Acute Respiratory Distress Syndrome

The 2024 global definition describes ARDS as an acute, diffuse, inflammatory lung injury precipitated by a risk factor such as pneumonia, in which the injury raises pulmonary vascular and alveolar epithelial permeability and fluid pours into the air sacs. The hallmark lesion is exactly that permeability: the alveolar-capillary membrane leaks fluid, proteins, neutrophils and red blood cells into the alveolar space, filling it with protein-rich inflammatory edema. This is non-cardiogenic pulmonary edema — a leak, not a pressure problem — which is why the same disorder is also called increased-permeability pulmonary edema. Heart failure can look identical on a chest X-ray, so an assessment of left-ventricular function (usually an echocardiogram) is what tells the two apart.

At the cellular level, type I alveolar cells are destroyed early and the basement membrane is stripped bare; the endothelium stays structurally intact but functionally impaired. Neutrophils recruited into the alveoli release proteases, reactive oxygen species and neutrophil extracellular traps, and platelets aggregate with them. Injury to type II cells cuts surfactant production, so alveoli collapse: surfactant dysfunction increases atelectasis, and atelectatic lung is then more vulnerable to mechanical injury. Because the injury is non-uniform and gravity pulls edema downward, the dependent bases collapse more than the apices — a fact that later explains why turning the patient prone works.

The insult can arrive from either side of the membrane. Direct (pulmonary) causes injure the epithelium first: pneumonia (bacterial and viral, including COVID-19; fungal less often), aspiration of gastric or oral contents, and inhalation injury. Indirect (extrapulmonary) causes injure the endothelium through a systemic inflammatory cascade: sepsis, major trauma, massive transfusion, pancreatitis, drowning, drug overdose and fat embolism. Pneumonia and sepsis are the primary causes, and a causative pathogen goes unidentified in 30–60% of cases. Onset is usually progressive — dyspnea and hypoxemia worsening within 6 to 72 hours of the insult, escalating to ventilator-requiring distress within 12 to 24 hours — but transfusion-related acute lung injury (TRALI) can be hyper-acute and fulminant.

Hypoxemia in ARDS comes from ventilation-perfusion mismatch and right-to-left intrapulmonary shunt. Edema-filled and collapsed alveoli are still perfused but no longer ventilated, so blood streams past them and returns to the left heart without ever meeting oxygen. That is the exam logic behind the most important bedside observation in the disorder: despite 100% oxygen, the patient stays hypoxemic. Raising the FiO₂ enriches gas in alveoli that already work; it cannot oxygenate blood that never reaches a working alveolus. The fix is therefore mechanical — reopen alveoli with PEEP and redistribute perfusion with prone positioning — while the FiO₂ is kept at the lowest level that achieves adequate oxygenation, because oxygen toxicity from reactive oxygen species adds injury of its own. Targets once ventilated are modest on purpose: SpO₂ 88–95% or PaO₂ 55–80 mm Hg.

The flooded lung is also a stiff lung. Compliance falls and alveolar dead space rises — and both are independent predictors of mortality. Consolidated regions accept no gas, so the aerated volume shrinks to a "baby lung": a tidal volume sized for a healthy adult now overdistends a much smaller aerated lung. That single idea is the rationale for low tidal volume scaled to predicted body weight, which tracks height rather than actual weight because lung capacity relates to height. For the patient, the stiff lung means tachypnea and visibly increased breathing effort — the earliest cues — with tachycardia from hypoxemia and sympathetic drive.

ARDS moves through phases. The first is alveolar-capillary damage — the exudative phase, about the first week: diffuse alveolar damage with interstitial and alveolar edema, hyaline membranes, alveolar hemorrhage, neutrophils, fibrin, atelectasis and epithelial necrosis. A proliferative phase follows, marked by type II cell hyperplasia and healing, lasting more than three weeks in survivors; interstitial fibrosis can begin here. A fibrotic phase signals the end of the acute process, but fibrosis develops in only a subset of patients, and even classic diffuse alveolar damage is found in only about 45% of clinical ARDS at autopsy. Most survivors regain most of their lung function, yet many report dyspnea on exertion long after the numbers normalize.

