Acute respiratory distress syndrome (ARDS) is best described as ‘leaky lung syndrome’ or ‘low pressure, non cardiogenic pulmonary oedema‘. It is an acute inflammatory lung injury, often in previously healthy lungs, mediated by a uniform pulmonary pathological process in response to a variety of direct (i.e. inhaled) or indirect (i.e. blood-borne) insults.
The pathophysiological response occurs in two stages:
• An initial acute inflammatory phase (days 3–10) in which cytokine-activated neutrophils and monocytes adhere to the alveolar epithelium and capillary endothelium and release pro-inflammatory mediators. These damage the integrity of the alveolar capillary membrane, increase permeability, and cause alveolar oedema. A reduction in surfactant causes alveolar collapse, hyaline membrane formation, and reduced lung compliance. Subsequent severe V/Q mismatch and shunting cause progressive hypoxaemia, respiratory failure, and pulmonary hypertension. Hyperventilation of remaining functional alveoli is usually sufficient to prevent significant hypercapnia.
• A later healing, fibroproliferative phase (>10 days) is initially associated with resolution of pulmonary infiltrates, increased type 2 pneumocytes (i.e. surfactant production), myofibroblasts, and early collagen formation. Ongoing pulmonary fibrosis distorts lung architecture, reduces compliance (stiff lungs), and is associated with pulmonary hypertension.
Definition and diagnosis of ARDS
The recent internationally agreed criteria for the diagnosis of ARDS (Berlin Definition 2012) include:
• Acute onset within 1 week of the causative insult.
• Bilateral diffuse opacities on CXR.
• Non-cardiac origin for pulmonary oedema (nor fluid overload).
Oxygenation (degree of hypoxaemia) defines the severity of the ARDS; mild ARDS equates to a PaO2/FiO2 (P/F) of 200–300 mmHg; moderate 100–200 mmHg; severe <100 mmHg (all with PEEP/CPAP ≤5cm H2O). P/F is calculated as follows; if PaO2 is 80 mmHg on 80% inspired oxygen; PaO2/FiO2 = 80/0.8 = 100 mmHg.
Mild ARDS equates to the previous definition of acute lung injury.
Common causes of ARDS
Direct pulmonary
Infective (e.g. pneumonia, TB)
Pulmonary trauma
Near drowning
Toxic gas inhalation
• Smoke
• NO2, NH3, Cl2
• Phosgene
Oxygen toxicity (FiO2 0.8)
Inhalation of gastric contents
Indirect pulmonary injury
Sepsis
Non-thoracic trauma
Haemorrhage/large
transfusions
Post-arrest
Burns
Bowel infarction, pancreatitis
Drugs (e.g. salicylates)
Uraemia, eclampsia, toxins
Anaphylaxis
Epidemiology and prognosis of ARDS
The incidence of moderate and severe ARDS is ~2–8 cases/105/year, but its precursor mild ARDS is more common and often unrecognized. Improvements in survival over the last 20 years are due to better supportive care, rather than modification of the inflammatory process. Mortality is ~27% in mild, ~32% in moderate, and ~45% in severe ARDS. It is worse with pulmonary causes (trauma <35%, sepsis ~50%, aspiration pneumonia ~80% mortality) and increases with age (>60 years) and sepsis.
Early deaths are usually due to the precipitating condition and later deaths to multiorgan failure (MOF) and complications, with <20% dying from hypoxaemia alone. Residual lung damage, including restrictive defects and reduced gas transfer, occurs in 50% of survivors but is often mild, reflecting the lungs’ regenerative capacity.
Clinical features of ARDS
ARDS usually presents 1–2 days after the onset of the precipitating cause (e.g. pancreatitis, trauma). Typical features include increasing dyspnoea, dry cough, tachypnoea, cyanosis, increased work of breathing due to reduced compliance, confusion, and rapidly progressive hypoxaemia. Rapidly increasing concentrations of supplemental oxygen are required. Chest examination reveals coarse crepitations on auscultation, and the CXR shows diffuse alveolar infiltration.
