The Henderson-Hasselbalch equation. Winter’s formula. The anion gap. The delta-delta ratio. The albumin correction. The osmolar gap. The Stewart approach. The strong ion difference. The strong ion gap. Compensation rules for every primary disorder. Mnemonics for every differential. This is the article that replaces a textbook chapter — and it does not leave a single concept behind.
“You cannot practice critical care medicine without mastering acid-base physiology. You cannot titrate a ventilator without understanding the relationship between PaCO₂ and pH. You cannot resuscitate a septic patient without understanding why their lactate is elevated and what that means for their bicarbonate. You cannot choose between normal saline and lactated Ringer’s without understanding what chloride does to the strong ion difference. Acid-base is not a topic in critical care. It is the substrate on which every other topic sits.”
A Message From Javier Amador-Castaneda, BHS, RRT, FCCM
Founder & CEO, Interprofessional Critical Care Network (ICCN)
This article exists because someone told me that what we publish at ICCN is “oversimplified.” So today, I am going to publish the opposite of oversimplified. I am going to publish a comprehensive, clinically rigorous, and practically applicable acid-base physiology guide that can be utilized for your understanding of acid base physiology.
This is not a summary. This is not a clinical pearl. This is a complete acid-base framework — from the chemistry to the bedside — that covers every concept a critical care clinician needs to interpret any arterial blood gas in any clinical context.
It starts with first principles. It ends with clinical cases. It covers the traditional Henderson-Hasselbalch approach, the compensation formulas for every primary disorder, the anion gap with albumin correction, the delta-delta ratio, the osmolar gap, and the full Stewart physicochemical approach. Nothing is left behind.
Save this article. Print it. Laminate it. Bring it to the ICU. This is the one you will reference at 2 AM when the ABG does not make sense and the patient is crashing.
PART 1: THE CHEMISTRY — WHY pH MATTERS AND WHAT CONTROLS IT
1.1 — The Hydrogen Ion and pH
pH is the negative logarithm of the hydrogen ion concentration:
pH = −log[H⁺]
Normal arterial pH is 7.35–7.45, corresponding to a hydrogen ion concentration of approximately 35–45 nmol/L. This is an extraordinarily narrow range — the body defends it aggressively because enzymatic function, oxygen-hemoglobin binding, and cellular metabolism are all exquisitely sensitive to hydrogen ion concentration.
A pH below 7.35 is acidemia. A pH above 7.45 is alkalemia. The processes that drive pH in those directions are called acidosis and alkalosis, respectively. A patient can have an acidosis without acidemia — if a concurrent alkalosis is offsetting the pH change. Understanding this distinction is the first step to mastering acid-base interpretation.
1.2 — The Henderson-Hasselbalch Equation
The relationship between pH, PaCO₂, and bicarbonate is defined by the Henderson-Hasselbalch equation:
pH = 6.1 + log([HCO₃⁻] / (0.03 × PaCO₂))
This equation tells you that pH is determined by the ratio of bicarbonate (the metabolic component) to dissolved CO₂ (the respiratory component) — not by the absolute value of either one. A bicarbonate of 12 with a PaCO₂ of 24 produces the same pH as a bicarbonate of 24 with a PaCO₂ of 48. The ratio is identical. The pH is identical.
This is why compensation works: when one component changes (e.g., bicarbonate falls in metabolic acidosis), the other component changes in the same direction (PaCO₂ falls through hyperventilation) to preserve the ratio and defend the pH.
1.3 — The Three Buffer Systems
The body maintains pH through three integrated buffer systems:
Chemical buffers (immediate, seconds): The bicarbonate-carbonic acid system (HCO₃⁻/H₂CO₃), proteins (especially hemoglobin and albumin), and phosphate provide immediate buffering capacity. The bicarbonate system is the most important extracellular buffer because both its components are independently regulated — HCO₃⁻ by the kidneys and CO₂ by the lungs.
Respiratory compensation (minutes to hours): Changes in ventilation alter PaCO₂. Acidosis stimulates hyperventilation (reducing PaCO₂). Alkalosis suppresses ventilation (raising PaCO₂). Respiratory compensation begins within minutes and reaches maximal effect within 12–24 hours.
