Acute Respiratory Distress Syndrome: Mechanisms, Diagnosis, Management, and Outcomes

Bottom Line

Acute respiratory distress syndrome (ARDS) is a life-threatening form of acute inflammatory lung injury and hypoxemic respiratory failure that affects approximately 10% of intensive care unit admissions globally, with hospital mortality rates of 35–46% depending on severity. Over the past three decades, the diagnostic framework has evolved from the 1994 American-European Consensus Conference criteria through the 2012 Berlin Definition to the 2023 Global Definition, each iteration improving reliability and inclusivity. The cornerstone of evidence-based management remains lung-protective mechanical ventilation (6 mL/kg tidal volume, plateau pressure ≤30 cm H₂O), with adjunctive therapies including prone positioning for severe cases, conservative fluid strategies, and cautious use of neuromuscular blockade and extracorporeal membrane oxygenation in selected patients. Major randomized trials have reshaped practice, but the evidence base remains marked by important controversies — particularly around the routine use of neuromuscular blockade following the ROSE trial, the role of recruitment maneuvers after the ART trial, and the optimal selection of patients for ECMO after the EOLIA trial. Long-term outcomes among survivors are dominated by persistent physical, cognitive, and neuropsychological impairments, and the COVID-19 pandemic both revealed distinctive vascular pathobiology and exposed deep health inequities in ARDS outcomes. This report provides a rigorous, evidence-grounded analysis of ARDS from definition through management and outcomes, distinguishing established recommendations from areas of active investigation.

---

1. Definition and Clinical Significance

ARDS is an acute, diffuse, inflammatory lung injury characterized by increased permeability of the alveolar–capillary barrier, resulting in non-cardiogenic pulmonary edema, reduced lung compliance, and severe hypoxemia. Clinically, it presents as acute-onset respiratory failure with bilateral radiographic opacities and a ratio of arterial oxygen partial pressure to fraction of inspired oxygen (PaO₂/FiO₂) ≤300 mm Hg, occurring within one week of a known clinical insult and not fully explained by cardiac failure or fluid overload.

The clinical significance of ARDS is profound: it is a common, frequently underrecognized, and highly morbid syndrome in critically ill patients. The LUNG-SAFE study, a prospective cohort involving 459 ICUs across 50 countries, found that ARDS accounted for 10.4% of ICU admissions and 23.4% of mechanically ventilated patients. Critically, approximately 40% of cases were not recognized by clinicians at the time of meeting diagnostic criteria, and underrecognition was associated with less aggressive use of evidence-supported therapies such as higher PEEP, neuromuscular blockade, and prone positioning.

---

2. Evolution of ARDS Definitions

2.1 The AECC Definition (1994)

The first consensus definition, developed by the American-European Consensus Conference (AECC), established four criteria: acute onset; PaO₂/FiO₂ ≤300 mm Hg for acute lung injury (ALI) and ≤200 mm Hg for ARDS; bilateral infiltrates on frontal chest radiograph; and pulmonary artery wedge pressure ≤18 mm Hg or no clinical evidence of left atrial hypertension.

While foundational, the AECC definition suffered from several limitations: the term "acute" was vaguely defined; inter-observer reliability in interpreting chest radiographs was poor; the PaO₂/FiO₂ criterion lacked standardization for positive end-expiratory pressure (PEEP); and the requirement for pulmonary artery catheterization was increasingly impractical as the use of pulmonary artery catheters declined.

2.2 The Berlin Definition (2012)

The Berlin Definition, published by an expert task force in JAMA in 2012, addressed these limitations through evidence-based refinement. Key changes included:

• Elimination of the term "ALI": ARDS was reclassified into three mutually exclusive severity categories rather than treated as the severe end of a spectrum.
• Explicit timing: "Acute" was defined as onset within one week of a known clinical insult or new/worsening respiratory symptoms.
• Oxygenation criteria with minimum PEEP: A minimum PEEP or CPAP of 5 cm H₂O is required to standardize the PaO₂/FiO₂ ratio measurement.
• Severity categories:
  • Mild: 200 mm Hg < PaO₂/FiO₂ ≤ 300 mm Hg
  • Moderate: 100 mm Hg < PaO₂/FiO₂ ≤ 200 mm Hg
  • Severe: PaO₂/FiO₂ ≤ 100 mm Hg
• Radiographic clarification: Bilateral opacities not fully explained by effusions, lobar/lung collapse, or nodules.
• Cardiac exclusion: Eliminated the specific pulmonary artery wedge pressure cutoff; requires that respiratory failure not be fully explained by cardiac failure or fluid overload.

