On this page
Study tools
Write your answer
Saved on this device. No marking — you are the marker.
Q1: Definition and classification (2 min)
Examiner: Define respiratory failure. And how do you classify it at the bedside — what is the single most useful distinction, and why?
Expected answer:
- Respiratory failure is the failure of the respiratory system to maintain adequate gas exchange, defined by the arterial blood gas: a PaO2 below 60 mmHg (8 kPa) while breathing room air, with or without a raised PaCO2.
- The single most useful distinction is Type 1 versus Type 2:
- Type 1 (hypoxaemic) — PaO2 below 60 mmHg with normal or low PaCO2. Mechanism V/Q mismatch or shunt. Causes: pneumonia, ARDS, pulmonary oedema, pulmonary embolism, asthma, pneumothorax, ILD.
- Type 2 (hypercapnic / ventilatory) — PaO2 below 60 mmHg with PaCO2 above 45 mmHg (6 kPa). Mechanism alveolar hypoventilation. Causes: COPD exacerbation, neuromuscular weakness, obesity hypoventilation, opiate/sedative overdose, brainstem stroke.
- The distinction matters because it determines the oxygen target (94 to 98 percent for Type 1; 88 to 92 percent for COPD/Type 2 to avoid oxygen-induced hypercapnia), the need for ventilatory support (BiPAP for acidotic Type 2), and the differential of causes (lung versus pump).
Follow-up: What are the extended Types 3 and 4? Type 3 — perioperative / atelectasis-related hypoxaemia; Type 4 — shock / low perfusion with mixed-venous desaturation. Both are managed within the Type 1/2 framework.
Q2: Pathophysiology — the five mechanisms and the A-a gradient (3 min)
Examiner: Take me through the five mechanisms of hypoxaemia. Which one does NOT correct with 100 percent oxygen, and why? And how does the A-a gradient help you?
Expected answer — the five mechanisms:
- Hypoventilation — PaCO2 rises; the alveolar gas equation lowers PAO2. A-a gradient is NORMAL — the lung is fine, the pump has failed. Corrects with oxygen, but the cause (opiates, neuromuscular, brainstem) must be fixed.
- V/Q mismatch — the commonest mechanism in disease. Blood perfuses under-ventilated alveoli. Largely corrects with supplemental oxygen. Underlies most Type 1 failure and most COPD exacerbations.
- Shunt — blood bypasses ventilated alveoli entirely (ARDS, lobar pneumonia, pulmonary AV malformation, right-to-left cardiac shunt). Does NOT correct with 100 percent oxygen — the blood never sees the alveolus. This is the bedside definition of shunt: refractory hypoxaemia on high FiO2.
- Diffusion impairment — thickened alveolar-capillary membrane (ILD, fibrosis, oedema). Corrects with oxygen (raises the gradient). Worse on exercise.
- Low inspired PaO2 — altitude, hypoxic gas mixture, rebreathing. A-a gradient normal. Corrects by restoring FiO2.
Why only shunt resists oxygen: because in shunt the blood does not contact ventilated gas at all, so raising FiO2 has no effect on the shunted fraction. V/Q mismatch and diffusion both respond because raising FiO2 raises the gradient across the (still-ventilated) alveolus.
The A-a gradient: PAO2 (calculated from the alveolar gas equation: PAO2 equals 150 minus PaCO2/0.8 on room air) minus the measured PaO2. Normal is below 15 mmHg in a young adult (rises about 3 mmHg per decade). A normal A-a gradient with hypoxaemia means pure hypoventilation — the lung is fine, look at the pump (opiates, neuromuscular, brainstem). A raised A-a gradient means lung disease or right-to-left shunt. This single calculation redirects the entire work-up.
Follow-up: Why is pulse oximetry a late detector of hypoventilation? Because the oxyhaemoglobin dissociation curve is flat above an SpO2 of about 90 percent, PaO2 can fall from 100 to 60 mmHg with barely a change in SpO2; the oximeter only starts to fall once PaO2 drops onto the steep shoulder of the curve. And in carbon monoxide poisoning and methaemoglobinaemia the oximeter reads falsely normal — confirm with an ABG and co-oximetry.