Renal compensation (hours to days): The kidneys regulate bicarbonate reabsorption and hydrogen ion excretion. In respiratory acidosis, the kidneys retain bicarbonate. In respiratory alkalosis, the kidneys excrete bicarbonate. Renal compensation is slow (full effect in 3–5 days) but powerful.
PART 2: THE SYSTEMATIC ABG INTERPRETATION — THE 6-STEP METHOD
Every arterial blood gas should be interpreted using a systematic approach. Shortcuts lead to missed diagnoses. Here is the method.
Step 1: Look at the pH — Determine Acidemia or Alkalemia
- pH < 7.35 → Acidemia (the net process is acidosis)
- pH > 7.45 → Alkalemia (the net process is alkalosis)
- pH 7.35–7.45 → Normal or a mixed disorder where opposing processes cancel each other
Step 2: Identify the Primary Disorder
Look at PaCO₂ (normal: 35–45 mm Hg) and HCO₃⁻ (normal: 22–26 mEq/L):
- Metabolic acidosis: Low HCO₃⁻ (< 22) with low pH
- Metabolic alkalosis: High HCO₃⁻ (> 26) with high pH
- Respiratory acidosis: High PaCO₂ (> 45) with low pH
- Respiratory alkalosis: Low PaCO₂ (< 35) with high pH
The primary disorder is the one that explains the direction of the pH. If pH is low and both HCO₃⁻ and PaCO₂ are abnormal, the one pushing pH in the acidotic direction is the primary disorder.
Step 3: Assess the Compensation — Is It Appropriate?
The body never overcompensates. If the compensation appears to overshoot — producing a pH on the opposite side of normal — there is a second primary disorder.
Compensation formulas:
For Metabolic Acidosis (Winter’s Formula):
Expected PaCO₂ = (1.5 × HCO₃⁻) + 8 ± 2
This is the formula that tells you whether the respiratory compensation for a metabolic acidosis is appropriate. If the measured PaCO₂ is higher than expected → there is a concurrent respiratory acidosis. If lower than expected → there is a concurrent respiratory alkalosis.
Alternative rapid estimate: Expected PaCO₂ ≈ last two digits of the pH. If pH is 7.25, expected PaCO₂ is approximately 25 mm Hg.
Another rapid estimate: Expected PaCO₂ ≈ HCO₃⁻ + 15.
For Metabolic Alkalosis:
Expected PaCO₂ = (0.7 × HCO₃⁻) + 21 ± 2
Alternative: Expected PaCO₂ rises approximately 0.7 mm Hg for each 1 mEq/L rise in HCO₃⁻.
Important: respiratory compensation for metabolic alkalosis is limited. PaCO₂ rarely rises above 55 mm Hg because hypoxemia from hypoventilation stimulates the peripheral chemoreceptors and prevents further ventilatory suppression.
For Acute Respiratory Acidosis:
Expected HCO₃⁻ rises 1 mEq/L for each 10 mm Hg rise in PaCO₂
(Minimal buffering — chemical buffers only, no renal compensation yet.)
For Chronic Respiratory Acidosis:
Expected HCO₃⁻ rises 3.5 mEq/L for each 10 mm Hg rise in PaCO₂
(Full renal compensation over 3–5 days.)
For Acute Respiratory Alkalosis:
Expected HCO₃⁻ falls 2 mEq/L for each 10 mm Hg fall in PaCO₂
For Chronic Respiratory Alkalosis:
Expected HCO₃⁻ falls 5 mEq/L for each 10 mm Hg fall in PaCO₂
If the measured values do not match the expected compensation → a mixed acid-base disorder is present.
Step 4: Calculate the Anion Gap
AG = Na⁺ − (Cl⁻ + HCO₃⁻)
Normal anion gap: 12 ± 4 mEq/L (laboratory-dependent; some references use 8–12)
The anion gap represents the unmeasured anions in plasma — primarily albumin, with contributions from phosphate, sulfate, and organic anions (lactate, ketoacids). An elevated anion gap indicates the presence of unmeasured anions — an acid that has been added to the blood and consumed bicarbonate in the process.
Step 5: Correct the Anion Gap for Albumin
This step is critically important and frequently missed — especially in the ICU, where hypoalbuminemia is nearly universal.
