Severe ARDS with right ventricular dysfunction: positioning, fluids and paralysis
A 58-year-old man is admitted to intensive care with severe community-acquired pneumonia. He has been intubated for 3 days and is sedated with propofol and fentanyl, ventilated with volume control: tidal volume 6 mL/kg predicted body weight, PEEP 14 cmH2O, FiO2 0.8 and respiratory rate 28/min. Arterial blood gas: PaO2 8.0 kPa, PaCO2 6.1 kPa, pH 7.28, lactate 3.6 mmol/L. Plateau pressure is 32 cmH2O and driving pressure 18 cmH2O. He is on noradrenaline 0.35 µg/kg/min with a mean arterial pressure of 68 mmHg, and urine output has been 0.3 mL/kg/h for 4 hours. Chest radiograph shows bilateral infiltrates. Echocardiography shows a moderately dilated right ventricle with reduced fractional area change. He is 178 cm tall and weighs 92 kg. There is no pneumothorax.
1. Is prone positioning indicated, and what practical issues determine whether it will help or harm?
Read the worked answer
Yes. The ratio of PaO2 to FiO2 is about 10 kPa (75 mmHg), which is severe ARDS, and he remains hypoxaemic despite a PEEP of 14 cmH2O and a reasonable tidal volume. Prone positioning for about 16 hours a day improves oxygenation and reduces mortality in severe ARDS when applied early and repeated, largely by recruiting dorsal lung and making ventilation more homogeneous. There is no listed contraindication: no unstable spine, open abdomen, recent sternotomy, raised intracranial pressure or haemodynamic instability that cannot be managed.
Success depends on execution. The team should plan the turn with sufficient staff, secure the tracheal tube and lines, protect eyes and pressure areas, and use a protocol for the face, breasts, knees and genitalia. Enteral feeding can continue with head-up tilt where possible. The tube and lines must be checked after the turn, because disconnection and tube migration are the main avoidable harms. Sedation must be deep enough to prevent dyssynchrony, and neuromuscular blockade may be needed for the turn itself.
Response should be assessed on oxygenation and driving pressure, not on a single blood gas immediately after turning. If there is no benefit after several sessions, or if the right ventricle deteriorates, the strategy should be reconsidered. Prone positioning is not a substitute for treating the pneumonia, avoiding fluid overload and managing the right ventricle.
2. How would you manage fluids and vasoactive drugs given the oliguria and right ventricular dysfunction?
Read the worked answer
The oliguria does not automatically mean he is hypovolaemic; after 3 days of sepsis and ventilation he may be fluid overloaded, and a dilated, poorly contracting right ventricle makes aggressive fluid loading dangerous because it can further distend the right ventricle and reduce left ventricular filling. I would assess fluid responsiveness dynamically: passive leg raise or stroke volume variation if he is in sinus rhythm and fully ventilated, together with echocardiography of ventricular size and inferior vena cava, and lung ultrasound for overload. A small 250 mL balanced crystalloid challenge with reassessment is reasonable if he looks fluid responsive; otherwise fluid should be restricted.
Maintain right ventricular perfusion with noradrenaline, targeting a mean arterial pressure around 65 mmHg and avoiding a fall in systemic pressure during any turn or procedure. If noradrenaline requirements rise above about 0.5 µg/kg/min, add vasopressin to reduce catecholamine load and pulmonary vascular resistance. Consider inhaled nitric oxide or inhaled epoprostenol for right ventricular failure if oxygenation and right ventricular function remain poor, and avoid hypoxia, hypercapnia, acidosis and high airway pressures, all of which raise pulmonary vascular resistance. Hydrocortisone should be considered for vasopressor-dependent septic shock. Diuresis or renal replacement therapy should wait until the shock has resolved.
3. What are the arguments for and against neuromuscular blockade in this patient?
Read the worked answer
The case for paralysis is that he is severely hypoxaemic with a high driving pressure, and dyssynchrony or vigorous respiratory effort can worsen lung injury through pendelluft and high transpulmonary pressures. Neuromuscular blockade for the first 48 hours improves survival in some trials of severe ARDS and makes prone positioning safer and more effective. It also reduces oxygen consumption and may improve right ventricular loading.
The case against is that the evidence is not unanimous, and later trials with lighter sedation did not reproduce the mortality benefit. Prolonged paralysis causes intensive care unit-acquired weakness, delays mobilisation, increases the risk of pressure injury and venous thromboembolism, and can mask inadequate sedation. The right ventricle may become more dependent on whatever respiratory contribution it has.
My approach would be to use a neuromuscular blocker as a time-limited intervention, for example cisatracurium for 24–48 hours, while deep sedation is established, the patient is proned and the ventilator settings are optimised. Use train-of-four monitoring, ensure adequate sedation with clinical assessment or processed electroencephalography, and set a clear daily review to stop the infusion. Review the sedation strategy, ventilator settings and treatable causes of deterioration each day, rather than allowing paralysis to become an open-ended default.