Arterial Blood Gas Sampling

Do you actually need an ABG?

• Venous and arterial pH agree closely (mean difference roughly 0.03–0.04 units) — a VBG pH can safely substitute for arterial pH in most non-shocked ED patients.
• Bicarbonate also agrees well between venous and arterial samples.
• pCO2 agreement is much weaker — limits of agreement of roughly ±20 mmHg mean a single venous pCO2 cannot be used interchangeably with arterial pCO2 when the precise number matters (e.g. borderline hypercapnic respiratory failure). A normal venous pCO2 has good negative predictive value for excluding arterial hypercarbia; an elevated one needs an arterial sample to quantify.
• Combine VBG + pulse oximetry (SpO2) as a first-line non-invasive assessment; reserve arterial puncture for when arterial oxygenation (PaO2) or a precise pCO2 is specifically needed — e.g. significant respiratory distress, suspected hypercapnic failure, or monitoring in critical illness.

The modified Allen test

• Traditionally taught before radial puncture: occlude both radial and ulnar arteries, have the patient clench and open the fist to blanch the hand, then release the ulnar artery and time colour return.
• Evidence for its predictive value is poor — studies have not shown a reliable relationship between an "abnormal" Allen test and subsequent hand ischaemia after radial puncture, and interobserver agreement is modest. It should not be relied upon as a definitive safety gate.
• State this honestly to learners: many departments still teach it as a quick bedside check, but a normal result does not guarantee safety and an abnormal result does not mandate avoiding the radial artery.

Local anaesthesia — evidence-backed and underused

• Small-volume intradermal lidocaine at the puncture site significantly reduces pain from radial arterial puncture in randomised trials, without reducing success rates.
• Despite this evidence, local anaesthesia before arterial puncture is inconsistently used in general (non-anaesthetic) practice — routinely offering it is good, evidence-based care.

Radial technique

1. Position the wrist

   Extend the wrist to roughly 30–45°, e.g. over a rolled towel, to bring the radial artery closer to the surface.

2. Palpate and anaesthetise

   Palpate the artery proximal to the wrist crease; infiltrate a small intradermal bleb of lidocaine if using.

   LAST rescue — AAGBI algorithm ↗

3. Insert at 45°, bevel up

   Advance the needle at roughly 45° to the skin, bevel up, against the direction of flow, watching for a flash of blood.

4. Self-filling syringe

   Use a heparinised self-filling arterial syringe — arterial pressure should fill it without active aspiration in most patients.

5. Withdraw and apply pressure

   Remove the needle and apply firm direct pressure for several minutes (longer if anticoagulated) until haemostasis is confirmed.

Femoral alternative

• Larger, more constant artery — useful when radial access fails, in shock with weak peripheral pulses, or for larger-volume sampling.
• Higher risk of significant bleeding and retroperitoneal haematoma if compression is inadequate; apply firm pressure for longer after puncture.

Sample handling

• Expel any air bubbles immediately — air in the sample falsely alters pO2/pCO2.
• Cap the syringe and mix gently (rolling or a mixing ball) to distribute heparin and prevent clotting.
• Analyse promptly — delay (especially without icing) alters pO2, pCO2 and lactate.

Complications

• Haematoma — commonest; minimised by adequate direct pressure.
• Arterial spasm — can make repeat attempts difficult; avoid multiple attempts at the same site.
• Nerve injury (median nerve at the wrist) — withdraw if the patient reports electric/shooting pain.
• Ischaemia — rare with radial puncture but is the theoretical concern behind the Allen test.

Interpreting the result

For structured blood gas interpretation, see bloodgas.resusdoc.uk.

References