Albumin is a negatively charged protein that accounts for a significant portion of the normal anion gap. When albumin falls, the “normal” anion gap falls with it. A patient with an albumin of 2.0 g/dL has a “normal” anion gap that is approximately 5 mEq/L lower than a patient with an albumin of 4.0 g/dL. Without correction, a high anion gap metabolic acidosis can be hidden behind a “normal-appearing” anion gap.
Corrected AG = Measured AG + 2.5 × (4.0 − measured albumin in g/dL)
For every 1 g/dL decrease in albumin below 4.0, add approximately 2.5 mEq/L to the measured anion gap.
Example: Na⁺ = 140, Cl⁻ = 105, HCO₃⁻ = 18, Albumin = 2.0 g/dL
- Measured AG = 140 − (105 + 18) = 17 → borderline elevated
- Corrected AG = 17 + 2.5 × (4.0 − 2.0) = 17 + 5 = 22 → significantly elevated
Without the albumin correction, this patient’s high anion gap acidosis could be underestimated or missed entirely. In the ICU, always correct the anion gap for albumin.
Step 6: Calculate the Delta-Delta (Δ-Δ) Ratio
The delta-delta ratio compares the change in anion gap to the change in bicarbonate:
Δ-Δ = (Measured AG − Normal AG) / (Normal HCO₃⁻ − Measured HCO₃⁻)
Use the albumin-corrected AG and a normal AG of 12 and normal HCO₃⁻ of 24 for calculation.
Interpretation:
- Δ-Δ < 1: The bicarbonate has fallen more than the anion gap has risen. This means there is a concurrent non-anion gap metabolic acidosis — an additional acid-base disorder beyond the HAGMA. Causes: hyperchloremia (saline resuscitation), renal tubular acidosis, diarrhea.
- Δ-Δ between 1 and 2: The change in anion gap matches the change in bicarbonate. This is a pure high anion gap metabolic acidosis with no concurrent metabolic disorder.
- Δ-Δ > 2: The anion gap has risen more than the bicarbonate has fallen. This means the bicarbonate is higher than expected — indicating a concurrent metabolic alkalosis. Causes: vomiting, nasogastric suction, diuretic use, volume contraction.
The delta-delta ratio is the single most powerful tool for detecting mixed metabolic disorders. It should be calculated on every HAGMA.
PART 3: HIGH ANION GAP METABOLIC ACIDOSIS — THE DIFFERENTIAL
The mnemonic MUDPILES captures the classic causes:
- M — Methanol
- U — Uremia (advanced renal failure, GFR < 20 mL/min)
- D — Diabetic ketoacidosis (also alcoholic and starvation ketoacidosis)
- P — Propylene glycol, Paraldehyde
- I — Isoniazid, Iron
- L — Lactic acidosis (Type A: hypoperfusion; Type B: non-hypoperfusion)
- E — Ethylene glycol
- S — Salicylates
A more contemporary and clinically complete mnemonic is GOLD MARK:
- G — Glycols (ethylene glycol, propylene glycol)
- O — Oxoproline (5-oxoproline, associated with chronic acetaminophen use)
- L — L-lactate
- D — D-lactate (short bowel syndrome, bacterial overgrowth)
- M — Methanol
- A — Aspirin (salicylates)
- R — Renal failure
- K — Ketoacidosis (diabetic, alcoholic, starvation)
The Osmolar Gap — When to Calculate It and What It Means
When the cause of a HAGMA is unclear, calculate the osmolar gap:
Calculated osmolality = 2(Na⁺) + (Glucose/18) + (BUN/2.8)
Osmolar gap = Measured osmolality − Calculated osmolality
Normal osmolar gap: < 10 mOsm/kg
An elevated osmolar gap in the setting of HAGMA suggests the presence of an osmotically active, unmeasured substance — classically:
- Methanol (metabolized to formic acid → optic nerve damage, blindness)
- Ethylene glycol (metabolized to glycolic and oxalic acid → calcium oxalate crystals, renal failure)
- Propylene glycol (IV lorazepam, IV diazepam carrier)
- Isopropyl alcohol (produces an osmolar gap without a HAGMA — metabolized to acetone, not an acid)
Critical clinical pearl: The osmolar gap may normalize as the toxic alcohol is metabolized to its acid metabolite. A normal osmolar gap does not exclude toxic alcohol ingestion if the patient presents late. Conversely, an elevated osmolar gap with a normal anion gap may represent early presentation before acid metabolites have accumulated.
