The Corrected Anion Gap — Why Albumin Matters

What is the corrected anion gap?

The corrected anion gap (also called the albumin-adjusted or adjusted anion gap) is the standard anion gap recalculated to account for the patient's serum albumin. It exists because albumin is roughly 75% of the normal anion gap in a healthy adult — the remaining "unmeasured anions" that the formula Na⁺ − (Cl⁻ + HCO₃⁻) cannot see are dominated by this single negatively charged plasma protein.

Because albumin contributes so much to the baseline gap, any change in albumin concentration shifts the entire reference range. A patient with albumin of 2.0 g/dL has an expected normal gap roughly 5 mEq/L lower than a patient with albumin of 4.0 g/dL. Interpreting both against the same textbook range of 8–12 mEq/L is a category error — it makes sick, hypoalbuminemic patients look fine. For a deeper look at why this is, see the role of albumin in acid–base balance.

The Figge equation

The standard correction is the Figge equation, derived from Figge and colleagues' quantitative analysis of albumin's charge contribution at physiological pH (Figge et al., Critical Care Medicine):

Corrected AG = Observed AG + 2.5 × (4.0 − albumin in g/dL)

The reasoning is clinical: the textbook 8–12 mEq/L band is calibrated for a patient with a normal serum protein level, so once albumin drifts downward that band no longer applies. The Figge adjustment adds the missing protein charge back in, returning the gap to a value you can safely read against the conventional interval. A full mathematical derivation of the corrected anion gap formula is on the formula page, and the corrected anion gap calculator applies it automatically.

Why albumin correction is essential

To understand why albumin matters in anion gap interpretation, consider what the gap actually measures. The "normal" gap of 8–12 mEq/L is not empty space — it is made up of real anions. At a healthy serum concentration, the bulk of that gap is protein-bound negative charge carried by albumin; phosphate, sulfate, and small organic anions round it out. Strip the protein away and the floor of the reference interval falls with it.

When albumin drops, two things happen simultaneously in a critically ill patient. First, the baseline gap falls because there is less albumin to fill it — by itself this is harmless electroneutrality at work. Second, and dangerously, the loss of albumin from the anion pool creates headroom: the gap can rise by several mEq/L from accumulating lactate or ketones and still fall within the textbook "normal" range. This is how hypoalbuminemia masks HAGMA, delaying diagnosis of sepsis, ketoacidosis, or toxin ingestion in exactly the patients who can least afford a delay.

When to apply the albumin correction

Correct the anion gap whenever the serum albumin is abnormal, and especially in these settings where hypoalbuminemia is common:

• Intensive care units. Critical illness, inflammation, and fluid resuscitation drive albumin down within hours of admission. Most ICU patients warrant correction by default.
• Liver disease (cirrhosis). Reduced hepatic synthesis lowers albumin chronically; the uncorrected gap underestimates acid burden in hepatic failure and hepatorenal syndrome.
• Nephrotic syndrome. Massive urinary albumin losses produce a persistently low albumin and a falsely reassuring "normal" gap.
• Malnutrition and malignancy. Chronic illness and cachexia lower albumin; the corrected value is the only reliable screen for occult lactic or ketoacidosis.
• Any metabolic panel in a sick patient. If albumin is reported on the electrolyte panel and is below 3.5 g/dL, correct before interpreting.

Worked examples

Example 1 — masked HAGMA. A septic ICU patient: Na⁺ 138, Cl⁻ 104, HCO₃⁻ 22, albumin 1.8 g/dL. The observed anion gap is 138 − (104 + 22) = 12 — "normal." But the corrected gap is 12 + 2.5 × (4.0 − 1.8) = 12 + 5.5 = 17.5. The true gap is markedly elevated, consistent with occult lactic acidosis that the raw number completely concealed.

Example 2 — correctly normal. A stable cirrhotic outpatient: Na⁺ 136, Cl⁻ 103, HCO₃⁻ 24, albumin 2.5 g/dL. Observed gap 136 − (103 + 24) = 9 — low-normal. Corrected: 9 + 2.5 × (4.0 − 2.5) = 9 + 3.75 = 12.75. Slightly above the upper limit — worth investigating, not reassuringly normal as the raw value suggested.

Example 3 — hypercalcemia caveat. The Figge factor of 2.5 assumes normal calcium and magnesium. In severe hypercalcemia or paraproteinemias (multiple myeloma), additional cationic charges alter the baseline gap and the simple albumin correction under- or over-adjusts — interpret the corrected value with caution in those settings.

Adjusted anion gap interpretation

Once corrected, interpret the adjusted anion gap the same way as the standard anion gap, against the conventional 8–12 mEq/L range: 13–20 mEq/L is mildly elevated (early HAGMA), above 20 mEq/L is markedly elevated (overt HAGMA), and below 8 mEq/L remains genuinely low (now excluding the common artifact of hypoalbuminemia). The corrected value is what should drive the decision to calculate the delta ratio, check a lactate, or order an osmolar gap.

For the full list of electrolyte reference values that accompany the anion gap calculation, see the electrolyte panel ranges page.