Acid-base disorders

Contents

The Six-Step Approach

A systematic approach prevents errors and ensures all disorders are identified.

Step 1: pH — Acidaemia or Alkalaemia?

  • pH <7.35: acidaemia
  • pH >7.45: alkalaemia
  • pH 7.35–7.45: normal (may still have mixed disorder)

Step 2: PaCO₂ — Respiratory Component?

  • PaCO₂ >6.0 kPa (>45 mmHg): respiratory acidosis component
  • PaCO₂ <4.7 kPa (<35 mmHg): respiratory alkalosis component

Step 3: Bicarbonate — Metabolic Component?

  • HCO₃⁻ >26 mmol/L: metabolic alkalosis component
  • HCO₃⁻ <22 mmol/L: metabolic acidosis component

Step 4: Is the Primary Disorder Consistent with the pH?

  • If pH acidaemic: primary disorder is whichever (respiratory or metabolic) is also acidotic
  • If pH alkalaemic: primary disorder is whichever (respiratory or metabolic) is also alkalotic

Step 5: Compensation — Is It Appropriate?

The body compensates for primary disorders to minimise pH change. If compensation is greater or less than expected, a second primary disorder exists.

Expected compensation formulas:

Primary disorder Expected compensation
Metabolic acidosis PaCO₂ = (1.5 × HCO₃⁻) + 8 ± 2 (Winter's formula)
Metabolic alkalosis PaCO₂ rises ~0.7 mmHg per 1 mmol/L rise in HCO₃⁻
Respiratory acidosis (acute) HCO₃⁻ rises 1 mmol/L per 10 mmHg rise in PaCO₂
Respiratory acidosis (chronic) HCO₃⁻ rises 3.5 mmol/L per 10 mmHg rise in PaCO₂
Respiratory alkalosis (acute) HCO₃⁻ falls 2 mmol/L per 10 mmHg fall in PaCO₂
Respiratory alkalosis (chronic) HCO₃⁻ falls 5 mmol/L per 10 mmHg fall in PaCO₂

Winter's formula example: HCO₃⁻ = 12 mmol/L → expected PaCO₂ = (1.5 × 12) + 8 = 26 mmHg. If measured PaCO₂ is 20 mmHg: there is a concurrent primary respiratory alkalosis.

Step 6: Anion Gap — If Metabolic Acidosis, Is There a HAGMA?

Calculate anion gap (AG) and delta-delta ratio where relevant (see below).

Primary Disorders and Compensation

Metabolic Acidosis

Primary ↓ HCO₃⁻; respiratory compensation → hyperventilation → ↓ PaCO₂

Type Key Features Examples
High anion gap (HAGMA) Unmeasured acids accumulate; AG >12 mmol/L Lactic acidosis, DKA, AKI, salicylate, methanol, ethylene glycol
Normal anion gap (NAGMA) HCO₃⁻ lost (gut or kidney); hyperchloraemia Diarrhoea, RTA, high-chloride IV fluids, Addison's

Metabolic Alkalosis

Primary ↑ HCO₃⁻; respiratory compensation → hypoventilation → ↑ PaCO₂

Causes:

  • Loss of acid (vomiting, NG suctioning: loss of HCl; self-perpetuating in hypovolaemia)
  • Gain of base (excessive NaHCO₃ administration, blood product transfusion — citrate converted to HCO₃⁻)
  • Mineralocorticoid excess: Conn's syndrome, exogenous steroids → ↑ Na⁺/H⁺ exchange → H⁺ loss in urine

Saline-responsive (urine Cl⁻ <20 mmol/L): vomiting, diuretics — respond to IV NaCl
**Saline-resistant** (urine Cl⁻ >20 mmol/L): Conn's, Cushing's — treat underlying cause

Respiratory Acidosis

Primary ↑ PaCO₂; renal compensation (slow — hours to days) → ↑ HCO₃⁻ retention

  • Acute: minimal metabolic compensation; pH falls sharply
  • Chronic: complete compensation; pH may be near-normal despite ↑ PaCO₂

Causes: reduced ventilatory drive, neuromuscular weakness, airway obstruction, severe COPD/asthma, oversedation

Respiratory Alkalosis

Primary ↓ PaCO₂; renal compensation → ↓ HCO₃⁻ excretion

Causes: anxiety, pain, fever, sepsis (early), salicylate poisoning (early), head injury, hypoxia (reflex hyperventilation), iatrogenic hyperventilation

Anion Gap and MUDPILES

Anion Gap = Na⁺ − (Cl⁻ + HCO₃⁻)
Normal: 8–12 mmol/L (some labs quote 8–16 when albumin not corrected)

Albumin correction: hypoalbuminaemia falsely lowers the AG — critical in ICU patients. Corrected AG = measured AG + 2.5 × (40 − albumin in g/L). Always correct for albumin in ICU.