Q3: Oxygen-induced hypercapnia in COPD — the Haldane effect and loss of HPV (2 min)
Examiner: You give high-flow oxygen to a CO2-retaining COPD patient and she becomes more hypercapnic and drowsy. Walk me through the mechanism. And what is the oxygen rule you should have used?
Expected answer — two mechanisms:
- The Haldane effect. Deoxyhaemoglobin carries CO2 (as carbamino compounds) and buffers H plus far better than oxyhaemoglobin. Oxygenating venous blood in the lung releases CO2 from haemoglobin into plasma, raising mixed venous and arterial CO2. In a patient with marginal ventilation, this extra CO2 load cannot be cleared and PaCO2 rises.
- Loss of hypoxic pulmonary vasoconstriction (HPV). In COPD, poorly ventilated lung units are kept locally vasoconstricted by hypoxia, diverting blood to better-ventilated units. Supplemental oxygen relieves the hypoxia, relaxes the vasoconstriction, and re-perfuses poorly ventilated units, worsening V/Q mismatch and increasing dead-space-equivalent CO2 retention.
The clinical consequence is CO2 narcosis: rising PaCO2 produces CSF acidosis, cerebral vasodilatation (headache), warm peripheries and a bounding pulse, a flapping tremor (asterixis), then depression of the central respiratory drive — confusion, drowsiness, coma, and respiratory arrest.
The oxygen rule: controlled oxygen via a fixed-performance Venturi mask (24 or 28 percent), target SpO2 88 to 92 percent, and repeat the arterial blood gas at 30 to 60 minutes (BTS oxygen guideline). Hypoxia kills faster than hypercapnia, so oxygen is never withheld — but it must be controlled in a known retainer.
Follow-up: What is the evidence for controlled oxygen in COPD? The Austin 2010 randomised trial of titrated versus high-flow oxygen in COPD exacerbations showed titrated (controlled) oxygen was safer — fewer deaths, less acidosis.
Q4: NIV in hypercapnic respiratory failure — indications, settings, evidence (2 min)
Examiner: When do you start non-invasive ventilation, and how do you set it up? Cite the evidence.
Expected answer — indications (BTS/ICS 2016):
- COPD exacerbation with pH 7.25 to 7.35 despite standard medical therapy — first-line, level-1 evidence. pH below 7.25 — NIV on the ICU, with a low threshold for intubation.
- Decompensated obesity hypoventilation syndrome with acidotic hypercapnia.
- Neuromuscular weakness / chest-wall disease with hypercapnia.
- Cardiogenic pulmonary oedema — CPAP (and BiPAP if hypercapnic); faster physiological improvement than standard oxygen.
- Weaning and peri-extubation in high-risk hypercapnic patients.
Settings (BiPAP for COPD): IPAP 10 to 15 cmH2O, EPAP 4 to 5 cmH2O, oxygen entrained to target SpO2 88 to 92 percent, full-face mask first. Titrate IPAP upward by 2 to 5 cmH2O every 10 minutes until the pH rises above 7.35 or PaCO2 falls; reassess at 1, 4 and 12 hours. Switch to a nasal interface once stable. Escalate to intubation if NIV fails (rising CO2, falling pH, exhaustion, intolerance) or the airway is threatened.
Evidence: the Plant 2000 multicentre randomised trial (Lancet) — early NIV on general respiratory wards for COPD exacerbations with pH 7.25 to 7.35 reduced in-hospital mortality, the need for intubation, and length of stay.
Follow-up: When do you go straight to intubation rather than NIV? Immediate intubation for: respiratory arrest or peri-arrest; failure of or contraindication to NIV; an unprotected airway (GCS below 8, copious secretions, vomiting risk); severe refractory hypoxia despite NIV/CPAP/HFNC; and haemodynamic instability requiring vasopressors alongside respiratory failure. The fatiguing patient must be intubated before the arrest.
Q5: ARDS — lung-protective ventilation and escalation (3 min)
Examiner: Define ARDS. How do you ventilate a patient with it, and when do you escalate to prone positioning or ECMO?
Expected answer — definition (Berlin 2012): onset within ONE WEEK of a known insult; BILATERAL opacities not fully explained by effusion, atelectasis or nodules; respiratory failure NOT FULLY EXPLAINED BY CARDIAC FAILURE or fluid overload; oxygenation by PaO2/FiO2 on PEEP at least 5 cmH2O — mild 200 to 300, moderate 100 to 200, severe below 100.
