6.3 Arterial Blood Gas (ABG) Analysis, Acid-Base Interpretation & Capnography in Critical Care
Key Takeaways
Assess pH, PaCO₂ and bicarbonate with the child’s baseline and respiratory support. A near-normal pH can occur with mixed disorders.
In metabolic acidosis, compare measured PaCO₂ with Winter’s estimate of 1.5 × bicarbonate + 8 mm Hg, with a range of about ±2 mm Hg.
Calculate the anion gap from simultaneous serum chemistry. Use the laboratory interval and consider albumin, lactate, ketones and treatment effects.
Phase I is early expiration. EtCO₂ is measured at the end of expiration, before fresh inspired gas produces the downstroke.
A waveform change calls for a patient, airway and circuit check. During CPR, interpret EtCO₂ trends with the rest of the clinical assessment.
Arterial blood gases and capnography
Use the arterial blood gas to assess oxygenation, ventilation and acid-base status. Use capnography for continuous trend and waveform information. When a result and the bedside picture disagree, check the specimen or sensor, current respiratory support and the child's trajectory.
Interpret an arterial blood gas in sequence
- Read the pH. The usual arterial interval is about 7.35–7.45. Apply the local laboratory's age- and specimen-specific ranges. A near-normal pH can occur when two processes move pH in opposite directions.
- Read the measured PaCO₂ and the calculated bicarbonate. Rising PaCO₂ drives pH down. Falling bicarbonate also drives pH down. In alkalemia, look for low PaCO₂ or high bicarbonate. Use the history and serial values if the primary process is unclear. Use the contemporaneous chemistry panel for the anion gap. Its total CO₂ is a measured surrogate for bicarbonate.
- Compare the second value with the expected response. An unexpectedly large or small response raises concern for another process. Check timing, the child's baseline, kidney function and whether ventilator settings determine PaCO₂. The estimates describe physiology and should not be used as ventilator targets.
- Check oxygenation in context. Interpret PaO₂ with inspired oxygen, support and the underlying disease. An acceptable PaO₂ on supplemental oxygen can coexist with inadequate ventilation or substantial oxygenation impairment.
| Primary process | Expected response and interpretation |
|---|---|
| Metabolic acidosis | Winter's formula estimates expected PaCO₂ in mm Hg as , with a range of roughly ±2 mm Hg. PaCO₂ above the range suggests additional respiratory acidosis and PaCO₂ below it suggests additional respiratory alkalosis. Use bicarbonate in mEq/L. |
| Metabolic alkalosis | PaCO₂ usually rises with bicarbonate through hypoventilation. The response varies. A falling PaCO₂ during alkalemia despite elevated bicarbonate supports a concurrent respiratory alkalosis. |
| Respiratory acidosis | As rough estimates, expect bicarbonate to rise by about 1 mEq/L per 10 mm Hg PaCO₂ rise acutely and by about 4–5 mEq/L per 10 mm Hg in the chronic steady state with functioning kidneys. Compare with prior values when available. |
| Respiratory alkalosis | For a 10 mm Hg fall in PaCO₂, bicarbonate falls roughly 2 mEq/L acutely and about 4 mEq/L in the chronic steady state. Reduced renal net acid excretion contributes to the later response. An unexpectedly low bicarbonate for the time course prompts evaluation for a metabolic acidosis. |
For example, when bicarbonate is 12 mEq/L, expected PaCO₂ is mm Hg, or about 24–28 mm Hg. A PaCO₂ of 36 mm Hg suggests an additional respiratory acidosis even though it is close to the usual laboratory interval. For respiratory acidosis with PaCO₂ of 60 mm Hg and prior values of 40 mm Hg and 24 mEq/L, bicarbonate near 26 mEq/L fits an acute rise and about 32–34 mEq/L fits a sustained rise. These are rough estimates from physiologic studies. Compare with prior blood gases and allow for ongoing treatment, time course and kidney disease. Children's Hospital of Philadelphia laboratory ranges · Metabolic acidosis compensation study · Metabolic alkalosis study · Acute and chronic hypercapnia studies and prospective chronic study · Acute hypocapnia study and sustained hypocapnia study
Classify metabolic acidosis with the anion gap
Using a chemistry panel and omitting potassium, calculate . Compare with the laboratory's reference interval for that method. Low albumin can mask an elevated gap. A commonly used estimate is . For an observed gap of 12 mEq/L with albumin of 2 g/dL, the corrected estimate is 17 mEq/L. Use the history and targeted measurements such as lactate, beta-hydroxybutyrate, kidney function and toxicology tests to determine the cause. Anion gap and albumin study · ISPAD pediatric DKA guidance
| Pattern | Bedside reasoning |
|---|---|
| Elevated gap | Consider ketoacidosis, lactic acidosis, reduced renal acid clearance, organic acidemias and selected toxic ingestions. Lactate can rise through multiple mechanisms, including poor perfusion and adrenergic stimulation. Examine perfusion and use trends and the rest of the workup. Salicylate poisoning can add a primary respiratory alkalosis. |
| Normal gap with low bicarbonate | The fall in bicarbonate accompanies a relative rise in chloride. Consider diarrhea or other gastrointestinal bicarbonate losses, impaired renal acid handling and chloride-rich fluid exposure. Distal renal tubular acidosis can cause an inappropriately high urine pH during systemic acidosis and is associated with nephrocalcinosis. Proximal tubular bicarbonate wasting can accompany Fanconi syndrome. Hyperkalemic RTA is associated with impaired aldosterone action and reduced ammonium excretion. Interpret urine results with kidney function, volume status and medication exposure. |
| Change during treatment | A child treated for DKA can have falling ketones but persistent hyperchloremic acidosis. Follow beta-hydroxybutyrate when available. In a child resuscitated for diarrhea, both stool bicarbonate loss and chloride-rich fluid can contribute. Reassess perfusion, renal function and ongoing losses. |
An increase in anion gap substantially smaller than the drop in bicarbonate can suggest an additional normal-gap acidosis. A gap rise larger than the bicarbonate drop can suggest coexisting alkalosis or a high baseline bicarbonate. For an elevated corrected gap of 24 and bicarbonate of 12 mEq/L, with baselines of 12 and 24 respectively, the changes are both 12 mEq/L and the ratio is 1. That supports a substantial gap acidosis. Factors such as prior hypercapnia, albumin, timing and renal excretion of organic anions affect the comparison, so use it as a clue to investigate. Observed acid-base patterns in DKA · ERKNet/ESPN RTA guidance · Pediatric crystalloid trial · Organic acidemias guideline · AHA poisoning guidance
Read the capnogram and verify the cause
Capnography shows exhaled carbon dioxide over time. Phase I is early expiration and represents gas from conducting airways with little CO₂. Phase II is the rising transition as alveolar gas reaches the sensor. Phase III reflects predominantly alveolar gas and may slope upward as regions empty at different rates. The end-expiratory reading is EtCO₂. The downstroke at the onset of inspiration marks wash-in of fresh gas. Teaching diagrams may label the inspiratory downstroke phase 0 or phase IV. During inspiration the trace ordinarily returns near the baseline when inspired gas contains little CO₂.
