How to Calculate the Anion Gap (With Albumin Correction)
The anion gap is Na⁺ minus the sum of Cl⁻ and HCO₃⁻, a quick way to screen a basic metabolic panel for high anion gap metabolic acidosis. A normal result is roughly 3 to 11 mmol/L without potassium, or 7 to 16 mmol/L if the lab folds potassium into the formula. The number only tells the full story once it’s corrected for albumin, which is where most bedside calculations go wrong.
The formula and why potassium is optional
The standard equation:
Anion Gap = Na+ - (Cl- + HCO3-)Some labs add potassium, since it’s also a measured cation:
Anion Gap = (Na+ + K+) - (Cl- + HCO3-)Both versions are correct, they just use different reference ranges. Adding K+ pushes the normal band up by about 4 mmol/L (3 to 11 becomes roughly 7 to 16), because you’ve added a cation that’s usually around 4 mmol/L to the top of the equation without a matching change on the bottom. All values here are in mmol/L, which for these monovalent ions is numerically the same as mEq/L. Check which version your lab uses before comparing a result against “normal” ranges you’ve seen elsewhere; a 12 that looks mildly high on the no-potassium scale is unremarkable on the with-potassium scale.
Albumin is the piece that’s easy to skip and the one that matters most in sick patients. Albumin is the largest unmeasured anion in blood, so the raw anion gap formula implicitly assumes a normal albumin level. When albumin drops, as it does routinely in ICU patients, cirrhosis, and malnutrition, the raw gap reads lower than it should, sometimes low enough to mask a real acidosis. The correction:
Corrected AG = Measured AG + 2.5 x (4.0 - albumin in g/dL)For every 1 g/dL that albumin sits below 4.0, add 2.5 mmol/L back to the measured gap.
Two worked examples
The first case is the straightforward kind: an elevated gap that’s obvious even without correction, because albumin is normal.
Case 1. Na 140, Cl 100, HCO3 10 mmol/L, albumin 4.0 g/dL (normal).
Raw AG = 140 − (100 + 10) = 30 mmol/L. That’s far above the normal range of roughly 3 to 11, a clear high anion gap metabolic acidosis, consistent with something like diabetic ketoacidosis, lactic acidosis, or a toxic ingestion. Since albumin is already normal, no correction changes the picture.
The second case is the one that actually catches people out.
Case 2. Na 138, Cl 106, HCO3 20 mmol/L, albumin 1.5 g/dL (a critically ill or cirrhotic patient).
Raw AG = 138 − (106 + 20) = 12 mmol/L, which reads as only mildly elevated, easy to shrug off. Correct it for the low albumin: 12 + 2.5 × (4.0 − 1.5) = 12 + 6.25 = 18.25 mmol/L. Once corrected, the gap is clearly elevated, not borderline.
| Case | Na | Cl | HCO3 | Albumin | Raw AG | Corrected AG | Read |
|---|---|---|---|---|---|---|---|
| 1 | 140 | 100 | 10 | 4.0 g/dL | 30 | 30 (no change) | Clearly high, needs a cause |
| 2 | 138 | 106 | 20 | 1.5 g/dL | 12 | 18.25 | Looks near-normal raw, actually high once corrected |
Case 2 is the reason albumin correction matters in practice. Read the raw gap alone in a hypoalbuminemic patient and you’d file it as unremarkable. The corrected value says otherwise, and in a sick or cirrhotic patient, that’s exactly the population where a hidden acidosis is most likely and most consequential to miss.
Calculate it with your own values
Plug in sodium, chloride, and bicarbonate for the raw gap, add potassium if your lab reports it that way, and enter albumin whenever it’s below normal, an ICU chemistry panel or a cirrhotic patient’s labs being the usual case.
What causes an abnormal gap, and where the mistakes happen
A high anion gap points to an unmeasured anion accumulating in the blood. The mnemonic MUDPILES covers the usual causes: Methanol, Uremia, Diabetic ketoacidosis, Propylene glycol, Iron or Isoniazid, Lactic acidosis, Ethylene glycol, Salicylates. The gap itself isn’t a diagnosis, it’s a signal that one of these (or something similar) needs to be tracked down.
A normal gap doesn’t rule out metabolic acidosis. Normal-gap, or hyperchloremic, metabolic acidosis is a distinct and common category, seen with diarrhea, renal tubular acidosis (RTA), ileostomy losses, and carbonic anhydrase inhibitors. In these conditions bicarbonate falls but chloride rises to compensate, so the gap stays put even though the patient is acidotic. If metabolic acidosis is suspected on clinical grounds, check bicarbonate or a blood gas directly rather than assuming a normal gap means normal acid-base status.
A low anion gap has its own set of causes, and hypoalbuminemia is by far the most common one, since albumin is the biggest unmeasured anion in the equation. Beyond that: multiple myeloma or other paraproteinemias (an abnormal cationic protein), lithium toxicity, bromide intoxication, and lab or instrument error.
Four mistakes account for most of the confusion at the bedside:
- Skipping the albumin correction in low-albumin patients. ICU, cirrhosis, and malnutrition are exactly the settings where this matters most, and exactly where it gets skipped most often, as Case 2 above shows.
- Not checking whether the lab’s reference range includes potassium. Comparing a with-potassium result against a no-potassium range (or vice versa) makes a normal value look abnormal, or the reverse.
- Treating the anion gap as a diagnosis instead of a clue. A high gap tells you to go looking with MUDPILES in mind, it doesn’t name the cause by itself.
- Assuming a normal gap rules out metabolic acidosis. Normal-gap acidosis is real and common; a normal anion gap only rules out the high-gap subtype.
Frequently asked questions
What is a normal anion gap?
Roughly 3 to 11 mmol/L when potassium is excluded from the formula, or roughly 7 to 16 mmol/L when potassium is included. The exact cutoffs vary slightly by lab, mainly because of differences in how chloride is measured.
Why does low albumin affect the anion gap, and how do you correct for it?
Albumin is the largest unmeasured anion normally present in blood, so the standard formula assumes a normal albumin level. When albumin is low, the raw gap reads artificially low and can hide a genuine acidosis. The correction adds 2.5 mmol/L to the measured gap for every 1 g/dL that albumin sits below 4.0 g/dL: Corrected AG = Measured AG + 2.5 × (4.0 − albumin).
What causes a high anion gap?
An unmeasured anion building up in the blood. The mnemonic MUDPILES lists the classic causes: Methanol, Uremia, Diabetic ketoacidosis, Propylene glycol, Iron or Isoniazid, Lactic acidosis, Ethylene glycol, Salicylates.
Does a normal anion gap rule out metabolic acidosis?
No. Normal-gap, or hyperchloremic, metabolic acidosis is a real and distinct category, caused by things like diarrhea, renal tubular acidosis, ileostomy fluid losses, and carbonic anhydrase inhibitors. In these cases bicarbonate drops but chloride rises to match, keeping the gap in range despite true acidosis. If acidosis is suspected, check bicarbonate or a blood gas directly instead of relying on the gap alone.