How should an arterial blood gas (ABG) result be interpreted in a critically ill patient in the intensive care unit (ICU)?

Medical Advisory BoardAll articles are reviewed for accuracy by our Medical Advisory Board
Educational purpose only • Exercise caution as content is pending human review
Article Review Status
Submitted
Under Review
Approved

Last updated: January 28, 2026View editorial policy

Personalize

Help us tailor your experience

Which best describes you? Your choice helps us use language that's most understandable for you.

Arterial Blood Gas (ABG) Interpretation in the ICU Setting

Core Sampling Protocol

In critically ill ICU patients, arterial blood samples from an indwelling arterial catheter should be your first-line approach for ABG analysis, with blood gas analyzers as the default measurement device to ensure accuracy and guide life-sustaining interventions. 1, 2

Sampling Hierarchy by Patient Acuity

  • For patients with invasive vascular monitoring (shock, vasopressors, severe edema, mechanical ventilation): Draw all samples from arterial lines; if temporarily unavailable, use central venous lines; never use capillary finger-stick samples as they are inaccurate in this population 1

  • For ECMO patients specifically: Obtain samples from the right radial arterial line, as this best represents cerebral perfusion and helps identify differential oxygenation syndromes 2

  • For less critically ill patients without invasive monitoring: Capillary samples may be acceptable, but arterial puncture with local anesthesia (after Allen's test for radial approach) remains preferred 1, 3

Critical Timing Requirements

  • Obtain ABG within 60 minutes of starting oxygen therapy 4
  • Repeat ABG within 60 minutes of any FiO₂ change, especially in patients at risk for hypercapnic respiratory failure (COPD, obesity hypoventilation) 1, 4
  • After return of spontaneous circulation post-cardiac arrest, obtain ABG immediately to guide oxygen therapy 2

Systematic ABG Interpretation Framework

Step 1: Assess Oxygenation Status

  • PaO₂ provides information that pulse oximetry cannot capture - normal SpO₂ does not exclude significant hypoxemia, acid-base disturbances, or hypercapnia 2, 3, 4

  • Target SpO₂ of 88-92% in all patients with acute hypercapnic respiratory failure, whether spontaneously breathing or on NIV 1

  • In patients on supplemental oxygen, ABG is mandatory even with normal saturation, as oxygen masks can hide underlying abnormalities 4

Step 2: Evaluate Ventilation (PaCO₂)

  • PaCO₂ reflects adequacy of alveolar ventilation and cannot be reliably estimated from venous samples in shock states 2, 5

  • Rising PaCO₂ with acidosis indicates respiratory failure requiring ventilatory support consideration 2

  • In stable patients, PaCO₂ can vary spontaneously by 3.0 ± 1.9 mm Hg, so base decisions on trends rather than isolated values 6

Step 3: Analyze Acid-Base Status Using the RoMe Technique

pH Analysis:

  • pH < 7.35 = acidemia
  • pH > 7.45 = alkalemia
  • pH 7.35-7.45 = normal or fully compensated 7

Identify Primary Disorder:

  • "Respiratory opposite": If pH and PaCO₂ move in opposite directions (pH↓ with PaCO₂↑, or pH↑ with PaCO₂↓), the primary disorder is respiratory 7
  • "Metabolic equal": If pH and HCO₃⁻ move in the same direction (both↓ or both↑), the primary disorder is metabolic 7

Assess Compensation:

  • Uncompensated: Only one system (respiratory or metabolic) is abnormal
  • Partially compensated: Both systems abnormal, pH still outside normal range
  • Fully compensated: Both systems abnormal, pH normalized 7

Step 4: Identify Specific Acid-Base Disorders

Metabolic Acidosis (pH↓, HCO₃⁻↓):

  • In cardiogenic shock, associated with poor outcomes and requires urgent intervention 2
  • Check lactate levels simultaneously to assess tissue perfusion 2

Respiratory Acidosis (pH↓, PaCO₂↑):

  • May indicate impending respiratory failure requiring NIV or intubation 2
  • In COPD patients on oxygen, suggests need for ventilatory support rather than just increased FiO₂ 1

Mixed Disorders:

