Stewart Acid-Base Calculator
Physicochemical approach to acid-base analysis — calculates apparent strong ion difference (SIDa), effective strong ion difference (SIDe), strong ion gap (SIG), and non-volatile weak acid contributions for a comprehensive quantitative assessment of acid-base status in critically ill patients.
Calculate Stewart Acid-Base Parameters
Enter electrolytes from a basic metabolic panel, arterial blood gas values, and serum protein levels to compute the apparent and effective strong ion differences and the strong ion gap. This calculator is intended for use in adult patients in the critical care setting.
The Stewart approach complements rather than replaces traditional acid-base analysis. It is most useful when conventional bicarbonate-based methods fail to explain the full acid-base picture, particularly in critically ill patients with hypoalbuminaemia, complex fluid therapy, or multi-organ dysfunction.
Understanding the Stewart Approach
Traditional acid-base analysis centres on the Henderson-Hasselbalch equation, treating bicarbonate as the key variable. The Stewart (physicochemical) approach, first described by Peter Stewart in 1983, takes a fundamentally different view: it identifies three independent variables that determine pH, while bicarbonate and hydrogen ion concentration are dependent variables that change only as a consequence of the independent ones.
The three independent determinants of pH in the Stewart model are the strong ion difference (SID), the total concentration of non-volatile weak acids (ATOT), and the partial pressure of carbon dioxide (PCO₂). A change in any of these three variables will alter the dissociation equilibria of water and weak acids, thereby changing pH and bicarbonate.
Apparent SID (SIDa)
SIDa = [Na⁺] + [K⁺] + [Ca²⁺ × 2] + [Mg²⁺] − [Cl⁻] − [Lactate]
Example: Na 140 + K 4 + iCa (1.15 × 2) + Mg 2 − Cl 102 − Lac 1.0 = 45.3 mEq/L
Normal range: ~38–42 mEq/L. Ionised Ca²⁺ is entered in mmol/L and multiplied by 2 (divalent cation) to convert to mEq/L. Mg²⁺ is entered directly in mEq/L.
Effective SID (SIDe)
SIDe = [HCO₃⁻] + [A⁻alb] + [A⁻phos]
Albumin charge: Alb (g/dL) × (1.2 × pH − 6.15)
Phosphate charge: PO₄ (mg/dL) × (0.097 × pH − 0.13)
Uses the Figge-Mydosh-Fencl equations to estimate the anionic charge contributed by weak acids at the measured pH.
The Strong Ion Gap (SIG) = SIDa − SIDe. In health, SIG is approximately zero (±2 mEq/L). An elevated SIG indicates the presence of unmeasured strong anions — conceptually similar to an elevated anion gap, but corrected for albumin and phosphate by design, making it more sensitive in hypoalbuminaemic patients.
The physicochemical approach was further refined by Figge, Mydosh, and Fencl (1992), who developed equations for the charge contribution of albumin and phosphate at varying pH levels, and by Kellum (2000), who simplified the SIG calculation for bedside use. The approach is especially valuable in the intensive care unit where saline-induced hyperchloraemic acidosis, hypoalbuminaemia, and unmeasured anions frequently coexist.
Interpretation of Stewart Parameters
Each Stewart parameter provides a distinct window into the acid-base disturbance. Interpreting them together reveals disorders that traditional bicarbonate-based analysis may miss.