PART 4: NON-ANION GAP METABOLIC ACIDOSIS (NAGMA) — THE OVERLOOKED DIAGNOSIS
NAGMA — also called hyperchloremic metabolic acidosis — is defined by a low bicarbonate, low pH, and a normal anion gap. It is remarkably common in the ICU: studies have shown that 19–41% of ICU patients demonstrate NAGMA at some point during their stay.
The pathophysiology: when bicarbonate is lost (through the kidneys or the GI tract) or when chloride is gained (through saline administration), the anion gap remains normal because the lost bicarbonate is replaced by chloride — maintaining electroneutrality without accumulating unmeasured anions.
The Urine Anion Gap — Distinguishing Renal from GI Causes:
Urine AG = Urine Na⁺ + Urine K⁺ − Urine Cl⁻
- Negative urine AG (e.g., −20): The kidneys are appropriately excreting ammonium (NH₄⁺), which carries Cl⁻ with it, making urine Cl⁻ exceed urine cations. This suggests an extrarenal cause — usually GI bicarbonate loss (diarrhea, fistulae, ileostomy output).
- Positive urine AG (e.g., +15): The kidneys are failing to excrete adequate ammonium. This suggests a renal cause — usually renal tubular acidosis (RTA) or early chronic kidney disease.
Renal Tubular Acidosis — The Three Types:
Type 1 (Distal RTA): Failure of the distal tubule to secrete H⁺. Urine pH is inappropriately high (> 5.5). Associated with hypokalemia, nephrocalcinosis, and autoimmune diseases (Sjögren’s).
Type 2 (Proximal RTA): Failure of the proximal tubule to reabsorb bicarbonate. The urine pH is initially high (as bicarbonate spills) but falls below 5.5 once the serum bicarbonate drops below the lowered threshold. Associated with Fanconi syndrome, multiple myeloma, and carbonic anhydrase inhibitors (acetazolamide).
Type 4 (Hypoaldosteronism): Impaired aldosterone secretion or action → decreased H⁺ and K⁺ secretion in the collecting duct. Associated with hyperkalemia, diabetes (hyporeninemic hypoaldosteronism), ACE inhibitors, ARBs, spironolactone, and adrenal insufficiency.
The iatrogenic cause: Large-volume normal saline resuscitation is the most common cause of NAGMA in the ICU. Normal saline contains 154 mEq/L of chloride — significantly higher than plasma chloride (96–106 mEq/L). Aggressive saline infusion raises serum chloride, narrows the strong ion difference (see Part 6), and produces a hyperchloremic, non-anion gap metabolic acidosis. This is why the 2026 SSC guidelines recommend balanced crystalloids over normal saline — and why understanding the Stewart approach matters at the bedside.
PART 5: METABOLIC ALKALOSIS — THE DISORDER EVERYONE UNDERTREATS
Metabolic alkalosis is the most common acid-base disorder in hospitalized patients — and the one most frequently undertreated, because it is often viewed as benign. It is not benign. Severe metabolic alkalosis (pH > 7.55) is associated with increased mortality, arrhythmias, seizures, impaired oxygen delivery (left-shifted oxyhemoglobin dissociation curve), hypokalemia, and ionized hypocalcemia.
Compensation formula for metabolic alkalosis:
Expected PaCO₂ = (0.7 × HCO₃⁻) + 21 ± 2
Classification: Chloride-Responsive vs. Chloride-Resistant
Urine Chloride < 25 mEq/L → Chloride-responsive (most common):
- Vomiting, nasogastric suction (loss of HCl)
- Diuretic use (after the drug’s effect has worn off)
- Post-hypercapnic alkalosis (chronic CO₂ retainer placed on a ventilator and acutely normalized — the kidneys had retained bicarbonate for compensation, and when CO₂ is rapidly corrected, the excess bicarbonate remains)
- Treatment: Volume repletion with normal saline (provides chloride), potassium repletion, and treatment of the underlying cause.
Urine Chloride > 40 mEq/L → Chloride-resistant:
- Primary hyperaldosteronism (Conn syndrome)
- Cushing syndrome
- Bartter syndrome, Gitelman syndrome
- Current diuretic use (while the drug is active)
- Severe hypokalemia (K⁺ < 2.0 mEq/L drives renal H⁺ secretion)
- Licorice ingestion (contains glycyrrhizin, a mineralocorticoid mimic)
- Treatment: Address the underlying cause. Chloride-resistant alkalosis does not respond to saline infusion.
