Anion Gap Calculator — Serum Anion Gap & Delta Ratio
The anion gap is a calculated value derived from serum electrolytes that helps differentiate types of metabolic acidosis and identify underlying causes.
About
The serum anion gap (AG) represents the difference between measured cations (sodium) and measured anions (chloride and bicarbonate) in the serum. It reflects the concentration of unmeasured anions — normally albumin, phosphate, sulfate, and organic acids. The concept was first introduced by Gamble in the 1920s and later refined by Emmett and Narins in 1977 who established the differential diagnosis for elevated AG metabolic acidosis using the MUDPILES mnemonic. An elevated AG indicates the presence of pathological unmeasured anions such as lactate, ketones (beta-hydroxybutyrate, acetoacetate), toxic alcohols (methanol, ethylene glycol), uremic toxins, or salicylates. The albumin-corrected AG accounts for the effect of hypoalbuminemia, which reduces AG by approximately 2.5 mEq/L per 1 g/dL drop in albumin — a crucial adjustment in critically ill patients where low albumin is common and can mask a significant AG elevation. The delta ratio (ΔAG/ΔHCO3) helps identify mixed acid-base disorders by comparing the change in AG to the change in bicarbonate. A pure high AG metabolic acidosis yields a delta ratio of 0.8–1.0, while values outside this range suggest a concomitant metabolic alkalosis or normal AG acidosis. The AG remains a cornerstone of acid-base interpretation, used daily in ICUs and emergency departments worldwide. Evidence level: Grade B, supported by extensive physiological validation.
Formula
AG = Na - (Cl + HCO3) | Corrected AG = AG + 2.5 × (4 - Albumin) | Delta Ratio = (AG - 12) / (24 - HCO3)
The anion gap is calculated as AG = Na⁺ − (Cl⁻ + HCO₃⁻). All values are in mEq/L. A normal AG is 8–12 mEq/L, using older laboratories with different measurement methods — modern ion-selective electrode methods often yield a lower normal range of 3–10 mEq/L, so clinicians must know their laboratory's normal range. When albumin is low, the AG should be corrected: Corrected AG = Measured AG + 2.5 × (4 − albumin in g/dL). For example, a patient with Na 140, Cl 105, HCO₃ 15 has an AG of 140 − (105 + 15) = 20 mEq/L (elevated). If albumin is 2.0 g/dL: Corrected AG = 20 + 2.5 × (4 − 2) = 20 + 5 = 25 mEq/L. The delta ratio is calculated as (AG − 12) / (24 − HCO₃). It compares the excess AG (AG above normal) to the bicarbonate deficit (HCO₃ below normal). A ratio of 0.8–1.0 confirms a pure high AG metabolic acidosis — the rise in unmeasured anions fully accounts for the fall in HCO₃⁻. A ratio <0.4 suggests a normal AG acidosis (e.g., diarrhea, renal tubular acidosis) where the AG is not elevated but HCO₃⁻ is low. A ratio >1.0 suggests a concurrent metabolic alkalosis or pre-existing compensated respiratory acidosis. The delta ratio is less reliable when the initial HCO₃⁻ is very high or very low, or in chronic acid-base disorders. For interpretation, the AG should always be assessed in context of the patient's clinical presentation and the serum electrolyte pattern.
Score Interpretation
The anion gap is essential for acid-base interpretation in both emergency and critical care settings. The Kidney Disease: Improving Global Outcomes (KDIGO) guidelines emphasize the AG in the evaluation of acute kidney injury and metabolic acidosis. The Surviving Sepsis Campaign recommends calculating the AG as part of initial sepsis evaluation — an elevated AG with lactate elevation signals tissue hypoperfusion requiring urgent resuscitation. In diabetic ketoacidosis (DKA), the AG is a more sensitive marker of treatment response than blood glucose or beta-hydroxybutyrate levels — resolution of DKA is defined by a normalized AG (≤12 mEq/L). The American Diabetes Association recommends monitoring AG every 2–4 hours during DKA treatment. In toxicology, an elevated AG with an osmolal gap is the hallmark of methanol and ethylene glycol poisoning, guiding the use of fomepizole and hemodialysis. The delta ratio helps clinicians distinguish between pure and mixed acid-base disorders, which is crucial for appropriate management. For example, a patient with DKA who develops vomiting has a delta ratio >1.0 due to concurrent metabolic alkalosis — treating the alkalosis is as important as treating the DKA. A low AG (<8 mEq/L) may indicate hypoalbuminemia, but when persistent without explanation, it should prompt investigation for multiple myeloma, as paraproteins are positively charged and reduce the AG. The AG also aids in detecting laboratory errors — a negative AG is almost always artefactual. In clinical decision-making, the albumin-corrected AG is particularly important in the ICU where hypoalbuminemia is present in up to 70% of patients. Without correction, a significant AG elevation can be missed, leading to delayed diagnosis of conditions such as lactic acidosis or uremic acidosis.
Normal Anion Gap — 8–12
AG between 8-12 mEq/L. If albumin is low, use corrected AG.
Management: No acute intervention needed for the anion gap itself. Reassess based on clinical context.
Elevated Anion Gap — 12.1+
AG >12 mEq/L indicates unmeasured anions — evaluate using MUDPILES mnemonic.
Management: Investigate causes: Methanol, Uremia, DKA, Paraldehyde, Isoniazid/Iron, Lactic acidosis, Ethylene glycol, Salicylates/Sepsis. Check serum osmolality, lactate, ketones, renal function.
