed a primary respiratory pH disturbance. A rise in CO2 levels will acidify the ECF – a respiratory acidosis while a fall in CO2 levels will alkalise the ECF – a respiratory alkalosis.
A pH disturbance due to a rise or fall in the concentration of metabolic base (bicarbonate) is termed a primary metabolic pH disturbance. A fall in bicarbonate will make the ECF acidic-a metabolic acidosis while a rise in bicarbonate will make the ECF alkali-a metabolic alkalosis.
This simple view is complicated by the fact that when a disease causes a primary change in one side of this equation the body can compensate for the change by altering the side of the equation it still controls.
So in a primary metabolic acidosis (too little bicarbonate), the body will oppose the change in pH by getting rid of respiratory acid (CO2) through the lungs (respiratory compensation). In a primary metabolic alkalosis (too much bicarbonate) the body will compensate by retaining respiratory acid. This logic gives us our first two footprints on the ABG (see figure below).
The situation in a primary respiratory acidosis (too much CO2) is only slightly more complicated although exactly the same logic applies. The compensatory change in bicarbonate levels mediated by the kidney (renal compensation) opposing the CO2 induced change in ECF pH takes at least two days to develop. So, in primary respiratory pH disturbances we have four patterns of changes in the concentration of the components of the bicarbonate buffer system measured on the ABG as illustrated below.
You will notice that I have left out the pH on the above!. What happens to the pH when the primary pH disturbance and compensation are both in full swing. Do they cancel each other out. No, not usually. Compensation is usually partial that is compensation usually fails to return the pH to normal. This gives us the six patterns shown below.
In theory, the fact that compensation is partial is crucial in allowing us to decipher the nature of a pH disturbance on the ABG. It means that the direction of change of the pH is usually a true reflection of the underlying pathological process. So, for example, if I see a pH below the normal range on the ABG, I know that the primary problem is acidifying the ECF. I then look at the CO2 and bicarbonate. If it is a primary respiratory acidosis, by definition the CO2 will be raised. If it is a primary metabolic acidosis, by definition the bicarbonate will be low. I can then look for evidence of compensation.
It becomes slightly more difficult if the pH has drifted into the normal range. However, in reality, all of this is a little contrived. You will come to realise with experience that, in the interpretation of the ABG, the clinical scenario is everything, ‘given what I know about this patient, what do I expect to see on the ABG’. Usually, that is what you will see. Einstein will be spinning in his metaphorical grave by now but the way of thinking outlined here works well for beginners.
back to course menu