Post-hypercapnic metabolic alkalosis deserves special attention in the ICU. A patient with chronic hypercapnia (e.g., severe COPD with a baseline PaCO₂ of 60) has a compensatory metabolic alkalosis (elevated HCO₃⁻ of approximately 32–35). If this patient is intubated and ventilated to a “normal” PaCO₂ of 40, the elevated bicarbonate is no longer compensatory — it becomes a primary metabolic alkalosis. The pH rises dramatically. This is why the ventilatory management of chronic CO₂ retainers must target the patient’s baseline PaCO₂, not the textbook normal.
PART 6: RESPIRATORY ACID-BASE DISORDERS — THE VENTILATOR INTERFACE
Respiratory Acidosis (Elevated PaCO₂)
Acute: pH drops approximately 0.08 for each 10 mm Hg rise in PaCO₂. HCO₃⁻ rises only 1 mEq/L per 10 mm Hg (chemical buffering only).
Chronic: pH drops approximately 0.03 for each 10 mm Hg rise in PaCO₂. HCO₃⁻ rises 3.5 mEq/L per 10 mm Hg (full renal compensation).
Common ICU causes: Hypoventilation (over-sedation, neuromuscular weakness, obesity hypoventilation), increased dead space (PE, ARDS), severe bronchospasm, and permissive hypercapnia during lung-protective ventilation.
Clinical pearl: In a ventilated patient, the PaCO₂ is a direct function of alveolar ventilation: PaCO₂ = (VCO₂ × 0.863) / VA, where VA = (VT − VD) × RR. If PaCO₂ is rising on a ventilated patient, either CO₂ production has increased (fever, sepsis, malignant hyperthermia), alveolar ventilation has decreased (reduced RR or VT, increased dead space), or both.
Respiratory Alkalosis (Low PaCO₂)
Acute: pH rises approximately 0.08 for each 10 mm Hg fall in PaCO₂. HCO₃⁻ falls 2 mEq/L per 10 mm Hg.
Chronic: pH rises approximately 0.03 for each 10 mm Hg fall in PaCO₂. HCO₃⁻ falls 5 mEq/L per 10 mm Hg.
Common ICU causes: Anxiety, pain, sepsis (early — respiratory alkalosis is often the first acid-base disturbance in sepsis), PE, CNS injury (central neurogenic hyperventilation), hepatic failure, pregnancy, and — as we discussed in our Saturday mythbusting edition — iatrogenic hyperventilation when a clinician increases the respiratory rate inappropriately.
Clinical pearl: Respiratory alkalosis is the only acid-base disorder where full compensation can return pH to normal. In chronic respiratory alkalosis, the pH may be 7.40–7.44 despite a PaCO₂ of 25. Do not assume a normal pH means no acid-base disorder.
PART 7: THE STEWART APPROACH — THE PHYSICOCHEMICAL MODEL
The traditional Henderson-Hasselbalch approach treats bicarbonate as an independent variable — something the body directly controls. The Stewart approach, introduced by Peter Stewart in 1983, argues that bicarbonate and hydrogen ions are dependent variables — determined by three independent variables that the body actually regulates:
1. PaCO₂ — controlled by the lungs 2. Strong Ion Difference (SID) — controlled by the kidneys and IV fluid administration 3. Total weak acids (A_TOT) — primarily albumin and phosphate
The Strong Ion Difference (SID)
SID = (Na⁺ + K⁺ + Ca²⁺ + Mg²⁺) − (Cl⁻ + Lactate⁻)
Normal apparent SID ≈ 40–42 mEq/L
When SID decreases (either by losing strong cations or gaining strong anions), pH falls — acidosis. When SID increases, pH rises — alkalosis.
Clinical applications of SID thinking:
- Normal saline resuscitation reduces SID: saline has a SID of zero (Na⁺ 154 = Cl⁻ 154). Infusing large volumes drives plasma SID toward zero → acidosis. This explains hyperchloremic acidosis mechanistically.
- Lactated Ringer’s has a SID of approximately 28 mEq/L (once lactate is metabolized, the effective SID increases). It is less acidifying than saline because its SID is closer to plasma SID.
