Introduction
Arterial blood gas (ABG) sampling is an essential clinical skill for all junior doctors. It provides real-time information about a patient's respiratory and metabolic status and is indispensable in the assessment and management of acutely unwell patients. An ABG can be obtained within minutes and interpreted at the bedside, making it one of the most powerful rapid diagnostic tools available.
Indications for ABG Sampling
- Acute respiratory failure or respiratory distress
- Assessment of oxygenation and ventilation (particularly in patients on oxygen therapy or NIV/CPAP)
- Metabolic disturbances — acidosis, alkalosis, electrolyte derangement
- Monitoring patients in intensive care or high-dependency units
- Titrating oxygen therapy (particularly in COPD patients at risk of hypercapnia)
- Monitoring cardiorespiratory resuscitation
- Unexplained altered consciousness
Equipment Required
- Pre-heparinised ABG syringe (most hospitals provide purpose-made kits containing syringe, needle, and cap)
- Needle (23G or 25G — a smaller gauge reduces patient discomfort)
- Gloves (non-sterile)
- Alcohol prep wipe (2% chlorhexidine in 70% alcohol)
- Sterile gauze or cotton wool
- Adhesive plaster or dressing
- Sharps bin positioned within reach
- Ice (if analysis machine is not immediately nearby — sample must be analysed within 15 minutes, or kept in ice for up to 30 minutes)
- Patient label for the syringe
Anatomy — Arterial Sites
Radial Artery (Preferred Site)
The radial artery is the preferred site for ABG sampling in most clinical settings. It is located at the wrist, running between the flexor carpi radialis tendon (medially) and the radial styloid process (laterally). It is palpated at the distal wrist crease on the lateral (thumb) side. Its advantages include: superficial and easily accessible; dual blood supply to the hand (with the ulnar artery via the palmar arch — hence, inadvertent occlusion of the radial artery does not cause hand ischaemia if the ulnar collateral circulation is adequate); and good patient tolerance.
Ulnar Artery
The ulnar artery runs on the medial side of the wrist and provides the dominant blood supply to the palmar arch in most people. It is technically more difficult to cannulate and is adjacent to the ulnar nerve — it is therefore reserved as a second option only. Do not sample from both the radial and ulnar artery of the same hand.
Brachial Artery
The brachial artery lies in the antecubital fossa and is sometimes used when radial access is not possible. There is a higher risk of complications as it is an end-artery in the arm — there is no adequate collateral circulation, so any prolonged spasm or thrombosis could compromise the distal limb.
Femoral Artery
Used in emergency situations only. Located at the mid-inguinal point (midway between the anterior superior iliac spine and the pubic symphysis). Risk of haematoma formation and femoral nerve damage is higher; prolonged pressure is required after sampling.
Allen's Test — Assessing Collateral Circulation
Allen's test is performed before radial artery sampling to confirm that ulnar collateral flow is adequate. It should be performed routinely, particularly before arterial line insertion.
- Ask the patient to make a fist and elevate their hand. Occlude both the radial and ulnar arteries simultaneously using your thumbs.
- Ask the patient to open their fist — the palm should appear pale (blanched).
- Release the ulnar artery while maintaining pressure on the radial artery.
- Observe the hand for colour return. If the palm flushes pink within 5–7 seconds, the test is positive, indicating adequate ulnar collateral circulation — it is safe to proceed with radial artery sampling.
- If flushing takes more than 15 seconds or does not occur, the test is negative — avoid radial artery sampling on that side and select an alternative site.
Technique — Step-by-Step
Assemble your ABG kit. Confirm the syringe is pre-heparinised (or load the heparin as instructed if using a manual syringe). Expel any excess heparin, leaving only enough to coat the syringe barrel — excess heparin dilutes the sample and artefactually lowers the pCO2 and bicarbonate.
Position the patient: The patient should be sitting or lying comfortably. Hyperextend the wrist over a towel or small pillow to bring the radial artery closer to the surface. This makes palpation and cannulation easier. Clean the skin with an alcohol wipe and allow to dry. Local anaesthetic (1% lidocaine subcutaneously) may be used to reduce discomfort — this is good practice, particularly in awake patients.
