Blood gas basicsAcidaemia or alkalaemiaAnalysing an ABG


An analysis of arterial blood gases can tell you a lot about what is going on inside a patient. Homeostatic mechanisms in the body try very hard to keep pH, PaO2 and PaCO2 at constant, physiological levels. The lungs and the kidney work collaboratively to ensure that this is the case in the short and long term, respectively. When pH, gases and bicarbonate ions (HCO3-) diverge from their usual ranges, the pattern of changes hint at what is wrong, and what the body is trying to do about it.

Blood gas basics

From the outset it is worth highlighting the normal ranges for PaO2 and PaCO2, and what is considered abnormal:
hypercapnia hypocapnia

There are four mechanisms that can lead to hypoxaemia (and possibly CO2 disturbances):


Note that for three mechanisms above, the effect on PaCO2 is “Normal (or low)”. This is because when PaO2 is low, a common response by the body is to increase ventilation, sometimes excessively so (see hyperventilation below). By contrast in hypoventilation, CO2 will accumulate to higher than normal levels. (For details of why the body is sometimes forced into hyperventilation, see the V̇A:Q̇c mismatch entry for a full working of this problem.) Alveolar diffusion defects occur when the diffusion barrier for gas exchange is greater than normal, such as in the disease alveolar proteinosis. Such problems are quite rare. V̇A:Q̇c mismatch is a component of many respiratory diseases, and shunt is a severe mismatch. A convenient definition of shunt is that while the V̇A:Q̇c mismatch can be overcome to some extent by putting a patient on supplemental O2, this is not the case in shunt.

There’s only a single way that hypocapnia evolves: when you ventilate excessively (beyond the metabolic needs of the body to eliminate CO2.


As we’ve discussed elesewhere (and it can never be emphasised enough), the adequacy of ventilation is determined by PaCO2, not PaO2.

Acidaemia or alkalaemia

The regulation of blood pH is a business that is run collaboratively by the lungs and the kidneys. They have quite distinct roles. It all comes down to the equation central to CO2 transport in blood:


The equation above describes an equilibrium. The reaction could occur in either direction depending on the amounts of the reaction components present. If you add CO2, you will drive the equation to the right, producing H+ (acidaemia). If you blow too much CO2 away on the other side of the equation you lower the H+ in the body (alkalaemia). This is what goes on in the body depending on how you stress it.

Say you are late and run to catch a train. Your body will call on anaerobic metabolism to produce some of the energy required, and lactic acid will be produced. We call this a metabolic acidosis, since it is the body's fault that acid accummulated. This adds H+ to the right hand side of the equation, driving it to the left and producing CO2. Your lungs will blow this off as you pant and recover on the train. The minor pH disturbance is rapidly resolved with a few deep breaths. Lungs are great at solving little problems like that.

When the lungs aren’t working well, the kidneys can help by changing the amount of HCO3- they excrete into urine. This is compensation. If you had a minor pulmonary embolism (say as the result of a deep vein thrombosis that dislodged from your leg and got stuck in a branch of the pulmonary artery) your lungs might blow off the CO2 excessively trying to raise PaO2 (hyperventilation). The result would be alkalaemia. If it was small, you might not ever know you had the embolism because the kidneys can rectify the alkalaemia by excreting HCO3-, pulling the equation to the right to produce acid (H+). A quick arterial blood gas analysis would tell the story though: your PaCO2 level would be low, your HCO3- level would also be low but your blood pH would be spot on. The alkalosis is respiratory in origin and compensated for by the kidneys.

The simple lesson here is that pH isn't determined by the absolute values of CO2 and HCO3-, but their ratio.

Interpreting the results of an arterial blood gas analysis

Step 1: Identify the primary disturbance

The first task is to identify the primary disturbance and work out if it is respiratory (PaCO2) or metabolic (HCO3-) in origin. Find the disturbance that explains the pH change. Don’t try to juggle the various values and their ups and downs to do this. Look at the pH and make a snap decision: is it an alkalosis or acidosis (only the pH can tell you that – ignore everything else) and who is to blame, the lungs or the body?

Figure 1: The initial approach to analysing arterial blood gas (ABG) results. Firstly, establish if pH is outside the normal range of around 7.35-7.45. If it is, identify the culprit: find the change in CO2 (acid, which lowers pH when in excess) or HCO3- (base, which raises pH when in excess) that can explain the change in pH. Ignore everything else for now. We can work out other - possibly confusing readings –once we’ve decided what the primary problem is.

Causes of respiratory acidosis and alkalosis are hypoventilation and hypoventilation, respectively. Causes of metabolic disturbances are more varied.

Let’s try this simple algorithm on an example ABG.

A 15 year old male has undergone reconstructive surgery to realign his previously fractured clavicle. His post-surgical pain relief includes morphine, delivered by a patient-controlled analgesia pump. His mother alerts the medical team that he has become drowsy and incoherent. An arterial blood gas analysis reveals the following:

His arterial blood pH is low, which is acidic. What is the source of the imbalance? PaCO2 is high (this fits the picture) while HCO3- is normal. So, this is a respiratory acidosis: he’s hypoventilating. Respiratory depression is a well-known side-effect of opioid drugs, like morphine.