- Hyperlactatemia directly reduces SID: lactate is a strong anion. Every mmol/L increase in lactate reduces SID by 1 mEq/L, directly lowering pH — independent of the bicarbonate system.
- Hyperchloremia reduces SID by adding strong anions. Hyponatremia reduces SID by removing strong cations. Both produce acidosis through the same mechanism.
The Strong Ion Gap (SIG)
SIG = SID_apparent − SID_effective
Where SID_effective is calculated from bicarbonate, albumin, and phosphate.
Normal SIG ≈ 0 ± 2 mEq/L
An elevated SIG indicates the presence of unmeasured strong anions — analogous to an elevated anion gap, but corrected for albumin, phosphate, and other weak acids. The SIG has been shown to predict ICU mortality independently of lactate, APACHE scores, and standard anion gap in multiple studies.
When to Use Stewart vs. Traditional
The traditional approach works well for most clinical scenarios. The Stewart approach adds diagnostic value in three specific situations:
- Hypoalbuminemia — the traditional anion gap underestimates unmeasured anions because the “normal” anion gap is lower when albumin is low. The Stewart approach accounts for this automatically through A_TOT.
- Hyperchloremia — the traditional approach identifies hyperchloremic acidosis but does not explain its mechanism. The Stewart approach explains it directly: chloride excess reduces SID, which reduces pH.
- Mixed disorders in complex ICU patients — the Stewart approach detects simultaneous metabolic acidosis and alkalosis that the traditional approach may miss. In a 2022 ICU study, the Fencl-Stewart method revealed frequent simultaneous metabolic acidosis and alkalosis missed by traditional bicarbonate-anion gap analysis.
PART 8: PUTTING IT ALL TOGETHER — THE COMPLETE ABG INTERPRETATION PROTOCOL
Here is the master checklist. Apply it to every ABG.
1. Look at pH → acidemia or alkalemia?
2. Identify the primary disorder → metabolic or respiratory?
3. Check compensation → use the appropriate formula. If compensation is inadequate or excessive → mixed disorder.
4. If metabolic acidosis → calculate the anion gap.
5. Correct the anion gap for albumin.
6. If HAGMA → calculate the delta-delta ratio. Δ-Δ < 1 = concurrent NAGMA. Δ-Δ > 2 = concurrent metabolic alkalosis.
7. If HAGMA with unknown cause → calculate the osmolar gap. Elevated = consider toxic alcohols.
8. If NAGMA → calculate the urine anion gap. Negative = GI loss. Positive = renal cause (RTA).
9. If metabolic alkalosis → check urine chloride. < 25 = chloride-responsive. > 40 = chloride-resistant.
10. If your patient is complex and hypoalbuminemic → consider the Stewart approach (SID, SIG) for unmeasured ion detection.
PART 9: THREE CLINICAL CASES — APPLYING THE FRAMEWORK
Case 1: The Crashing DKA Patient
ABG: pH 7.12, PaCO₂ 22, HCO₃⁻ 7 Lytes: Na⁺ 135, K⁺ 5.8, Cl⁻ 98, BUN 32, Glucose 580, Albumin 3.5
Step 1: pH 7.12 → severe acidemia. Step 2: HCO₃⁻ 7 (very low) → metabolic acidosis is primary. Step 3: Winter’s formula: Expected PaCO₂ = (1.5 × 7) + 8 ± 2 = 18.5 ± 2 = 16.5–20.5. Measured PaCO₂ = 22 → slightly higher than expected → possible concurrent mild respiratory acidosis (patient may be tiring, consider intubation readiness). Step 4: AG = 135 − (98 + 7) = 30 → markedly elevated. Step 5: Corrected AG = 30 + 2.5 × (4.0 − 3.5) = 31.25 → remains elevated (albumin near-normal). Step 6: Δ-Δ = (30 − 12) / (24 − 7) = 18/17 = 1.06 → pure HAGMA. No concurrent metabolic disorder.
Diagnosis: High anion gap metabolic acidosis from diabetic ketoacidosis with borderline inadequate respiratory compensation. Monitor respiratory effort closely — failure to maintain compensatory hyperventilation may signal impending respiratory failure.
