Venous Blood Gas

Venous blood gas (VBG) analysis measures pH, partial pressure of carbon dioxide (pCO2), bicarbonate (calculated), base excess, lactate, and often electrolytes (sodium, potassium, chloride, ionised calcium), glucose, haemoglobin and haematocrit, on a heparinised venous sample, by a point-of-care or laboratory analyser.

It is quicker and less painful than arterial sampling, can be taken from a peripheral or central vein, and correlates well with arterial values for pH (venous pH is 0.03 to 0.04 lower), bicarbonate (1 to 2 mmol/L higher) and lactate.

Venous pCO2 is 4 to 6 mmHg (0.5 to 0.8 kPa) higher than arterial pCO2, and venous pO2 and oxygen saturation do not reflect arterial oxygenation, so VBG cannot assess hypoxaemia.

VBG is the first line test for acid base assessment in emergency and critical care, diabetic ketoacidosis (VBG replaces arterial puncture for pH and bicarbonate), sepsis (lactate), acute kidney injury, poisoning and respiratory failure (screening for hypercapnia). A venous pCO2 under 6.5 kPa (49 mmHg) excludes clinically important hypercapnia in most patients, and the venous pH and lactate trends direct resuscitation.

Indications. Suspected acidosis or alkalosis. Diabetic ketoacidosis and hyperosmolar states. Sepsis, shock and cardiac arrest (lactate). Acute kidney injury and hyperkalaemia. Poisoning (salicylate, methanol, ethylene glycol, metformin, paracetamol, carbon monoxide). Chronic obstructive pulmonary disease exacerbation and type 2 respiratory failure (pCO2 screening).

Vomiting, diarrhoea and dehydration. Monitoring of non-invasive ventilation. Neonates and children where arterial puncture is difficult. Not for oxygenation: use arterial blood gas or pulse oximetry.

Specimen. 1 to 2 mL of venous blood in a heparinised syringe (balanced lithium heparin, dry electrolyte), taken without prolonged tourniquet use (under 1 minute, as stasis lowers pH and raises lactate and potassium).

Remove all air bubbles, cap the syringe, mix gently and analyse within 10 to 15 minutes at room temperature (within 30 minutes if kept on ice or in a plastic syringe, glass preferred for delayed analysis). Label as venous with the site, the patient's temperature and oxygen therapy. Central venous samples reflect mixed or regional venous blood. Avoid samples drawn upstream of an infusion.

Reference values (venous, adult). pH 7.31 to 7.41. pCO2 5.0 to 6.7 kPa (38 to 50 mmHg). Bicarbonate 22 to 28 mmol/L. Base excess minus 2 to plus 3 mmol/L. Lactate 0.5 to 2.0 mmol/L. Anion gap 8 to 16 mmol/L (without potassium). Potassium 3.5 to 5.0 mmol/L, sodium 135 to 145 mmol/L, chloride 98 to 107 mmol/L, ionised calcium 1.15 to 1.30 mmol/L, glucose 3.9 to 5.8 mmol/L (fasting).

Interpretation.

• Low pH and low bicarbonate: metabolic acidosis. Calculate the anion gap: raised gap (lactic acidosis, ketoacidosis, renal failure, toxic alcohols, salicylate, uraemia) or normal gap (diarrhoea, renal tubular acidosis, saline infusion, acetazolamide).
• Low pH with raised pCO2: respiratory acidosis (hypoventilation, COPD, neuromuscular weakness, sedation, obesity hypoventilation); chronic cases show raised bicarbonate.
• High pH with low pCO2: respiratory alkalosis (anxiety, pain, sepsis, pulmonary embolism, pregnancy, salicylate early, hypoxia, liver disease).
• High pH with raised bicarbonate: metabolic alkalosis (vomiting, diuretics, nasogastric loss, mineralocorticoid excess, hypokalaemia, milk alkali), with urine chloride separating chloride responsive from resistant causes.
• Compensation: expected pCO2 in metabolic acidosis is 1.5 times bicarbonate plus 8 (plus or minus 2). Higher indicates additional respiratory acidosis, and lower indicates additional respiratory alkalosis.
• Delta gap (change in anion gap compared with change in bicarbonate) shows mixed disorders.
• Lactate over 2 mmol/L: hypoperfusion, sepsis, seizures, liver failure, metformin, thiamine deficiency, or regional ischaemia; over 4 mmol/L carries high mortality; failure to clear (less than 10% fall in 2 hours) indicates poor resuscitation.
• Venous to arterial relationship: if pCO2 venous is under 6.5 kPa and pH over 7.35, hypercapnic respiratory failure is excluded; otherwise arterial blood gas is required.
• Electrolyte results (potassium, sodium, glucose, ionised calcium) provide quick bedside information but are confirmed by laboratory measurement when decisions are critical.

