When to Use Arterial Blood Gas Instead of Venous Blood Gas
For critically ill patients or those with shock or hypotension (systolic blood pressure <90 mm Hg), arterial blood gas sampling is mandatory as the initial blood gas measurement 1.
Absolute Indications for Arterial Blood Gas
Critical Illness and Hemodynamic Instability
- Shock states (systolic BP <90 mm Hg)
- Critically ill patients requiring intensive care
- Severe hypoxemia or when accurate oxygenation assessment is essential
The BTS guideline is unequivocal on this point: arterial sampling must be obtained first in these high-acuity scenarios where mortality risk is elevated 1.
When Oxygenation Assessment is Required
Arterial blood gas is essential whenever you need to assess PO2 accurately. Venous blood gas cannot reliably evaluate oxygenation status 2, 3. While pulse oximetry can supplement monitoring, it has significant limitations:
- Normal SpO2 doesn't exclude abnormal pH or PCO2
- Unreliable in patients on supplemental oxygen
- Subject to artifact and errors 1
Recent research confirms that oxygenation parameters (SpO2, PaO2, O2Hb) show weak correlation between arterial and venous samples 4.
Specific Clinical Scenarios Requiring ABG
- Congestive heart failure
- Congenital heart disease
- Neonatal seizures
- Shock states (any etiology)
These conditions showed inappropriately low validity or weak clinical agreement (<20%) when VBG was compared to ABG 5.
When Venous Blood Gas is Acceptable
Acid-Base Status Screening
For most non-critically ill patients requiring blood gas sampling, venous blood gas can reliably assess pH and PCO2 1. The correlation is strong (r > 0.8) for:
- pH (bias 0.044, excellent agreement)
- PCO2 (bias -6.2 mmHg)
- Bicarbonate
- Base excess 6
Specific Conditions Where VBG is Adequate
- Diabetic ketoacidosis - diagnosis can be made with VBG 2, 5
- Respiratory distress syndrome
- Neonatal sepsis
- Renal failure
- Pneumonia
- Status epilepticus 5
Monitoring Therapeutic Response
Once initial arterial sampling confirms the diagnosis in critically ill patients, subsequent monitoring can often use venous samples to track treatment response 6.
Important Caveats and Pitfalls
PCO2 Interpretation
While venous PCO2 correlates with arterial values, the average difference is 8.6 mm Hg with wide limits of agreement (-7.84 to 25.05 mm Hg) 7.
For COPD exacerbations: VBG using a screening cutoff of 45 mm Hg has 100% sensitivity for detecting arterial hypercapnia, but insufficient agreement exists to determine the precise degree of hypercapnia 7.
Severe Metabolic Acidemia
Venous-to-arterial conversion methods (v-TAC) are unreliable in severe metabolic acidemia. One case demonstrated arterial pH 7.07 with venous pH 7.04, but calculated pH incorrectly showed 7.42 8. In severe acidosis, always obtain arterial samples.
Lactate Measurement
Venous lactate shows moderate agreement with arterial values and can be used at the emergency department as an independent prognostic marker for mortality 2. However, venous lactate is consistently higher than arterial (bias -0.07 mEq·L-1) 6.
Practical Algorithm
Start with these questions:
Is the patient critically ill, in shock, or hypotensive? → ABG required
Do you need to assess oxygenation or suspect severe hypoxemia? → ABG required
Does the patient have heart failure, congenital heart disease, or is in shock? → ABG required
Are you screening for acid-base disturbance in a stable patient? → VBG acceptable
Is this follow-up monitoring after initial ABG confirmed the diagnosis? → VBG acceptable
Do you suspect severe metabolic acidemia (pH <7.10)? → ABG required (conversion methods unreliable)
The guideline emphasizes that arterialised earlobe blood gases may be used as an alternative to arterial sampling for pH and PCO2 in non-critically ill patients, though PO2 underestimates by 0.5-1 kPa 1.