EtCO₂ depends on CO₂ production, delivery by pulmonary blood flow, alveolar ventilation and the quality of gas sampling. For example, with a simultaneous PaCO₂ of 52 mm Hg and EtCO₂ of 32 mm Hg, the measured gradient is 20 mm Hg. Increased physiologic dead space, reduced pulmonary perfusion, circuit leaks and sampling conditions can contribute to a discrepancy. Use paired readings when needed and reassess if the child's circulation or lung mechanics change. AARC capnography guideline · Society for Pediatric Sedation capnogram figure · Pediatric paired CO₂ studies and PICU gradient variation
| Finding | Interpretation and action |
|---|---|
| Slow upstroke and sloping plateau | Uneven emptying can produce a shark-fin appearance. Evaluate bronchospasm, secretions and tube or circuit obstruction and compare the response to treatment. |
| Downward notch in the plateau | A curare cleft may reflect an inspiratory effort during controlled ventilation. Correlate with chest movement, ventilator graphics, pulse timing and the clinical assessment of blockade. Reassess comfort, synchrony and the ongoing indication for blockade before changing it. |
| Elevated inspired baseline | Check for rebreathing and sampling error. Inspect circuit valves and, when applicable, fresh gas flow or CO₂ absorber function. |
| Sudden disappearance | Assess the child, ventilation and circulation immediately. Check tube position and patency, circuit connections and the sampling line. If circulation is lost, begin or continue resuscitation. |
| Change during CPR | Follow waveform and trends alongside compression assessment, ventilation and arterial pressure when available. A sudden sustained rise can suggest return of spontaneous circulation. Verify with the rhythm, arterial waveform or an appropriate pulse check. A single EtCO₂ cutoff should never determine when to stop pediatric resuscitation. |
AARC waveform and airway monitoring · 2025 AHA/AAP pediatric advanced life support
A 5-year-old with septic shock from perforated appendicitis has an arterial pH of 7.23 and PaCO₂ of 22 mm Hg. Simultaneous chemistry shows sodium 138, chloride 101 and total CO₂ 9 mEq/L. Albumin is 4.0 g/dL, lactate is 8 mmol/L and the laboratory’s anion gap reference interval is 8–12 mEq/L. What does the anion gap and expected respiratory response show?
Primary respiratory alkalosis with an anion gap of 9 mEq/L.
Normal anion gap metabolic acidosis because the bicarbonate is low.
High anion gap metabolic acidosis with PaCO₂ in the expected compensation range.
High anion gap metabolic acidosis with additional respiratory acidosis because 22 mm Hg exceeds the expected range.
A ventilated 9-year-old who received rocuronium during intubation now has a discrete downward notch in the expiratory capnogram plateau. The clinician sees a brief inspiratory effort with the notch, which is not synchronized to the pulse. Inspired CO₂ remains near zero. Which pattern and response fit?
Obstructive shark-fin pattern. Give continuous albuterol based on the notch.
Cardiogenic oscillations. Change the trigger based on pulse-synchronous deflections.
CO₂ rebreathing. Replace the expiratory valve based on the inspired baseline.
A curare cleft associated with inspiratory effort. Assess comfort, synchrony and the ongoing need for blockade.
A 2-year-old with profuse diarrhea received several 0.9% saline boluses with reassessment for hypovolemic shock. Perfusion has improved and lactate has fallen to 1.2 mmol/L. Arterial pH is 7.29 and PaCO₂ is 32 mm Hg. Simultaneous chemistry shows sodium 140, chloride 115 and total CO₂ 15 mEq/L with albumin 4.0 g/dL. The laboratory’s anion gap interval is 8–12 mEq/L. What most likely explains the persistent acidemia?
High anion gap lactic acidosis caused by persistent severe hypoperfusion.
Normal anion gap metabolic acidosis, with gastrointestinal bicarbonate loss and chloride from saline both plausible contributors.
Diabetic ketoacidosis established by the low serum bicarbonate.
Proximal renal tubular acidosis established by the diarrheal illness.
Sections you finish are checked off in the contents.