Arterial Blood Gases in a Four-Step Reading Order

An arterial blood gas is four questions asked in a fixed order, and the order is the part people skip. Ask them out of sequence and a compensated disorder looks normal. Ask them in sequence and most gases resolve in under a minute, including the ones built to look like traps.

The sequence, in the order the source gives it

StatPearls sets out gas interpretation as four steps. Read the pH and decide acidemia or alkalemia. Compare the carbon dioxide and the bicarbonate against that pH to decide whether the problem is respiratory or metabolic. Check whether the second value has moved in a direction that does not fit the pH, which means compensation. Then assess oxygenation separately.

That last word carries weight. Oxygenation is assessed separately because it answers a different physiological question, and folding it into the acid base reasoning is how a hypoxic patient with a normal pH gets called fine.

If you want the compressed version for a quick review, reading ABGs in three steps is the short form. This post is the fuller, sourced one.

Step one: the pH decides which side you are on

The published normal arterial pH runs between 7.35 and 7.45, averaging 7.40. Below 7.35 is acidemia. Above 7.45 is alkalemia. The pH is a logarithmic scale and carries no units.

Read it first, every time, because everything after this is a question about who caused the pH and who is trying to fix it. Without a side, those questions have nothing to attach to.

A normal pH does not mean a normal gas, which is the most common misread in the whole topic. Step one: does the pH say acidemia or alkalemia takes that apart properly.

Step two: carbon dioxide or bicarbonate

Now compare the two candidate drivers against the pH. The published PaCO2 range is 35 to 45 mmHg, and the published bicarbonate on a gas is 22 to 26 mEq/L.

Whichever of those has moved in the direction that would produce the pH in front of you is the primary driver. If the carbon dioxide is doing it, the disorder is respiratory. If the bicarbonate is doing it, the disorder is metabolic.

A units trap sits right here. The bicarbonate on a gas and the carbon dioxide row on a basic metabolic panel are different rows in different units, and bicarbonate on a BMP versus HCO3 on an ABG keeps them apart. The full step is in step two: deciding respiratory versus metabolic.

Step three: is anything compensating

Compensation shows up as the second value moving in a direction that does not fit the pH. A body correcting an acidemia pushes the other value the opposite way, and that mismatch is the only signal you need.

This is where a normal pH stops being reassuring. A fully compensated disorder can land the pH back inside its range while both of the other values sit well outside theirs. Step three: spotting compensation in the numbers shows what that looks like on real values.

Step four: oxygenation is its own question

The published PaO2 range is 75 to 100 mmHg, and oxygen saturation is 95 to 100%. Some sources round the floor upward and print 80 to 100 mmHg instead, which is a rounding convention rather than a disagreement.

Oxygenation does not tell you the acid base disorder, and the acid base disorder does not tell you the oxygenation. They are read side by side. Sometimes the oxygen value is the urgent one and the pH can wait, and step four: oxygenation is a separate question works through when that happens.

Every published range on one screen

Two caveats belong with that list. Reference ranges vary slightly between laboratories and across age groups, which the source says in its own words. And base excess deserves more than a row, because it and the four physiologic buffer systems explain why the metabolic side behaves as it does, which is what base excess and the four buffer systems is for.

A worked gas

Take a set of values: pH 7.28, PaCO2 58 mmHg, bicarbonate 25 mEq/L, PaO2 72 mmHg, saturation 92%.

Step one. The pH is below 7.35, so this blood is acidemic.

Step two. The carbon dioxide is above its range and would push the pH down, so it is the driver and the disorder is respiratory. The bicarbonate sits inside its published range, so it is not the cause of anything.

Step three. The bicarbonate has not moved to help, so nothing is compensating yet.

Step four. The PaO2 is below its published range and the saturation is below its own, so oxygenation is inadequate as well.

One sentence covers it: respiratory acidosis, uncompensated, with inadequate oxygenation. No acronym was involved, and the same four questions would have worked on any other set of values.

Where the practical nursing track picks this up

PN items usually stop short of full interpretation and sit closer to recognising, reporting and monitoring, which is a different task rather than a smaller one. What PN candidates need from an ABG question sets out where those items stop.

The bedside findings that appear before any gas comes back are worth knowing on both tracks, and they are collected in acid base changes you can see at the bedside.

Why a sequence rather than an acronym

Plenty of acronyms circulate for this topic, and some of them work. The reason we publish the sequence instead is simple. The sequence sits in a source you can open and read, and the acronyms do not.

That matters more than it sounds. When two study sources disagree about a mnemonic you have no way to settle it. When they disagree about a sequence taken from a published article, you can go and look.

There is a second reason. An acronym hands you a label, and the four step order hands you a method that still works on an unusual gas, which is precisely where labels fail.

What this post deliberately leaves out

Causes are not in here. Naming which conditions produce which pattern, and what those patients look like at the bedside, belongs to the disorder posts rather than the method posts, starting with respiratory acidosis and what the patient looks like and metabolic acidosis causes you will meet in a stem.

Keeping method and causes apart is deliberate. A method post that starts listing causes becomes a memory test, and the point of a sequence is that it works on a gas you have never seen.

Practising it

Take any set of gas values and say the four answers out loud in order. Acidemia or alkalemia. Respiratory or metabolic. Compensating or not. Oxygenating or not.

Four short answers. That is the entire interpretation, and everything else in this pillar hangs off one of the four.