Critical values and findings needing immediate action.

• pH under 7.20 (or over 7.55): urgent assessment and treatment of cause; consider bicarbonate only in selected severe acidosis (pH under 7.0 to 7.1 with hyperkalaemia or toxin) as specialist decisions.
• pCO2 over 8 kPa (60 mmHg) with pH under 7.25: acute hypercapnic respiratory failure, controlled oxygen, non-invasive ventilation or intubation.
• Lactate over 4 mmol/L: septic shock or severe hypoperfusion, fluid resuscitation, antibiotics within an hour, source control, and vasopressors.
• Potassium over 6.5 mmol/L or under 2.5 mmol/L: ECG and immediate treatment.
• Glucose under 3.0 mmol/L: hypoglycaemia, immediate glucose; ketoacidosis with pH under 7.0 or ketones over 6 mmol/L: intensive treatment.
• Ionised calcium under 0.9 mmol/L: calcium replacement; high anion gap acidosis with unexplained osmolar gap: toxic alcohol ingestion, fomepizole and dialysis.

Interferences and pitfalls.

• Venous values depend on tissue perfusion, so in shock or cardiac arrest peripheral venous pCO2 is much higher than arterial and pH lower (arterial puncture or central venous gas is better).
• Air bubbles raise pO2 and lower pCO2; delay at room temperature in plastic syringes lowers pH, raises pCO2 and lactate through cell metabolism (marked with leucocytosis or thrombocytosis).
• Prolonged tourniquet, fist clenching and sampling from the arm with infusion cause errors in potassium, lactate and glucose.
• Excess liquid heparin dilutes the sample and lowers pCO2 and bicarbonate.
• Haemolysis falsely raises potassium on whole blood analysis (less detected than in serum).
• Co-oximetry is needed to measure carboxyhaemoglobin and methaemoglobin; conventional pulse oximeters are misleading in carbon monoxide poisoning.
• Whole blood glucose differs from plasma glucose by about 10 to 15%, and haematocrit extremes influence point-of-care results.
• Temperature correction is not used routinely for pH and pCO2 interpretation.
• Different analysers have different reference ranges, and each laboratory should publish its own.
• A normal VBG does not exclude hypoxaemia or early compensated disorders; oxygenation must be assessed separately.

Special groups.

• Children and neonates: heel prick capillary gases give similar pH and pCO2 to venous samples, normal values differ by age (neonatal pH 7.25 to 7.35 in the first hours), and lactate is higher at birth.
• Pregnancy: compensated respiratory alkalosis (pCO2 about 4.0 kPa, bicarbonate 18 to 21 mmol/L) is normal.
• Chronic lung disease: baseline raised pCO2 and bicarbonate, so interpret change from baseline.
• Older adults and renal failure: baseline acidosis and abnormal electrolytes.
• Critically ill patients: central venous or arterial lines allow repeated sampling, and trends matter more than single values.
• Regions: point-of-care analysers are increasingly used in remote and low resource hospitals for sepsis, malaria acidosis, severe malnutrition and obstetric emergencies, where cartridges and quality control are limited.

Follow up tests. Arterial blood gas for oxygenation and when VBG is abnormal or the patient is in shock, serum electrolytes, urea, creatinine and glucose, ketones (beta hydroxybutyrate), serum osmolality (osmolar gap), toxicology screen, salicylate and paracetamol levels, carboxyhaemoglobin, urine chloride, repeat lactate at 2 hours, ECG, chest radiograph, cultures in sepsis, and monitoring of pH and potassium during insulin therapy.