MUDPILES (HAGMA Causes)

  • Methanol
  • Uraemia (acute kidney injury — accumulation of phosphate, sulfate, organic acids)
  • Diabetic ketoacidosis
  • Paraldehyde
  • Isopropanol (does not cause acidosis directly — osmol gap)
  • Lactic acidosis (most common in ICU)
  • Ethylene glycol
  • Salicylate / Sepsis

Delta-Delta Ratio (Δ/Δ)

In a pure HAGMA, for every 1 mmol/L rise in AG, HCO₃⁻ should fall by 1 mmol/L.

Δ/Δ = (AG − 12) / (24 − HCO₃⁻)

  • Δ/Δ 0.8–2.0: pure HAGMA — the HCO₃⁻ fall is proportionate to the AG rise
  • Δ/Δ <0.8: concurrent NAGMA — HCO₃⁻ is falling more than the AG rise explains → additional process lowering HCO₃⁻ (e.g., diarrhoea + sepsis)
  • Δ/Δ >2.0: concurrent metabolic alkalosis — HCO₃⁻ is higher than expected for the AG rise → pre-existing metabolic alkalosis masking the acidosis (e.g., chronic vomiting + DKA)

Stewart Approach

The Stewart (physicochemical) approach views acid-base status as determined by three independent variables:

  1. SID (Strong Ion Difference) = (Na⁺ + K⁺ + Ca²⁺ + Mg²⁺) − (Cl⁻ + lactate + other strong anions) — normally ~40 mEq/L
  2. ATOT (Total weak acid concentration) — mainly albumin and phosphate
  3. PaCO₂ — controlled by ventilation

Key insight: [H⁺] (and pH) is a dependent variable — it is determined by SID, ATOT, and PaCO₂, not by HCO₃⁻ directly (which is also dependent).

Clinical Utility of Stewart in ICU

  • Hyperchloraemic acidosis from IV saline: traditional approach labels this a normal AG acidosis; Stewart explains it mechanistically — high Cl⁻ reduces SID (Na⁺ − Cl⁻) → ↓ [SID] → acidosis
  • Hypoalbuminaemia: low albumin (weak acid) → ↓ ATOT → alkalising effect; may mask a coexistent metabolic acidosis (corrected AG reveals this)
  • Dilutional acidosis: giving large volumes of any chloride-containing fluid dilutes SID and raises Cl⁻ relative to Na⁺

Common ICU Acid-Base Scenarios

1. Lactic Acidosis

  • Type A (tissue hypoxia): septic shock, cardiac arrest, haemorrhage — most common in ICU
  • Type B (no tissue hypoperfusion): metformin (rare — usually only in severe AKI), liver failure, thiamine deficiency, mitochondrial toxins
  • Lactate >2 mmol/L: elevated; >4 mmol/L: severe (associated with higher mortality in sepsis)
  • Treatment: address underlying cause (antibiotics, vasopressors, source control) — sodium bicarbonate only for severe acidosis (pH <7.1) with haemodynamic compromise; limited evidence

2. Respiratory Acidosis on ICU — Permissive Hypercapnia

In lung-protective ventilation for ARDS: low tidal volumes (6 mL/kg) → CO₂ may accumulate → respiratory acidosis. Permissive hypercapnia (PaCO₂ up to 8–10 kPa) is acceptable to avoid volutrauma (target pH ≥7.20). Avoid aggressive hyperventilation.

3. Metabolic Alkalosis — Common and Underappreciated

  • Vomiting or NG suction: loss of gastric acid (HCl) → HCO₃⁻ rises; hypovolaemia, hypochloraemia, hypokalaemia perpetuate it (renal conservation of Cl⁻ at expense of H⁺)
  • Treatment: IV normal saline (restores Cl⁻ and volume); replace K⁺; stop NG suctioning if possible
  • Severe metabolic alkalosis (HCO₃⁻ >40 mmol/L): respiratory depression, cardiac arrhythmias

4. Mixed Acid-Base Disorders

Very common in ICU — a patient with sepsis may have: metabolic acidosis (lactic) + metabolic alkalosis (from diuretics) + respiratory alkalosis (sepsis → hyperventilation) simultaneously. The delta-delta ratio and compensation formulas identify multiple coexisting disorders.