| Parameter | Normal Range | Low / Narrow | High / Wide |
|---|---|---|---|
| SIDa | 38–42 mEq/L | Acidosis (excess strong anions or deficit of strong cations — e.g. hyperchloraemia, lactic acidosis) | Alkalosis (deficit of strong anions or excess of strong cations — e.g. hypochloraemia, free-water excess) |
| SIDe | 38–42 mEq/L | Reduced buffering capacity (low HCO₃⁻ and/or low ATOT) | Increased buffering (metabolic alkalosis component) |
| SIG | 0 ± 2 mEq/L | Unmeasured cations or lab error (rare) | Unmeasured strong anions (ketoacids, uraemic toxins, toxin metabolites, sepsis-related anions) |
| ATOT charge | ~12–16 mEq/L | Hypoalbuminaemia → alkalinising effect (masks acidosis) | Hyperphosphataemia → acidifying effect |
A patient with a normal anion gap may still have a significantly elevated SIG if they are hypoalbuminaemic. The traditional anion gap falls by approximately 2.5 mEq/L for every 1 g/dL decrease in albumin. The SIG inherently corrects for this, making it a more reliable marker of unmeasured anions in critically ill patients where albumin levels are often low.
SIG and Clinical Outcomes
Multiple studies in ICU populations have associated an elevated SIG with increased mortality. A SIG greater than 5 mEq/L at ICU admission has been associated with higher hospital mortality across mixed surgical, trauma, and sepsis cohorts. While the SIG is not a prognostic tool on its own, it may add independent predictive value when combined with other severity scores.
Causes of Abnormal Stewart Parameters
Disruptions in any of the three independent variables — SID, ATOT, or PCO₂ — produce characteristic acid-base patterns. The following differentials focus on the SID and ATOT components, since PCO₂ disorders are usually apparent from the blood gas.
A narrowed apparent SID indicates an excess of strong anions relative to strong cations. The most common causes in the ICU include:
- Hyperchloraemic acidosis — often iatrogenic from large-volume normal saline (0.9% NaCl) resuscitation. The chloride load directly narrows the SID. Corrected chloride (Cl⁻ × 140 / Na⁺) above 106 mEq/L suggests relative hyperchloraemia.
- Lactic acidosis — lactate is a strong anion and directly reduces SIDa. Type A (hypoperfusion) and type B (metabolic, drug-related, or mitochondrial dysfunction) lactic acidosis are distinguished clinically.
- Renal tubular acidosis — impaired renal chloride excretion or bicarbonate reclamation. In the Stewart framework, this manifests as failure to maintain adequate renal SID excretion.
- Unmeasured anions (elevated SIG) — ketoacids (diabetic or alcoholic ketoacidosis), uraemic solutes, toxic alcohols (formate, glycolate, oxalate), and sepsis-related anions all contribute unmeasured strong anions that widen the SIG while narrowing the true SID.
A widened apparent SID indicates a relative excess of strong cations over strong anions. Causes include:
- Hypochloraemic alkalosis — from vomiting, nasogastric suction, or diuretic therapy. Loss of chloride without proportionate sodium loss widens the SID.
- Sodium excess — administration of sodium bicarbonate, citrate (from massive transfusion or continuous renal replacement therapy anticoagulation), or sodium acetate widens the SID because the accompanying anion is metabolised, leaving behind sodium.
- Free-water deficit — contraction alkalosis concentrates sodium more than chloride, effectively widening the SID. This is common with dehydration and volume depletion.
From the Stewart perspective, the kidneys regulate acid-base status primarily by adjusting chloride excretion (and thus the urinary SID), rather than by bicarbonate reclamation per se.
A SIG significantly above 2 mEq/L indicates the presence of strong anions not accounted for by chloride and lactate. The differential mirrors the traditional elevated anion gap causes, using the mnemonic GOLDMARK:
- G — Glycols (ethylene glycol, propylene glycol)
- O — Oxoproline (5-oxoproline, from chronic paracetamol use)
- L — L-lactate (already measured in SIDa, so SIG elevation implies other anions)
- D — D-lactate (not measured by standard lactate assays; seen in short bowel syndrome)
- M — Methanol (formate accumulation)
- A — Aspirin / salicylates
- R — Renal failure (uraemic anions: sulphate, hippurate, phosphate in excess of normal)
- K — Ketoacidosis (β-hydroxybutyrate, acetoacetate)
In sepsis, an elevated SIG may reflect accumulation of endogenous strong anions whose identity is incompletely characterised — citrate cycle intermediates, sulphate, and other anions associated with organ dysfunction.