Low Anion Gap — 0–7.9
AG <8 mEq/L — less common. Consider hypoalbuminemia, paraproteinemia (multiple myeloma), or laboratory error.
Management: Check albumin, serum protein electrophoresis. Consider multiple myeloma if low AG persists.
Reference Ranges
| Population | Normal Range | Notes |
|---|---|---|
| Normal AG (standard) | 8 – 12 mEq/L | |
| Albumin-corrected AG | 8 – 12 mEq/L | Use if albumin <4 g/dL |
| Delta ratio interpretation | 0.4 – 2.0 | <0.4: Normal AG acidosis | 0.8-1.0: Pure AG acidosis | >1.0: Mixed disorder |
Dr. Omar Farouk
Dr. Omar Farouk is a board-certified internist with expertise in acid-base disorders and electrolyte management.
View medical review board & editorial policy →Example Calculation
A 28-year-old woman with type 1 diabetes presents to the emergency department with a 2-day history of polyuria, polydipsia, nausea, and vomiting. She has been unable to eat and missed two insulin doses. On examination, she appears unwell, with dry mucous membranes, tachycardia (110 bpm), Kussmaul breathing at 24 breaths/min, and a fruity odor on her breath. Her blood pressure is 105/70 mmHg. Laboratory studies show: Na 136 mEq/L, Cl 100 mEq/L, HCO₃ 10 mEq/L, glucose 480 mg/dL, albumin 4.5 g/dL, lactate 1.2 mmol/L, and β-hydroxybutyrate 6.8 mmol/L. Step 1 — Calculate AG: AG = 136 − (100 + 10) = 26 mEq/L (normal 8–12). This is elevated, consistent with unmeasured anions (ketones). Step 2 — Correct for albumin: Albumin is normal at 4.5 g/dL, so corrected AG = 26 + 2.5 × (4 − 4.5) = 26 − 1.25 = 24.75 mEq/L. The correction is minimal since albumin is normal. Step 3 — Calculate delta ratio: (AG − 12) / (24 − HCO₃) = (26 − 12) / (24 − 10) = 14/14 = 1.0. This confirms a pure high AG metabolic acidosis. Step 4 — Interpretation: The elevated AG of 26 with a delta ratio of 1.0 indicates pure high AG metabolic acidosis from DKA. The normal lactate and absent clinical history of toxic alcohol ingestion support DKA as the sole cause. Step 5 — Management: The patient is admitted for IV fluids (0.9% saline at 15 mL/kg), IV insulin infusion at 0.1 units/kg/hr, and close monitoring of electrolytes, glucose, and AG every 2 hours. The target is to reduce the AG to ≤12 mEq/L, which signals resolution of ketosis. Potassium replacement is anticipated as insulin drives potassium intracellularly. The patient is diagnosed with DKA precipitated by missed insulin doses, with a pure high AG metabolic acidosis confirmed by the delta ratio.
Related Conditions
Related Medications
Common Mistakes
Interpreting AG without correcting for albumin
For every 1 g/dL drop in albumin below 4 g/dL, add 2.5 to the calculated AG. Low albumin is common in hospitalized patients.
Not calculating delta ratio in elevated AG acidosis
Always calculate the delta ratio when AG is elevated to detect mixed acid-base disorders. A ratio outside 0.8-1.0 suggests a concurrent metabolic alkalosis or normal AG acidosis.
Failing to consider lab error in unexpected AG values
Very low AG (<3 mEq/L) or negative AG is often due to lab error, lipemia, or bromide interference. Repeat electrolytes and check for pre-analytical issues.
Using delta ratio in chronic metabolic alkalosis
The delta ratio assumes 24 mEq/L as normal HCO₃⁻. In chronic disorders, the "normal" HCO₃⁻ may be different. Use the Winters formula or a full blood gas analysis with expected compensation.
Ignoring the effect of hyperglycemia on AG
Hyperglycemia increases serum osmolality and can slightly lower measured sodium, which indirectly affects the AG calculation. Always correct sodium for glucose when interpreting AG in hyperglycemic patients.
Frequently Asked Questions
What is the MUDPILES mnemonic for elevated AG?
Can the anion gap be negative?
How do I interpret the delta ratio?
When should I use albumin-corrected AG?
What is a normal AG in newborns?
Can AG be normal in DKA?
How does AG differ between modern and older lab methods?
References
- Gabow PA, Kaehny WD, Fennessey PV, et al. Diagnostic importance of an increased serum anion gap. N Engl J Med. 1980;303(15):854-858. PubMed
- Kraut JA, Madias NE. Serum anion gap: its uses and limitations in clinical medicine. Clin J Am Soc Nephrol. 2007;2(1):162-174. PubMed
- Figge J, Jabor A, Kazda A, Fencl V. Anion gap and hypoalbuminemia. Crit Care Med. 1998;26(11):1807-1810. PubMed
- Emmett M, Narins RG. Clinical use of the anion gap. Medicine (Baltimore). 1977;56(1):38-54. PubMed
- Kamel KS, Halperin ML. An approach to the patient with metabolic acidosis. Nephrol Dial Transplant. 2021;36(9):1579-1587. PubMed
- Kidney Disease: Improving Global Outcomes (KDIGO). Clinical Practice Guideline for Acute Kidney Injury. Kidney Int Suppl. 2012;2(1):1-138.
- Lolekha PH, Vanavanan S, Lolekha S. Update on value of the anion gap in clinical diagnosis. Clin Chim Acta. 2001;307(1-2):33-36. PubMed