Case 2: The Septic Patient on Saline Resuscitation
ABG: pH 7.28, PaCO₂ 28, HCO₃⁻ 13 Lytes: Na⁺ 140, K⁺ 4.2, Cl⁻ 112, Lactate 5.2, Albumin 2.0
Step 1: pH 7.28 → acidemia. Step 2: HCO₃⁻ 13 → metabolic acidosis is primary. Step 3: Winter’s formula: Expected PaCO₂ = (1.5 × 13) + 8 ± 2 = 27.5 ± 2 = 25.5–29.5. Measured PaCO₂ = 28 → appropriate compensation. No concurrent respiratory disorder. Step 4: AG = 140 − (112 + 13) = 15. Step 5: Corrected AG = 15 + 2.5 × (4.0 − 2.0) = 15 + 5 = 20 → significantly elevated (masked by hypoalbuminemia). Step 6: Δ-Δ = (20 − 12) / (24 − 13) = 8/11 = 0.73 → < 1 → concurrent NAGMA in addition to the HAGMA.
Diagnosis: Mixed metabolic acidosis — high anion gap component from lactic acidosis (lactate 5.2, sepsis) PLUS non-anion gap component from hyperchloremic acidosis (Cl⁻ 112, from aggressive normal saline resuscitation). Without the albumin correction, the anion gap of 15 would have appeared near-normal — and the lactic acidosis would have been missed.
This is exactly why albumin correction and the delta-delta ratio matter in the ICU.
Case 3: The Intubated COPD Patient
ABG: pH 7.48, PaCO₂ 35, HCO₃⁻ 30 History: Intubated for acute exacerbation. Baseline PaCO₂ approximately 55. Set on ventilator at RR 18, VT 450 mL.
Step 1: pH 7.48 → alkalemia. Step 2: HCO₃⁻ 30 → elevated → metabolic alkalosis is primary. Step 3: But wait — PaCO₂ is 35 (normal). Is this respiratory compensation? Expected PaCO₂ for metabolic alkalosis: (0.7 × 30) + 21 ± 2 = 42 ± 2 = 40–44. Measured PaCO₂ = 35 → lower than expected → concurrent respiratory alkalosis.
Diagnosis: Post-hypercapnic metabolic alkalosis with iatrogenic respiratory alkalosis. The patient’s kidneys had retained bicarbonate to compensate for chronic hypercapnia (baseline PaCO₂ 55). The ventilator acutely normalized PaCO₂ to 35 — but the excess bicarbonate remained. The ventilator is now causing a combined metabolic and respiratory alkalosis.
Fix: Target the patient’s baseline PaCO₂ of 55 — not 40. Reduce the respiratory rate and/or tidal volume to allow PaCO₂ to rise to the patient’s chronic baseline. The renal bicarbonate excess will correct over 24–48 hours once PaCO₂ returns to its chronic level.
The Bottom Line
Acid-base physiology is not optional knowledge for the critical care clinician. It is the physiological foundation on which ventilator management, fluid resuscitation, electrolyte correction, and hemodynamic optimization all depend. Every ABG tells a story. The clinician who can read that story — who can calculate the anion gap, correct it for albumin, apply the delta-delta ratio, check Winter’s formula, and recognize when the Stewart approach adds diagnostic value — is the clinician who catches the mixed disorder that everyone else misses, who identifies the toxic alcohol that the triage team overlooked, who recognizes that the saline resuscitation is causing a hyperchloremic acidosis on top of the lactic acidosis it was meant to treat.
This is not oversimplified. This is not a clinical pearl. This is the comprehensive, systematic, evidence-based framework that every member of the interprofessional team deserves access to.
I will see you at the bedside.
— Javier Amador-Castaneda, BHS, RRT, FCCM
Medical Disclaimer: The content published in ICCN is intended solely for educational and informational purposes for healthcare professionals. It does not constitute medical advice, clinical guidelines, or a standard of care, and should not be used as a substitute for the independent professional judgment of a licensed clinician. All clinical decisions must be individualized to the patient and made by qualified healthcare providers. ICCN assumes no liability for any clinical outcomes arising from the information presented herein.
© 2026 Interprofessional Critical Care Network (ICCN). All rights reserved. Unauthorized reproduction or redistribution of this content is prohibited. Subscribers may share excerpts with proper attribution to ICCN and the author.