Albumin is the dominant non-volatile weak acid in plasma. When albumin is low — as it is in nearly all critically ill patients — the total weak acid buffer (ATOT) decreases. In the Stewart framework, this has an independent alkalinising effect, because fewer anionic charges from weak acids require fewer hydrogen ions to maintain electrical neutrality.
- For every 1 g/dL fall in albumin, the effective SIDe decreases by approximately 2.5–3 mEq/L, narrowing the SIG and potentially masking a co-existing SIG acidosis.
- In a patient with albumin of 2.0 g/dL, the “normal” anion gap may be as low as 6–7 mEq/L, whereas the SIG remains corrected.
- Hypoalbuminaemic alkalosis is one of the most common overlooked acid-base disorders in the ICU and can mask a significant underlying anion-gap acidosis when assessed by traditional methods alone.
Elevated phosphate increases the total non-volatile weak acid concentration, which has an acidifying effect in the Stewart model. This is distinct from the acidosis caused by strong anions and is not captured by the SIG.
- Renal failure — phosphate retention is a common contributor to the acidosis of chronic and acute kidney injury. The phosphate component may account for 3–5 mEq/L of the acidifying effect in severe hyperphosphataemia.
- Tumour lysis syndrome — massive phosphate release from cell lysis.
- Rhabdomyolysis — muscle breakdown releases intracellular phosphate.
Recognising the ATOT contribution separately from the SIG allows the clinician to distinguish between acidosis from unmeasured anions and acidosis from weak acid excess — which have different aetiologies and treatments.
Think of the Stewart model as decomposing metabolic acid-base status into three compartments: (1) the SID effect (water and chloride disorders), (2) the ATOT effect (albumin and phosphate), and (3) the SIG (unmeasured anions). Each compartment has its own differential and its own treatment approach.
Special Populations & Considerations
Pregnancy, paediatric, and neonatal populations have different normal ranges for albumin, electrolytes, and buffering capacity. The Stewart equations used in this calculator are validated primarily for adult ICU populations. Apply with caution outside this group and use age/population-specific reference ranges.
Systematic Stewart Acid-Base Approach
Use the following stepwise approach to interpret Stewart acid-base results at the bedside. This complements (rather than replaces) the traditional pH → PCO₂ → HCO₃⁻ → AG → Delta-Delta approach.
Begin with the arterial blood gas. Identify the primary process (acidaemia vs alkalaemia) and determine whether the PCO₂ is contributing (respiratory component). The PCO₂ is one of the three independent Stewart variables — a respiratory acid-base disorder is addressed the same way in both traditional and Stewart frameworks.
If the PCO₂ is abnormal, assess whether the compensation is appropriate using standard rules (e.g., Winter’s formula for metabolic acidosis). Inappropriate compensation suggests a mixed respiratory-metabolic disorder.
The apparent SID reflects the balance of measured strong ions. A normal SIDa is approximately 38–42 mEq/L.
- SIDa < 38: Excess strong anions — look for hyperchloraemia (check corrected Cl⁻) and elevated lactate. If chloride and lactate do not fully explain the low SID, suspect unmeasured anions (SIG will confirm).
- SIDa > 42: Relative excess of strong cations — consider hypochloraemia (from vomiting, diuretics, NG suction), free-water deficit, or sodium-containing infusions (bicarbonate, citrate).
Check the corrected chloride (Cl⁻ × 140/Na⁺) to determine whether chloride is disproportionately high or low relative to sodium.
The SIG (SIDa − SIDe) reveals unmeasured strong anions. It is inherently corrected for albumin and phosphate, unlike the traditional anion gap.
- SIG 0 ± 2: No significant unmeasured anions. Any acidosis is explained by chloride, lactate, or weak acid changes.