Locate the pulse: Using the index and middle fingers of your non-dominant hand, palpate the radial pulse. Identify the point of maximal pulsation. Position your fingers on either side of the artery to guide the needle accurately and prevent the artery from rolling.
Insert the needle: Hold the ABG syringe like a pen in your dominant hand. Insert the needle at approximately 45 degrees to the skin, directly over the point of maximal pulsation, bevel up. Advance slowly in small increments. If you are in the artery, the pulsatile pressure will cause blood to fill the syringe automatically without needing to aspirate — this is the hallmark of arterial (as opposed to venous) blood. The blood is typically brighter red than venous blood.
Collect the sample: Allow the syringe to fill to the 1–2 mL mark through arterial pressure alone. Then withdraw the needle smoothly and immediately apply firm pressure to the puncture site using dry gauze. The patient or a colleague should maintain firm pressure for a minimum of 5 minutes (or longer in patients on anticoagulants — aim for 10–15 minutes). Never apply a pressure dressing as a substitute for adequate manual compression time.
Prepare the sample: Remove any air bubbles from the syringe immediately — air will equilibrate with the sample and falsely raise the pO2. Cap the syringe, label it with the patient's details, and note the time, FiO2 (fraction of inspired oxygen), and temperature. Analyse the sample within 15 minutes or place on ice for up to 30 minutes.
Dispose of sharps: Discard the needle directly into the sharps bin. Never re-sheathe the needle.
Normal ABG Values (at sea level, breathing room air)
| Parameter | Normal Range | What it measures |
|---|---|---|
| pH | 7.35 – 7.45 | Acid-base status of the blood |
| pCO2 | 4.7 – 6.0 kPa (35–45 mmHg) | Partial pressure of CO2 — respiratory component |
| pO2 | 10.6 – 14.0 kPa (80–100 mmHg) | Partial pressure of oxygen — adequacy of oxygenation |
| HCO3- | 22 – 26 mmol/L | Bicarbonate — metabolic component (renal regulation) |
| Base Excess (BE) | -2 to +2 mmol/L | Overall metabolic component; negative = metabolic acidosis |
| SaO2 | > 95% | Oxygen saturation of haemoglobin |
Systematic Approach to ABG Interpretation
Use the following stepwise approach for every ABG to avoid missing abnormalities:
Step 1 — Is the Patient Adequately Oxygenated?
Look at the pO2 (and SpO2). Is it acceptable given the fraction of inspired oxygen (FiO2)?
- On room air (FiO2 = 0.21): pO2 should be at least 10 kPa in a young adult; above 8 kPa is a minimum acceptable threshold
- A simple rule: on a given FiO2, the pO2 should approximately equal FiO2 x 70 (in kPa) in a patient with healthy lungs. For example, on 40% oxygen (FiO2 0.4), expected pO2 = 28 kPa; a measured value of 10 kPa would indicate significant impairment
- pO2 below 8 kPa = type 1 or type 2 respiratory failure
Step 2 — What is the pH?
- pH < 7.35 = acidosis
- pH > 7.45 = alkalosis
- pH 7.35–7.45 = normal (but compensation may have occurred — check pCO2 and HCO3)
Step 3 — What is the Respiratory Component (pCO2)?
- pCO2 > 6.0 kPa with acidosis = respiratory acidosis (CO2 retention — hypoventilation)
- pCO2 < 4.7 kPa with alkalosis = respiratory alkalosis (hyperventilation — anxiety, pain, pulmonary embolism)
- The respiratory system compensates rapidly (within minutes to hours) for metabolic disturbances by adjusting ventilation
Step 4 — What is the Metabolic Component (HCO3 and Base Excess)?
- HCO3 < 22 mmol/L with BE < -2 = metabolic acidosis (e.g. diabetic ketoacidosis, lactic acidosis, renal failure, diarrhoea)
- HCO3 > 26 mmol/L with BE > +2 = metabolic alkalosis (e.g. vomiting, diuretic use, Conn's syndrome)
- The kidneys compensate more slowly for respiratory disturbances (over hours to days) by retaining or excreting bicarbonate
Step 5 — Is There Compensation?