  • Common in ICU patients with multiple organ dysfunction 8, 9
  • Require assessment of both respiratory and metabolic components independently 9

Technical Accuracy Considerations

Device Selection and Standards

  • Blood gas analyzers (central lab or ICU-based) must perform to ±0.4 mmol/L or ±8% above 5 mmol/L for critically ill patients 1, 4

  • Point-of-care glucose meters have significant limitations in ICU patients with anemia, hypoxia, shock, or on vasopressors - arterial blood gas analyzers are more reliable 1

Common Sources of Error

  • Hypotension, shock, vasopressors: Cause slow glucose equilibration and unreliable capillary samples; arterial/venous sampling mandatory 1

  • Contamination from IV fluids: Use only 0.9% sodium chloride (with or without heparin) for arterial line flush; avoid sampling from multilumen catheters with running infusions 3, 4

  • Physiologic variability: Even in stable ICU patients, PaO₂ can vary by 16.2 ± 10.9 mm Hg spontaneously over 50 minutes 6


Critical Pitfalls to Avoid

The "Normal SpO₂" Trap

  • Never assume adequate gas exchange based solely on pulse oximetry - patients can have normal SpO₂ with severe acidosis, hypercapnia, or anemia 2, 3, 4

  • Pulse oximetry measures oxygen saturation but provides no information about ventilation (PaCO₂) or acid-base status (pH) 2, 3

Sampling Site Errors

  • Never use capillary samples in patients with shock, on vasopressors, or with severe edema - these produce factitious results that can lead to dangerous treatment errors 1

  • In mechanically ventilated or insulin-infused patients, arterial or venous whole blood sampling is mandatory 1

Interpretation Errors

  • Base therapeutic decisions on trends, not isolated values - spontaneous variation occurs even in stable patients 6

  • At extremes of blood glucose concentration (<75 or >200 mg/dL), verify with central laboratory to avoid insulin dosing errors 1


Special ICU Populations

Patients on Non-Invasive Ventilation (NIV)

  • Continuous SpO₂ monitoring is essential, but repeated ABG measurement required to assess PaCO₂ and pH 1

  • Can use capillary sampling or intermittent arterial puncture, though arterial lines allow safer repeated sampling 1

  • Transcutaneous CO₂ monitoring may supplement but not replace ABG analysis 1

Acute Heart Failure

  • ABG helps differentiate cardiac versus pulmonary causes of respiratory distress 2

  • Assess effectiveness of CPAP therapy through serial ABG analysis showing improved oxygenation and reduced work of breathing 2

Post-Cardiac Arrest

  • Immediate ABG after ROSC is mandatory to guide oxygen therapy and avoid hyperoxia 2

  • Serial measurements guide ventilator management and assess adequacy of resuscitation 2

References

Guideline

Guideline Directed Topic Overview

Dr.Oracle Medical Advisory Board & Editors, 2025

Guideline

Role of Arterial Blood Gas Analysis in CVICU Management

Praxis Medical Insights: Practical Summaries of Clinical Guidelines, 2025

Guideline

Arterial Blood Gas and Venous Blood Gas Analysis

Praxis Medical Insights: Practical Summaries of Clinical Guidelines, 2025

Guideline

Arterial Blood Gas Analysis in Critically Ill Patients

Praxis Medical Insights: Practical Summaries of Clinical Guidelines, 2026

Research

Analysing arterial blood gas results using the RoMe technique.

Nursing standard (Royal College of Nursing (Great Britain) : 1987), 2024

Research

Interpretation of arterial blood gas.

Indian journal of critical care medicine : peer-reviewed, official publication of Indian Society of Critical Care Medicine, 2010

Professional Medical Disclaimer

This information is intended for healthcare professionals. Any medical decision-making should rely on clinical judgment and independently verified information. The content provided herein does not replace professional discretion and should be considered supplementary to established clinical guidelines. Healthcare providers should verify all information against primary literature and current practice standards before application in patient care. Dr.Oracle assumes no liability for clinical decisions based on this content.

Have a follow-up question?

Our Medical A.I. is used by practicing medical doctors at top research institutions around the world. Ask any follow up question and get world-class guideline-backed answers instantly.