The Strong Ion Difference (Stewart)
In 1981 the Canadian physiologist Peter Stewart published a quantitative acid-base framework that re-derives plasma pH from just three independent variables. It is more rigorous than the anion gap — and it explains hyperchloremic acidosis without invoking any "unmeasured" anion at all.
The Stewart approach to acid-base
The conventional acid-base framework treats bicarbonate, chloride, sodium, and albumin as if they were independent handles the clinician can pull to move pH. Stewart's insight was the opposite: under the law of electroneutrality, dissolved CO₂, the strong-ion charge difference, and the total weak-acid concentration are the only three genuinely independent variables. Everything else — including bicarbonate, which the traditional approach treats as primary — is a dependent variable that the body computes in order to satisfy electroneutrality and water dissociation simultaneously.
Stewart called these the three independent variables:
- SID — the strong ion difference. The net charge of all "strong" ions (those that are fully dissociated at physiological pH: Na⁺, K⁺, Ca²⁺, Mg²⁺ on the cation side; Cl⁻ and the anions of strong acids on the anion side).
- ATot — the total concentration of non-volatile weak acids (in plasma, essentially albumin plus inorganic phosphate).
- PaCO₂ — the partial pressure of CO₂ (set by ventilation).
Given these three inputs and the constraints of electroneutrality and mass conservation on water, plasma pH is fully determined. There is no fourth variable. This reframing is not a matter of taste — it changes which interventions you recognize as acid-base-active.
The strong ion difference, defined
The apparent SID (SIDa) is the difference between the sum of strong cations and the sum of strong anions. At the bedside it is approximated as:
Normal plasma SID is approximately 40 mEq/L (the bedside form, Na⁺ − Cl⁻, gives ~140 − 100 = 40). The crucial point is that this number is always positive — there is always more strong cation than strong anion in plasma, because the deficit is balanced by the weak acids (albumin, phosphate) and by bicarbonate. SID is therefore not a measure of "extra" anything; it is the fixed charge that the weak-acid / buffer system must neutralize.
How SID moves pH
Because SID is positive, lowering it makes the plasma more acidic, and raising it makes the plasma more alkalotic. This is the Stewart explanation for several disturbances the traditional framework handles awkwardly:
| Disturbance | SID change | pH effect |
|---|---|---|
| Large-volume normal saline (0.9% NaCl) | Cl⁻ rises more than Na⁺ (saline SID = 0), so plasma SID falls | Hyperchloremic metabolic acidosis |
| Diarrhea (loss of Na⁺-rich, HCO₃⁻-rich fluid, Cl⁻ preserved) | SID falls | Normal-anion-gap metabolic acidosis |
| Vomiting (loss of Cl⁻-rich gastric fluid) | SID rises | Metabolic alkalosis |
| Diuretics (loop / thiazide) | Cl⁻ lost disproportionately, SID rises | Metabolic alkalosis |
| Renal failure | Unmeasured strong anions accumulate, SID falls | High-anion-gap metabolic acidosis |
Notice that Stewart explains hyperchloremic (normal-anion-gap) metabolic acidosis without any reference to bicarbonate at all: saline lowers SID because saline itself has a SID of zero (equal Na⁺ and Cl⁻), so infusing it pulls plasma SID toward zero, and pH falls as a direct consequence. The fall in measured HCO₃⁻ is the dependent output of the same equation, not the driver. You can compute this at the bedside with the strong ion difference calculator.
Traditional anion gap vs Stewart SID
The two frameworks are not contradictory — they are different partitions of the same electroneutrality constraint. The traditional anion gap formula Na⁺ − (Cl⁻ + HCO₃⁻) isolates the unmeasured anion pool, which is excellent for screening HAGMA (is there a new organic acid?). Stewart's SID isolates the strong-ion charge difference, which is excellent for explaining why chloride-driven and dilutional acidoses happen. The two are complementary:
- Anion gap asks: is there an unmeasured anion accumulating? Best for HAGMA screening (lactate, ketones, toxins).
- SID asks: has the balance of strong ions shifted? Best for NAGMA / hyperchloremia, fluid-induced acidosis, and complex ICU disorders.
- Strong ion gap (SIG) — the Stewart analogue of the anion gap — corrects SID for ATot and is more sensitive than the conventional albumin-corrected gap for detecting unmeasured anions in critical illness.
In the ICU, the traditional anion gap misses roughly a third of acid-base disturbances because critically ill patients have deranged albumin, phosphate, and chloride simultaneously. Stewart's framework (or its bedside shortcut, the SID) accounts for all of these at once and is the method of choice for analyzing mixed disorders in sepsis, post-resuscitation saline loading, and renal failure.
Practical takeaways
For routine clinical work, the albumin-corrected anion gap remains the workhorse: it is fast, familiar, and sufficient for diagnosing the great majority of metabolic acidoses. But reach for the SID framework — and especially the strong ion gap — whenever you see (1) a hyperchloremic acidosis you cannot fully explain, (2) a mixed disorder in a critically ill patient with abnormal albumin or phosphate, or (3) a post-resuscitation acidosis after large-volume crystalloid. In those settings Stewart's three independent variables will resolve the picture that the anion gap alone leaves ambiguous.
Reference: Kellum JA, "Disorders of acid-base balance," Crit Care Med and the StatPearls Stewart / quantitative acid-base chapter; Stewart PA, "Independent and dependent variables of acid-base control" (1981); the LITFL Stewart acid-base page. Normal SID ≈ 40 mEq/L is the consensus value used across these sources.