In a compensated disturbance, the pH is normal (or near-normal) because the secondary system has corrected it. However, the primary abnormality and compensatory response are still visible:
- Compensated respiratory acidosis: Low pH (or normal if fully compensated), elevated pCO2, elevated HCO3 (kidney retaining bicarbonate to buffer the acid)
- Compensated metabolic acidosis: Low pH (or normal), low HCO3, low pCO2 (lungs blowing off CO2 — Kussmaul breathing)
Common ABG Patterns — Clinical Summary
| Condition | pH | pCO2 | HCO3 | BE | pO2 |
|---|---|---|---|---|---|
| Respiratory acidosis (COPD exacerbation) | Low | High | Normal/High | +ve | Low |
| Respiratory alkalosis (hyperventilation, PE) | High | Low | Normal/Low | -ve | Normal/Low |
| Metabolic acidosis (DKA, lactic acidosis) | Low | Low | Low | -ve | Normal |
| Metabolic alkalosis (vomiting, diuretics) | High | High | High | +ve | Normal |
| Type 1 respiratory failure (pneumonia, PE) | Normal/High | Low/Normal | Normal | Normal | Low (<8 kPa) |
| Type 2 respiratory failure (COPD, neuromuscular) | Low | High | High | +ve | Low |
Type 1 vs Type 2 Respiratory Failure
- Type 1 (hypoxaemic) respiratory failure: pO2 <8 kPa with normal or low pCO2. The patient has impaired gas exchange (V/Q mismatch or shunt) but is still able to increase ventilation to blow off CO2. Causes: pneumonia, pulmonary oedema, ARDS, PE, asthma.
- Type 2 (hypercapnic/ventilatory) respiratory failure: pO2 <8 kPa with pCO2 >6 kPa. The patient cannot maintain adequate alveolar ventilation. Causes: COPD exacerbation, acute severe asthma (exhaustion), neuromuscular disease (Guillain-Barré, MND), chest wall deformity, obesity hypoventilation syndrome.
Oxygen Therapy in COPD
In patients with type 2 respiratory failure due to COPD, oxygen therapy must be carefully controlled. Some COPD patients have a chronic, compensated respiratory acidosis (chronically elevated pCO2 with high bicarbonate). These patients may depend on their hypoxic drive to breathe. Administering high-flow oxygen can suppress this drive, causing further CO2 retention and worsening acidosis.
The British Thoracic Society (BTS) recommends targeting SpO2 of 88–92% in patients with known or suspected type 2 respiratory failure, using controlled low-flow oxygen (24–28% via Venturi mask). Regular ABG monitoring is essential to guide therapy.
Frequently Asked Questions
Why must air bubbles be removed from an ABG syringe?
Air bubbles in the syringe will equilibrate with the blood sample over time. Since the pO2 of atmospheric air (~21 kPa) is higher than most patients' arterial pO2, air contamination will falsely elevate the measured pO2 — which could give a falsely reassuring result in a hypoxic patient. Conversely, air has a lower pCO2 than blood, so air contamination will falsely lower the pCO2 reading. Bubbles must be expelled immediately after sampling by tilting the syringe with the needle pointing upward and tapping the barrel to collect air at the tip, then expelling it carefully.
What is the difference between a venous blood gas and an arterial blood gas?
A venous blood gas (VBG) is taken from a peripheral vein (usually the antecubital fossa) or a central line, and is much easier and less painful to obtain than an ABG. A VBG gives useful information about pH, bicarbonate, base excess, lactate, and electrolytes — making it suitable for assessing metabolic disturbances and for calculating the anion gap. However, the venous pO2 reflects tissue oxygen extraction (not lung function) and the venous pCO2 is approximately 5–6 mmHg higher than arterial. VBGs should not be used to assess oxygenation or ventilation — for these, a true ABG is required. In clinical practice, a VBG is now used in many acute scenarios as an initial screen, with an ABG reserved for when respiratory function assessment is the primary concern.
What is the anion gap and when should I calculate it?
The anion gap (AG) = Na+ - (Cl- + HCO3-). Normal range: 8–16 mmol/L (some labs use 8–12 depending on albumin correction). An elevated anion gap indicates the presence of unmeasured anions in the blood and is always calculated when a metabolic acidosis is present. Causes of a high anion gap metabolic acidosis can be remembered with the mnemonic MUDPILES: Methanol, Uraemia, Diabetic ketoacidosis, Paracetamol/Propylene glycol, Isoniazid/Iron, Lactic acidosis, Ethylene glycol, Salicylates. A normal anion gap metabolic acidosis (hyperchloraemic) has different causes including diarrhoea, renal tubular acidosis, Addison's disease, and ureteric diversion.