- SIG > 2: Unmeasured strong anions are present. Work through the GOLDMARK differential — ketoacids, uraemic solutes, toxic alcohols, salicylate, 5-oxoproline, D-lactate, or sepsis-related anions.
- SIG < −2: Suggests unmeasured cations (e.g., lithium toxicity), laboratory error, or significant paraproteinaemia. This is uncommon and warrants repeating the labs.
Look at the albumin and phosphate charges separately. These represent the weak acid buffering compartment.
- Low albumin: Reduces ATOT → alkalinising effect. A patient with albumin of 2.0 g/dL loses approximately 5–6 mEq/L of anionic buffer charge compared to a patient with albumin of 4.0 g/dL. This can completely mask a co-existing metabolic acidosis.
- High phosphate: Increases ATOT → acidifying effect. Particularly relevant in renal failure, tumour lysis, and rhabdomyolysis.
The key clinical question is: “Is hypoalbuminaemia hiding an acidosis?” If the SIG is elevated but the pH appears near-normal, the answer is likely yes.
The power of the Stewart approach is the ability to identify co-existing acid-base disorders that offset each other:
- Example 1: A patient with sepsis may have a low SIDa (lactic acidosis) + elevated SIG (unmeasured anions) + low ATOT (hypoalbuminaemia). The hypoalbuminaemic alkalosis partially offsets the SIG acidosis, making the pH appear less deranged than the underlying pathology.
- Example 2: A patient receiving large-volume saline may have a low SIDa (hyperchloraemic acidosis) with a normal SIG — confirming the acidosis is entirely chloride-driven and suggesting balanced crystalloids might have been preferable.
- Example 3: A patient with chronic liver disease may have a wide SIDa (hyponatraemic alkalosis) + elevated SIG (unmeasured anions) + very low ATOT — a triple disorder invisible to traditional AG analysis.
Common Pitfalls & Limitations
The SIDa requires the ionised (free) calcium concentration, not the total serum calcium. Total calcium includes protein-bound and complexed fractions that do not contribute to the strong ion balance. In hypoalbuminaemic patients, total calcium may be low while ionised calcium is normal — using total calcium would artificially narrow the SIDa.
Always use the ionised calcium from the arterial blood gas analyser. If only total calcium is available, the calculator will be less accurate, and this should be acknowledged in the interpretation.
One of the principal advantages of the Stewart approach is that the SIG is inherently corrected for albumin. Clinicians who use only the traditional anion gap may miss significant unmeasured anion acidosis when albumin is low. For every 1 g/dL decrease in albumin below 4.0, the expected normal anion gap decreases by approximately 2.5 mEq/L.
The albumin-corrected anion gap (AGcorr = AG + 2.5 × [4.0 − Alb]) is a partial solution, but the Stewart SIG is more physiologically rigorous because it also accounts for phosphate and uses pH-dependent charge calculations.
The Stewart calculation combines data from an arterial blood gas (pH, PCO₂, HCO₃⁻, ionised calcium, lactate) and a basic metabolic panel (Na⁺, K⁺, Cl⁻) and serum proteins (albumin, phosphate). These are often drawn at different times and from different sites. Significant haemodynamic or metabolic changes between sample collections can produce discordant results and an inaccurate SIG.
For best accuracy, use a simultaneously drawn ABG and chemistry panel. If samples were drawn more than 30 minutes apart in a haemodynamically unstable patient, interpret results with caution.
The SIG is a calculated value derived from multiple measured inputs, each with its own laboratory imprecision. Propagation of measurement errors means the SIG has an inherent uncertainty of approximately ± 2–3 mEq/L. A SIG of 3 mEq/L is likely within the noise range, while a SIG above 5 mEq/L is more clinically meaningful.
Trending the SIG over time (using consistent sampling methods) is more informative than interpreting a single isolated value. A rising SIG suggests worsening unmeasured anion accumulation even if the absolute value is modest.