Lung-protective ventilation (ARDS Network 2000): the trial that changed practice — tidal volume 6 mL/kg PREDICTED (ideal) body weight versus traditional 12 mL/kg reduced mortality from 40 to 31 percent. The bundle: tidal volume 6 mL/kg ideal body weight (reduce to as low as 4 mL/kg if plateau pressure exceeds 30 cmH2O); plateau pressure below 30 cmH2O; PEEP titrated to FiO2 (higher-PEEP strategy in moderate to severe, ALVEOLI/Brower 2004); permissive hypercapnia (allow pH at least 7.20); oxygen target SpO2 88 to 95 percent (PaO2 55 to 80 mmHg); conservative fluid strategy (keep the lung dry). The rationale is to minimise ventilator-induced lung injury (volutrauma, barotrauma, atelectrauma, biotrauma).
Escalation:
- Prone positioning at least 16 hours per day in severe ARDS (P/F below 150) — PROSEVA 2013 reduced mortality from 33 to 16 percent.
- ECMO for refractory hypoxaemia (P/F below 80 despite optimised ventilation) or uncontrollable respiratory acidosis (pH below 7.20 with PaCO2 above 80) — centre-of-excellence referral (CESAR/EOLIA framework).
- Neuromuscular blockade (cisatracurium) for 48 hours in severe ARDS (ACURASYS) is used selectively.
Follow-up: Why ideal and not actual body weight? Because ARDS preferentially injures a small 'baby lung' of normally aerated tissue; ventilation must be sized to the lung available, not to the patient's body habitus. Using actual body weight in an obese patient over-distends the small aerated lung and causes volutrauma.
Q6: The fatiguing patient, neuromuscular failure, and the pearls (2 min)
Examiner: You are assessing a breathless patient at the bedside. How do you recognise the fatiguing patient who must be intubated NOW? And how does that differ in neuromuscular failure?
Expected answer — the fatiguing patient: first tachypnoeic and agitated, with accessory-muscle use and inability to speak in full sentences; then the work of breathing cannot be sustained — the respiratory rate falls, breaths become shallow, accessory-muscle use wanes, and paradoxical (see-saw / abdominal) breathing appears. The patient becomes drowsy. A falling respiratory rate in a tachypnoeic patient is fatigue, not improvement. This is the pre-arrest state — intubate before the arrest, not after.
Neuromuscular respiratory failure is different: the patient is often calm, not tachypnoeic early, because the problem is a slowly failing pump (Guillain-Barre, myasthenia, motor neuron disease). SpO2 is a late and dangerous marker — by the time it falls, the patient is in trouble. Monitor the FVC, the maximum inspiratory pressure (MIP), and the maximum expiratory pressure (MEP). Thresholds to support: FVC below 20 mL/kg (below 1 L), MIP below minus 30 cmH2O, MEP below 40 cmH2O, or a fall in FVC above 30 percent. Start NIV (BiPAP) and arrange cough augmentation (mechanical insufflation-exsufflation); intubate electively when bulbar weakness or fatigue supervenes. Treat the underlying disease (IVIG or plasma exchange for GBS; cholinesterase inhibitors and immunosuppression for MG).
The pearls that decide a respiratory-failure answer: (1) respiratory failure equals PaO2 below 60 mmHg on room air; type 1 versus type 2 by PaCO2; (2) the five mechanisms — only shunt fails to correct with 100 percent oxygen; (3) a normal A-a gradient with hypoxaemia equals pure hypoventilation; (4) oxygen target 94 to 98 percent type 1, 88 to 92 percent COPD/type 2; (5) type 2 plus acidosis means BiPAP (Plant 2000, BTS/ICS 2016); (6) ARDS means lung-protective 6 mL/kg ideal body weight, plateau below 30 cmH2O, prone in severe (PROSEVA), ECMO if refractory; (7) HFNC reduces intubation and mortality in P/F at or below 300 (FLORALI 2015); (8) pulse oximetry is falsely reassuring in carbon monoxide poisoning and methaemoglobinaemia.