Arterial blood gases made easy

Key points

• Most doctors struggle with arterial blood gas (ABG) interpretation

• ABG interpretation is easy

• Break it down into steps

• The first priority for the respiratory system is pH

• If partial pressure of carbon dioxide (pCO2) goes down, partial pressure of oxygen (pO2) should go up

Mistakes in arterial blood gas (ABG) interpretation are common in clinical practice. The following is a simplified explanation of ABGs, including a practical method for interpreting results. It is simple, perhaps simplistic, but it will hopefully arm the reader with the tools (and confidence) to make better sense of ABG results in future. This is not for the dedicated physiologist. Overly complex explanations can be a barrier to a working understanding of the basics.

On acute medical take, ABGs are used to determine the nature as well as the severity of a problem. The results often have a direct bearing on management. Yet most doctors struggle with interpretation of this common test.

Arterial blood gas measurements

Unlike other ‘blood tests’, which are either ‘high or low’, ABGs present the doctor with six numbers that need to be interpreted as ‘one result’. Given that this can be difficult, there is a need for a simple algorithm for systematically handling each of the numbers in turn, as discussed below. However, I begin with a few basic points to understand.

Respiratory and metabolic systems – in balance

ABGs tell us about activity in two systems; the respiratory system and the ‘metabolic’ system. If one system is disturbed, the other tries to restore balance. Both systems are primarily concerned with keeping blood pH in the normal range. Even for the respiratory system, pH (rather than oxygen) is the priority.

The respiratory system – oxygenation vs pH

In health, we are driven to take our next breath by the arterial partial pressure of carbon dioxide (P_aCO₂), which is intimately linked to pH. Although there is an additional receptor (the hypoxic centre) in the brain stem that monitors P_aO₂, it spends most of it time ‘asleep’ and is rather unconcerned about minor fluctuations in the level of oxygenation. This is because individuals generally live at a level of oxygenation well above that which is required to sustain life. This ‘margin of oxygen safety’ enables the respiratory system to focus on pH and to adjust ventilation (to ‘blow off’ or retain CO₂) without the fall in oxygenation that underventilation would bring causing any difficulties. Therefore, if, for example, a metabolic alkalosis were to develop, ventilation would fall (at the expense of a small reduction in oxygenation) to retain CO₂ and, thus, return pH to the normal range. Only when hypoxia is more severe (approximately P_aO₂ <8 kPa) does the hypoxic centre ‘wake up’ and take note. Only then, will it drive ventilation to prevent harmful levels of hypoxia.

Respiratory and metabolic systems – the speed of response

The respiratory system can respond quickly to a metabolic derangement, with changes occurring to the blood gases within seconds to minutes. However, the metabolic system (largely regulated by the kidneys excreting or retaining acid or bicarbonate) is much slower and changes can take hours to days.

A step-by-step method for interpreting arterial blood gases

First – look at the pH

Decide whether this is an ‘acidosis’ or ‘alkalosis’ (if it is within the normal range, note whether it is sitting towards the ‘acidotic’ or ‘alkalotic’ end of that range). Fix that fact in your mind because it will not change, no matter what the other numbers are!

Second – look at the P_aCO₂

Ask the question: is the P_aCO₂ contributing to, or attempting to compensate for, the problem. If, for example, the problem is an acidosis and the P_aCO₂ is low, then clearly the respiratory system is attempting to compensate. Thus, one can conclude that the problem is metabolic (similarly with other combinations). Therefore, after looking at only two numbers (pH and P_aCO₂), most of the interpretation is done.

The other numbers (actual bicarbonate [aHCO₃], base excess [BE], P_aO₂ and so on) might do nothing more than confirm this conclusion. However, they can sometimes add information about time course or provide information on additional derangements, but they will not contradict the conclusion that has already been reached.

Third – look at the ‘base picture’

Actual bicarbonate (aHCO₃) vs standard bicarbonate (sHCO₃) – what's the difference? What is perhaps surprising is that, after many years of looking at ABGs, those intelligent, enquiring minds have seemingly never once pondered that question.

aHCO₃ is the actual measurement of bicarbonate in that actual blood sample (hence the name). The problem with this measurement is that it is markedly affected by P_aCO₂. If the P_aCO₂ is high, the aHCO₃ is dragged higher and vice versa. What one would like to know is what the HCO₃ would have been had the P_aCO₂ been normal. It is this value that would provide a direct handle on what the metabolic system is doing. One can calculate the value if aHCO₃ and P_aCO₂ are known, although most blood gas machines do this automatically, known as the sHCO₃.