The Stewart approach is a complementary framework, not a replacement for the Henderson-Hasselbalch and anion gap methods. Both approaches ultimately describe the same physiology from different mathematical perspectives. The Stewart model excels in complex ICU scenarios but is unnecessarily cumbersome for straightforward acid-base questions.
In practice, a hybrid approach works best: use the traditional method for initial assessment (pH, PCO₂, HCO₃⁻, AG) and layer on the Stewart analysis when the picture is unclear — particularly when albumin is low, multiple disorders coexist, or the AG fails to explain the clinical picture.
Quick Reference Summary
| Finding | Stewart Diagnosis | Common Causes |
|---|---|---|
| Low SIDa, Normal SIG | Hyperchloraemic / strong-anion acidosis | Saline resuscitation, RTA, diarrhoea |
| Low SIDa, High SIG | Unmeasured anion acidosis | Ketoacidosis, uraemia, toxic alcohols, sepsis |
| High SIDa, Normal SIG | Strong-ion alkalosis | Vomiting, diuretics, hypochloraemia |
| Low ATOT charge | Hypoalbuminaemic alkalosis | Critical illness, liver disease, nephrotic syndrome |
| High ATOT charge | Hyperphosphataemic acidosis | Renal failure, tumour lysis, rhabdomyolysis |
The Golden Rule: When the anion gap looks normal but the patient looks sick — check the albumin and calculate the SIG. Hypoalbuminaemia masks acidosis in the traditional framework, and the Stewart approach reveals it.
Disclaimer & References
For Educational Purposes Only. This calculator and the accompanying clinical information are intended as educational tools for healthcare professionals. They do not replace clinical judgement. Results should be interpreted in the full clinical context. Lab reference ranges vary by institution — verify with your own laboratory. Drug dosages should be confirmed against current prescribing information.
References
- Stewart PA. Modern quantitative acid-base chemistry. Can J Physiol Pharmacol. 1983;61(12):1444–1461. DOI: 10.1139/y83-207
- Figge J, Mydosh T, Fencl V. Serum proteins and acid-base equilibria: a follow-up. J Lab Clin Med. 1992;120(5):713–719. DOI: 10.5555/uri:pii:002221439290092H
- Kellum JA. Determinants of blood pH in health and disease. Crit Care. 2000;4(1):6–14. DOI: 10.1186/cc644
- Kellum JA, Kramer DJ, Pinsky MR. Strong ion gap: a methodology for exploring unexplained anions. J Crit Care. 1995;10(2):51–55. DOI: 10.1016/0883-9441(95)90016-0
- Dubin A, Menises MM, Masevicius FD, et al. Comparison of three different methods of evaluation of metabolic acid-base disorders. Crit Care Med. 2007;35(5):1264–1270. DOI: 10.1097/01.CCM.0000259536.11943.90
- Kaplan LJ, Kellum JA. Initial pH, base deficit, lactate, anion gap, strong ion difference, and strong ion gap predict outcome from major vascular injury. Crit Care Med. 2004;32(5):1120–1124. DOI: 10.1097/01.CCM.0000125517.28517.74
- Moviat M, van Haren F, van der Hoeven H. Conventional or physicochemical approach in intensive care unit patients with metabolic acidosis. Crit Care. 2003;7(3):R41–R45. DOI: 10.1186/cc2184
- Morgan TJ. The Stewart approach — one clinician’s perspective. Clin Biochem Rev. 2009;30(2):41–54. PMID: 19565024
- Figge J, Jabor A, Kazda A, Fencl V. Anion gap and hypoalbuminemia. Crit Care Med. 1998;26(11):1807–1810. DOI: 10.1097/00003246-199811000-00019
- Kellum JA. Clinical review: Reunification of acid-base physiology. Crit Care. 2005;9(5):500–507. DOI: 10.1186/cc3789