Step 2: look for compensation

Once you’ve identified the culprit (metabolism versus ventilation) that is causing the pH disturbance, the other numbers will either look normal (as in the example above), or may be completely contradictory. Don’t discard your initial conclusion about the primary disturbance, though! Odd values are probably some kind of compensation by the kidney for the lungs or the lungs for the kidneys. It’s that or the ABG analyser is on the blink.

If chronic PaCO2 is the primary disturbance, the kidneys will slowly compensate over hours to days by causing a similar change in HCO3- . In an acute condition, there may be no time for this renal compensation because this process involves synthesis of new enzymes. However, respiratory compensation for metabolic problems will be near-instantaneous, since the acid-base regulation by the lungs is performed on a breath-by-breath basis. You can produce a respiratory acidosis by holding your breath, but you’d asphyxiate long before the kidneys could do anything about it!

Rather than making a huge flowchart considering all the possibilities (there’s one here, if you really want one to run over), let’s just think about respiratory and metabolic disturbances in simple terms:

Figure 2: Figure 2: Analysing most AGB readings is a process of identifying the primary disturbance (respiratory or metabolic) and then looking for compensation (if there is any). You don’t really need a complex flow chart to perform an analysis, once you understand it.

Once you understand the possibilities and their effects on pH, you can reduce all this to a simple two-step approach that works most of the time:

  • Step 1: Establish which abnormal reading matches the observed disturbance to pH (alkalosis or acidosis): PaCO2 (respiratory) or HCO3- (metabolic)? You’ve got the primary disturbance. Hold firm.
  • Step 2: Any changes that now look completely contrary to what you decided in step 1 are probably compensation. Check that the numbers make sense. After all, blood gas analysers can be misused or poorly calibrated.

Where this may fall down is in a situation where compensation has returned pH to the normal range. This is why it’s important to look at the bigger picture and not follow a flowchart approach that might stop you at the normal-looking pH and not spot the compensation system keeping pH in check.

Let’s look at some more complex examples and apply this simple procedure.

As part of her routine care in HDU, a 72-year-old patient with pneumonia has an ABG analysis from her arterial line every 4 hours. Her most recent analysis shows:

  • Step 1: pH is high = alkalosis. The PaCO2 is low, which matches this well. This is a respiratory alkalosis.
  • Step 2: There isn’t any renal compensation, although the HCO3- is on the lowest side of the normal range, suggesting that compensation may not be far away. The medical team will be hoping that the next ABG is slightly less interesting.

This patient is also hypoxaemic, which doesn’t sit so obviously with hyperventilation, where we might expect good oxygenation of the blood. Old hands will recognise this common pattern which arises from ventilation:perfusion (V:Q) mismatch. You can read up on this common finding in respiratory disease here. Emily, who is suffering from a serious asthma exacerbation, is a good example of VQ mismatch leading to hypoxaemia and hyperventilation.

A 16-year-old happy-go-lucky Type 1 diabetic, who inadvertently left her insulin at home, collapses at the end of a school sports day and is brought to A&E. Her ABG values are:

  • Step 1: Low pH is acidosis. PaCO2 is low, which can’t explain the pH disturbance. However, her HCO3- is low, and that does. This is a metabolic acidosis. In this scenario it almost certainly diabetic ketoacidosis.
  • Step 2: Here we see some significant respiratory compensation. Without this hyperventilation, pH would be even lower. Catherine is a typical case of diabetic ketoacidosis with respiratory compensation in the form of Kussmaul breathing.

A 19-year-old woman is brought to A&E by ambulance having had a severe asthma exacerbation at a party, and did not responded to the rescue inhaler (salbutamol) that she always carries. Friends called an ambulance when she couldn’t speak properly without gasping for breath. She is now moribund. On admission, her ABG showed:

  • Step 1: Respiratory acidosis. Only the high CO2 can explain the low pH.
  • Step 2: No renal compensation. This is fits with the history of an acute respiratory crisis.

This patient is in type II respiratory failure: she’s can’t get enough O2 into her blood and she can’t get the CO2 out (type I respiratory failure is just hypoxia). This is the worst case scenario. Earlier, as her exacerbation evolved, she would probably have had a respiratory alkalosis arising from breathing off too much CO2 – a bit like the pneumonia patient above. By the time she’s made it to A&E, though, her bronchospasm has limited airflow so globally in the lung that she can’t even blow off excess CO2 and she’s retaining it.

A 69-year-old male patient is brought to A&E by his wife and daughter with an exacerbation of chronic obstructive pulmonary disease. He is too breathless to talk. His wife confirms that he has been increasingly unwell for several days, and has had similar exacerbations at this cold time of year over the last 5 years. His ABG results are:

  • Step 1: His arterial pH is borderline acidosis. His elevated PaCO2 fits the picture well, but is severe enough that we should expect a more profound acidosis.
  • Step 2: His HCO3- is elevated. This renal compensation is almost certainly why his arterial pH isn’t more profoundly disturbed than it should be, considering the CO2 he is retaining.

This gentleman also has profound hypoxaemia. He’s in type II respiratory failure (like the asthma patient above), but his kidneys have been compensating for the last few days, keeping his arterial pH on the low side of normal. These aren’t unusual ABG readings for a patient with COPD. Mr Khan is an interesting case of the respiratory deterioration and renal compensation seen in this disease.