Viva Questions

1. A patient has pH 7.32, PaCO₂ 3.8 kPa, HCO₃⁻ 14 mmol/L, Na⁺ 138, Cl⁻ 105, albumin 22 g/L, lactate 4.2 mmol/L. Interpret this systematically.

Step 1: pH 7.32 — acidaemia. Step 2: PaCO₂ 3.8 kPa (≈29 mmHg) — low, suggesting respiratory alkalosis component or appropriate compensation. Step 3: HCO₃⁻ 14 mmol/L — low, metabolic acidosis. Step 4: Primary disorder is metabolic acidosis (pH acidaemic, HCO₃⁻ low). Step 5: Compensation — Winter's formula: expected PaCO₂ = (1.5 × 14) + 8 = 29 mmHg = ~3.9 kPa. Measured PaCO₂ is 3.8 kPa — appropriate compensation, no concurrent respiratory disorder. Step 6: Anion gap — standard AG = 138 − (105 + 14) = 19 mmol/L. But albumin is 22 g/L (normal 40) → corrected AG = 19 + 2.5 × (40−22) = 19 + 45 = 64? That doesn't look right — let me recalculate: 2.5 × 18 = 45 → corrected AG = 19 + 45 = 64 — this is very high; or more commonly: corrected AG = 19 + 2 × (40−22) = 19 + 36 = 55. Either way: markedly elevated AG, confirming HAGMA. The lactate is 4.2 mmol/L — lactic acidosis is the primary cause. Delta-delta: (55−12) / (24−14) = 43/10 = 4.3 → >2: concurrent metabolic alkalosis (likely from previous diuresis or vomiting). Interpretation: severe lactic acidosis (likely septic shock or shock state), with concurrent metabolic alkalosis partially masking the acidosis; the true acid load is greater than the pH suggests.

2. Explain why hypoalbuminaemia matters in ICU acid-base interpretation.

Albumin is a weak acid — it carries a negative charge at physiological pH and is an important component of the body's total weak acid buffer (ATOT in Stewart terminology). In the traditional Henderson-Hasselbalch approach, hypoalbuminaemia is often ignored; however, low albumin has a direct alkalising effect (less weak acid = higher pH), which creates a "cryptic" metabolic alkalosis that partially offsets coexisting metabolic acidosis. Critically, hypoalbuminaemia also lowers the measured anion gap. If albumin is 20 g/L (severely hypoalbuminaemic, as in critical illness), the measured AG may be 8 mmol/L — apparently normal. But the corrected AG (adjusting for the deficit of albumin) may be 20+ mmol/L, revealing a significant HAGMA that the uncorrected value hides. Failure to correct the AG for albumin in ICU patients (where hypoalbuminaemia is near-universal) leads to systematic underestimation of acidotic disorders and missed diagnoses (missed lactic acidosis, missed uraemic acidosis). Always correct the AG for albumin in ICU.

3. Why is sodium bicarbonate not routinely given for lactic acidosis in septic shock?

The intuitive rationale for bicarbonate in lactic acidosis — neutralising H⁺ to raise pH — does not translate into clinical benefit. There are several reasons. First, the bicarbonate buffer reaction (H⁺ + HCO₃⁻ → H₂CO₃ → H₂O + CO₂) generates CO₂, which crosses cell membranes freely and can paradoxically worsen intracellular acidosis even while extracellular pH rises. Second, lactic acidosis is a Type A disorder — a marker of tissue hypoperfusion; correcting the pH chemically without addressing the underlying cause (restoring perfusion, cardiac output, oxygen delivery) achieves nothing physiologically. Third, sodium bicarbonate is a hypertonic sodium load — it causes fluid overload, hypernatraemia, and hyperosmolality. Fourth, alkalosis from bicarbonate shifts the oxyhaemoglobin dissociation curve left (Bohr effect) — worsening oxygen offloading to tissues already hypoxic. However, the BICAR-ICU trial (Jaber, Lancet 2018) showed a benefit of bicarbonate (targeting pH ≥7.30) in patients with severe acidaemia (pH ≤7.20) and AKI stage 2–3 — specifically reducing the need for RRT and 28-day mortality in the AKI subgroup. Current practice: sodium bicarbonate is not routinely recommended in lactic acidosis; consider for pH <7.15–7.20 with haemodynamic compromise or severe AKI, recognising the limitations.