Early CT scans often demonstrate dependent consolidation and later scans pneumothoraces, pneumatoceles, and fibrosis. These features are not diagnostic and must be differentiated from heart failure, diffuse alveolar haemorrhage, vasculitis, infection, and infiltrative disorders like lymphangitis carcinomatosis.
During the healing, fibroproliferative phase pneumothoraces are common. Secondary nosocomial infections affect 50% of cases and require constant surveillance (e.g. bronchoalveolar lavage).
Other complications include venous thromboembolism, gastrointestinal haemorrhage, and myopathy associated with steroids, paralysing agents, and poor glycaemic control.
Management of ARDS
Initially, the precipitating cause must be identified and treated. In mild ARDS, supportive therapy with supplemental oxygen, diuretics to minimize alveolar oedema, and physiotherapy aimed at optimizing gas exchange and V/Q matching will maintain gas exchange whilst awaiting recovery.
In some cases, non-invasive, mask-delivered respiratory support with continuous positive airways pressure (CPAP) to correct hypoxaemia or bilevel positive airways pressure (BIPAP), which also aids ventilation by reducing work of breathing, may avoid the need for mechanical ventilation (MV).
In severe ARDS, intubation and mechanical ventilation are usually indicated. High peak inspiratory pressures (PIP) are required to achieve normal tidal volumes (Tv) due to reduced lung compliance. These high pressures cause alveolar damage termed ‘barotrauma’ (e.g. pneumothorax). The reduced compliance of diseased or consolidated alveoli also results in the available Tv causing overdistension and damage to otherwise healthy alveoli, an effect termed ‘volutrauma’.
• Mechanical ventilation aims to limit pressure-induced lung damage and volutrauma, prevent oxygen toxicity (i.e. FiO2 <80%), optimize alveolar recruitment and oxygenation, and avoid circulatory compromise due to high intrathoracic pressures. A ‘protective’ lung ventilation strategy of low Tv (6 mL/kg) and low PIP (<30 cmH2O) prevent lung damage, whilst high positive end-expiratory pressures (PEEP; >10 cmH2O) and long inspiratory to expiratory (I:E) times (i.e. 2:1 instead of the normal 1:2) recruit collapsed alveoli. No ventilatory mode is proven to be superior, although pressure-controlled modes which tend to improve alveolar recruitment are generally favoured. The resulting CO2 retention, termed ‘permissive hypercapnia‘, resulting from this low Tv strategy, is usually tolerated with adequate sedation.
• Conservative fluid management reduces alveolar oedema caused by increased alveolar permeability. The aim is to maintain organ perfusion whilst reducing the pulmonary capillary hydrostatic pressure which generates alveolar oedema, often by using inotropes or vasoactive drugs, rather than with excessive fluid administration. In the acute phase, diuretics reduce pulmonary oedema and may improve oxygenation.
• General measures include good nursing care, physiotherapy, nutrition, sedation, and infection control. Bronchoscopy improves ventilation and V/Q matching by removing sputum plugs and secretions. Minimize metabolic demand and associated oxygen consumption (VO2) by preventing fever (VO2 increases by >10% for every degree rise above 37ºC) and shivering (can double or triple VO2) and controlling agitation with sedatives.
• Pharmacotherapy. No drug therapy, including steroids, anti-inflammatory agents, or surfactant, has been consistently effective in clinical ARDS trials. The role of corticosteroids in late-stage ARDS remains controversial and generally cautious. One consensus group has supported a ‘weak’ recommendation for use of low-to-moderate doses in ARDS of <14 days duration. If corticosteroids are used, strict infection surveillance, avoidance of neuromuscular blockers, and gradual weaning are recommended.
• Future developments include inhaled nitric oxide to increase perfusion of ventilated alveoli by vasodilating surrounding vessels, improving V/Q matching, and reducing overall shunt fraction. Unfortunately, the initial PaO2 improvement is not sustained, and there is no survival benefit. As consolidation is usually dependent and blood flow is greatest in the dependent areas, prone positioning aims to improve V/Q match by turning the patient so that non-consolidated ventilated lung is dependent. Although oxygenation is improved, there is no survival benefit. Extracorporeal membrane oxygenation (ECMO), techniques to oxygenate blood or remove CO2 are effective in children and increasing recent evidence suggests significant benefit in adults.