Anion gap

Calculate the serum anion gap, with optional correction for hypoalbuminaemia.

What is this for?

The anion gap is the difference between measured cations and measured anions, representing the unmeasured anions in serum. It is the pivotal branch point in evaluating a metabolic acidosis: a raised gap points to added acid, a normal gap to bicarbonate loss or chloride gain.

How to use it

Enter sodium, chloride, and bicarbonate from the same basic metabolic panel.
Add the albumin if it is available — in hospitalised patients it very often is, and the correction matters.
Use the differential table below the result to narrow the cause once you know which category the gap falls into.

Worked example

Sodium 140, chloride 100, bicarbonate 24 mEq/L. Calculate the anion gap.

Answer: 140 − 100 − 24 = 16 mEq/L, which is above the usual range and indicates a high anion gap acidosis warranting a search for an added acid.

Clinical pearls & pitfalls

Correct for albumin whenever it is low. A patient with an albumin of 2 g/dL has a baseline gap about 5 mEq/L lower than normal, so a "normal" gap of 11 is really 16 — an elevated gap that would otherwise be missed.
Potassium is excluded from the common formula. Some laboratories include it, which raises the reference range by about 4 mEq/L. Know which convention your institution uses.
The reference range has drifted downward over recent decades because modern ion-selective electrodes measure chloride differently. Ranges quoted in older textbooks as 8–16 are too high for current assays.
In a high-gap acidosis, compare the rise in the gap with the fall in bicarbonate (the delta-delta). If they do not match, a second metabolic process is present alongside.

Assumptions & limitations

Represents unmeasured anions in aggregate; it does not identify which anion is responsible.
The reference range is laboratory- and method-specific. Use your own laboratory's range in preference to any published one.
Falsely lowered by hypoalbuminaemia, hypercalcaemia, hypermagnesaemia, lithium, and paraproteins; falsely raised by severe dehydration and some assay interferences.
The albumin correction factor of 2.5 mEq/L per g/dL is itself an approximation from a small derivation study.

References

Figge J, Jabor A, Kazda A, Fencl V. Anion gap and hypoalbuminemia. Crit Care Med. 1998;26(11):1807-1810.
Kraut JA, Madias NE. Serum anion gap: its uses and limitations in clinical medicine. Clin J Am Soc Nephrol. 2007;2(1):162-174.
Berend K, de Vries AP, Gans RO. Physiological approach to assessment of acid–base disturbances. N Engl J Med. 2014;371(15):1434-1445.