The ventilator that keeps the patient alive can wound the baby lung four ways. Volutrauma is alveolar overdistension from large tidal volumes; barotrauma is rupture from high inflation pressure, producing pneumothorax, pneumomediastinum and subcutaneous emphysema; atelectrauma is the shear of collapsed units snapping open and shut with every breath; and biotrauma is the inflammatory response that mechanical injury provokes — which feeds back into the leak. The lung-protective bundle exists to break that cycle. ATS (2017, reaffirmed 2024) strongly recommends tidal volumes of 4–8 mL/kg predicted body weight with plateau pressure below 30 cm H₂O; ESICM 2023 makes the same strong low-tidal-volume recommendation. PEEP keeps alveoli open at end-expiration, reducing atelectasis and atelectrauma, improving oxygenation, and lowering the work of breathing in a low-compliance lung. Small volumes let CO₂ climb — permissive hypercapnia tolerates a raised PaCO₂, with pH allowed down to about 7.20, as the price of protecting the lung.

The nurse reads the pressures. Plateau pressure, measured with an inspiratory pause, reflects alveolar pressure and is kept below 30. Driving pressure is plateau minus PEEP; when it climbs past 14 cm H₂O the tidal volume is trimmed toward 4 mL/kg. A high peak pressure with a normal plateau points to airway resistance — secretions, bronchospasm, a kinked tube, biting — while high peak and high plateau together mean a compliance problem: worsening ARDS, pulmonary edema, a mainstem intubation, or a pneumothorax. Barotrauma announces itself as a sudden rise in peak or plateau pressure, a sudden fall in delivered tidal volume and high-pressure alarms, then acute desaturation, hypotension, tachycardia, unilateral absent breath sounds, tracheal deviation or crepitus. Tension pneumothorax can kill within minutes; it is a clinical diagnosis treated with urgent needle decompression and then a chest tube — do not wait for the chest radiograph. Barotrauma complicated 6.5% of ARDS in one large series and raised mortality from 39% to 51%.

Prone positioning attacks the same physics from the outside. Lying prone improves ventilation-perfusion matching in the dorsal lung — the regions that were gravity-dependent and collapsed while supine — relieves the weight of the heart and mediastinum on that dorsal lung, reduces ventral overdistension and regional stress and strain, and unloads the right ventricle. Oxygenation improves in about 50–70% of patients, often quickly enough to reduce FiO₂ and PEEP. In PROSEVA, sessions of 16 hours or more started early in moderate-to-severe ARDS cut 28-day mortality from 32.8% to 16.0%. ATS strongly recommends proning for more than 12 hours per day in severe ARDS; ESICM 2023, as reported in the 2026 Intensive Care Medicine prone-positioning review, applies it to intubated patients with PaO₂/FiO₂ below 150 mm Hg on PEEP ≥ 5 despite optimized settings, using prolonged sessions of 16 consecutive hours or more.

Positive pressure does not stay in the chest. PEEP raises intrathoracic pressure, which narrows the gradient between the veins and the right atrium; venous return and preload fall, right-atrial filling falls, and cardiac output falls — and cardiac output falling lowers mean arterial pressure unless systemic vascular resistance rises to compensate. The effect is worst in distributive shock or hypovolemia, which describes most septic ARDS patients. So after every PEEP increase and every sedation bolus the nurse expects and watches for hypotension, and hypotension that appears on positive-pressure ventilation is investigated for pneumothorax before it is attributed to anything else.

The right ventricle bears the second load. Hypoxic vasoconstriction, edema compressing vessels, inflammatory remodeling, endothelial dysfunction with less nitric oxide and prostacyclin, microthrombi, and overdistension raising transpulmonary pressure all raise pulmonary vascular resistance; hypercapnia worsens the vasoconstriction, so permissive hypercapnia has a cost. Acute cor pulmonale is present in 21–50% of mechanically ventilated ARDS patients and carries markedly higher mortality. That is why vasopressors — norepinephrine and epinephrine raise right-ventricular contractility — support cardiac output when positive pressure, deep sedation and sepsis pull it down, and why the conservative fluid strategy is deliberately sequenced after shock resolves: the FACTT trial withheld diuresis until the patient had held a mean arterial pressure above 60 mm Hg off vasopressors for 12 hours. Fluid overload after resuscitation worsens the edema; high PEEP in a hypovolemic patient collapses the blood pressure. The nurse lives between those two errors.