What is the oxygen-haemoglobin dissociation curve and why does it matter?
The oxygen-haemoglobin dissociation curve describes the relationship between the partial pressure of oxygen (pO2) and haemoglobin saturation (SaO2). It is sigmoid-shaped (S-shaped). Key clinical points: the upper flat portion means that above pO2 ~10 kPa, SaO2 remains high even with further increases in pO2 (meaning pulse oximetry gives little warning of early respiratory deterioration in a patient receiving supplemental oxygen). The steep portion below pO2 ~8 kPa means that small falls in pO2 cause large drops in SaO2. Factors shifting the curve to the right (decreasing Hb-O2 affinity — more O2 released to tissues): acidosis, fever, hypercapnia, 2,3-DPG. Factors shifting left (increased Hb-O2 affinity): alkalosis, hypothermia, carbon monoxide, fetal haemoglobin.
How do I know whether an acid-base disturbance is compensated or a mixed disorder?
After identifying the primary disturbance, calculate the expected compensation: For metabolic acidosis: expected pCO2 = 1.5 x HCO3 + 8 (±2) — "Winter's formula". If the measured pCO2 is lower than expected, there is a concurrent respiratory alkalosis. For metabolic alkalosis: expected pCO2 = 0.7 x HCO3 + 21 (±2). For respiratory acidosis (acute): for each 1 kPa rise in pCO2, HCO3 rises by ~1 mmol/L. For chronic respiratory acidosis: for each 1 kPa rise in pCO2, HCO3 rises by ~3.5 mmol/L. If the actual bicarbonate is higher than expected for the degree of CO2 retention, there is a concurrent metabolic alkalosis. These calculations require practice — the key is to always cross-check whether the compensation seems appropriate for the primary disturbance.
What are the complications of radial artery puncture?
Complications of radial artery puncture for ABG sampling are uncommon when performed correctly. They include: local haematoma formation (most common; prevented by adequate pressure for 5 minutes post-procedure); arterial spasm (self-limiting; avoid repeated attempts in the same site); infection (rare if aseptic technique is used); thrombosis of the radial artery (very rare; of concern in patients with inadequate collateral flow — hence the importance of Allen's test); arteriovenous fistula or pseudoaneurysm formation (rare); nerve injury (rare; particularly at the wrist near the superficial branch of the radial nerve). Pain during the procedure is common and can be reduced significantly with local anaesthetic.
What oxygen saturation target should I use in different patient groups?
The British Thoracic Society (BTS) Emergency Oxygen Guidelines recommend: For most acutely ill patients — target SpO2 of 94–98%. For patients with known or suspected type 2 respiratory failure (COPD, neuromuscular disease, obesity hypoventilation) — target SpO2 of 88–92%, using controlled oxygen via Venturi mask (24–28%). For patients at risk of hypercapnia but whose target range has not yet been established — start at 28% oxygen via Venturi mask, take an ABG after 30–60 minutes, and adjust the target based on the result. Never use high-flow oxygen (15 L/min via non-rebreather mask) in a patient with chronic CO2 retention without seeking senior review.
References
- O'Driscoll BR, Howard LS, Earis J, Mak V; British Thoracic Society Emergency Oxygen Guideline Group. BTS guideline for oxygen use in adults in healthcare and emergency settings. Thorax. 2017;72(Suppl 1):ii1–ii90.
- National Institute for Health and Care Excellence. Acutely ill adults in hospital: recognising and responding to deterioration. NICE Clinical Guideline CG50. NICE, 2007 (updated 2020).
- Thomas CP, Mayer BN. Arterial blood gas. In: StatPearls [Internet]. StatPearls Publishing, 2023. ncbi.nlm.nih.gov/books/NBK536919
- Marino PL. The ICU Book. 4th ed. Wolters Kluwer Health, 2014.
- Davidson AC, Banham S, Elliott M, et al. BTS/ICS guideline for the ventilatory management of acute hypercapnic respiratory failure in adults. Thorax. 2016;71(Suppl 2):ii1–ii35.