What is the base excess?

Base excess (BE) measures all bases, not just bicarbonate. However, because bicarbonate is the greater part of the base buffer, for most practical interpretations, BE provides essentially the same information as bicarbonate. The major advantage of BE is that its normal range is really easy to remember. One could probably have guessed that the expected value of BE was zero (the clue is in the word: ‘excess’). Therefore, a tight range around zero (−3 to +3) is normal. In simple terms, a high BE excess is the same as a high HCO₃.

What does the base picture tell us?

If the pH and P_aCO₂ led to the conclusion that the problem was primarily metabolic, then sHCO₃ (or BE) will do little more than confirm that; sHCO₃ being high in an alkalosis, low in an acidosis. If one has established that problem is respiratory, then the BE can tell us something of the duration of the problem.

If, for example, in a respiratory acidosis, the sHCO₃ has shown no sign of responding (still within the normal range), the probable explanation is that there has not yet been time to respond (ie the problem is an acute respiratory acidosis). In a respiratory ‘acidosis’ (perhaps with the pH in the lower half of the normal range), a high sHCO₃ would indicate a longer time course (ie the problem is a chronic -respiratory acidosis). A respiratory acidosis with a low sHCO₃ would indicate a combined respiratory and metabolic -acidosis.

Box 1. Example.

Remember that one cannot live for long with pH outside of the normal range. If the pH is outside the normal range, one should never fall into the trap of assuming the problem is ‘probably all chronic’ (no matter how high the bicarbonate). An abnormal pH means there has to be an acute component to the problem.

Finally, look at the oxygen

It is sometimes thought that type 2 respiratory failure is simply a more severe version of type 1. However, this is not the case. Type 1 and type 2 respiratory failures are due to entirely different mechanisms.

Type 2 respiratory failure is extremely an issue of ventilation, that is, the business of pumping air in and out of the lungs. When underventilation occurs, for what ever reason (eg muscular weakness or opiate overdose), the P_aCO₂ will increase (the definition of underventilation) and P_aO2 must decrease (even if the lungs are perfectly healthy). Type 2 respiratory failure results from underventilation, which can occur even in the context of healthy lungs.

Lung (or pulmonary vascular) disease disturbs the delicate ventilation–perfusion (V/Q) matching system. In such circumstances, oxygen delivered to the lungs by ventilation is handled inefficiently and P_aO₂ falls. However, provided that overall ventilation is normal, P_aCO₂ is maintained. When P_aO₂ is low yet P_aCO₂ normal, type 1 respiratory failure is present, and such a result implies lung (or pulmonary -vascular) disease.

Type 1 and type 2 respiratory failure can occur simultaneously. Indeed, the combination is common in severe chronic obstructive pulmonary disease, for example. Given that the two conditions result from entirely different mechanisms, with implications for treatment, one should be able to distinguish between them. When the only derangement is P_aO₂, clearly the failure is type 1. However, when the P_aCO₂ is high, one has to work out whether the low P_aO₂ can be accounted for by underventilation alone or whether there is an additional type 1 problem (ie whether there is anything wrong with the lungs).

To do this, one needs to measure the alveolar–arterial gradient, that is, the difference between the alveolar partial pressure of oxygen (P_AO₂) and the P_aO₂. The P_aO₂ is measured in the ABG, the P_AO2 has to be calculated using the alveolar gas equation:

where P_IO₂ is the partial pressure of oxygen in the inspired air (approximately 21 kPa when breathing room air, but 24 kPa when using a 24% Venturi mask and so on) and 0.8 is the ‘respiratory quotient’ (ie the ratio between the CO₂ produced and the O₂ utilised).

The alveolar–arterial gradient (P_AO₂–P_aO₂) can then be calculated. In healthy young adults, the difference should be less than 2 kPa. If the patient is older, breathing higher concentrations of O₂ or over ventilating, then the gap can widen, although in healthy patients this would not usually be expected to be greater than 4 kPa. If the alveolar–arterial gradient is higher than it should be, then a type 1 respiratory failure is present. This implies a problem with V/Q matching (ie a problem with either the lungs or the pulmonary vasculature).

Remember!

ABG interpretation is not difficult. Break down the task into steps and do them in order.

For a more detailed review of arterial blood gas interpretation, see Ref 1. Box 1 provides an example of a patient presenting with breathlessness, where ABGs form an important diagnostic test.