The task force empirically validated the definition using patient-level meta-analysis of 4,188 patients, demonstrating superior predictive validity for mortality compared to the AECC definition (area under the receiver operating curve 0.577 vs. 0.536).

Limitations of the Berlin Definition include its reliance on arterial blood gas sampling (limiting applicability where ABG is unavailable), the requirement for invasive or noninvasive mechanical ventilation (excluding patients managed with high-flow nasal oxygen), and its inapplicability in resource-limited settings that lack PEEP-capable ventilators. Additionally, the ancillary variables proposed for severe ARDS (respiratory system compliance ≤40 mL/cm H₂O, PEEP ≥10 cm H₂O, corrected expired volume per minute ≥10 L/min) were removed from the final definition because they did not improve predictive validity for mortality.

2.3 The Global Definition of ARDS (2023)

The 2023 Global Definition, endorsed by major critical-care societies, updates the Berlin framework to reflect modern clinical practice and global health equity:

• High-flow nasal oxygen (HFNO): Patients treated with HFNO at ≥30 L/min are now included, recognizing its widespread use since the FLORALI trial and the COVID-19 pandemic.
• SpO₂/FiO₂ criteria: Pulse oximetry-derived SpO₂/FiO₂ ≤315 (when SpO₂ ≤97%) is accepted as an alternative to PaO₂/FiO₂, broadening diagnostic access where ABG is unavailable.
• Lung ultrasound: Ultrasound identifying bilateral B-lines or consolidations by trained operators is now accepted as an imaging modality.
• Resource-limited settings: The requirement for PEEP, specific flow rates, or mechanical ventilation devices is removed.

The conceptual framework remains consistent: ARDS is an acute, diffuse, inflammatory lung injury precipitated by a known risk factor, resulting in pulmonary edema and impaired gas exchange not primarily explained by cardiogenic causes.

---

3. Pathophysiology

3.1 Alveolar–Capillary Barrier Injury

The alveolar–capillary membrane is the primary structural target in ARDS. Direct pulmonary insults (e.g., pneumonia, aspiration) primarily damage the alveolar epithelium, while indirect systemic insults (e.g., sepsis, pancreatitis) predominantly target the pulmonary vascular endothelium. Injury disrupts intercellular junctional complexes, leading to paracellular hyperpermeability that allows protein-rich fluid, neutrophils, and red blood cells to leak into the alveolar space. Impaired epithelial sodium and water transport mechanisms further compromise alveolar fluid clearance, perpetuating alveolar flooding.

3.2 The Inflammatory Cascade

The inflammatory response in ARDS involves a coordinated but dysregulated interplay of cellular and molecular mediators:

• Alveolar macrophages are activated by the initial stimulus (pathogen-associated molecular patterns or damage-associated molecular patterns) and release pro-inflammatory cytokines including IL-1β, IL-6, IL-8, and TNF-α.
• Neutrophils are recruited via chemotaxis and undergo respiratory burst, releasing reactive oxygen species, proteases (including matrix metalloproteinases and neutrophil elastase), and pro-inflammatory mediators that propagate tissue injury.
• Neutrophil extracellular traps (NETs) — web-like chromatin structures released by dying neutrophils — promote platelet aggregation, microvascular thrombosis, and further alveolar–capillary barrier damage.
• Cytokine amplification drives a self-sustaining cascade that recruits additional inflammatory cells and activates the coagulation system.

3.3 Pulmonary Edema and Surfactant Dysfunction

The hallmark of ARDS is protein-rich, non-cardiogenic pulmonary edema. Unlike hydrostatic edema, ARDS edema results from increased vascular permeability rather than elevated capillary pressure. The accumulation of proteinaceous fluid in the interstitium and alveoli exceeds lymphatic clearance capacity.

Concurrently, epithelial injury depletes surfactant-producing type II alveolar epithelial cells, and the protein-rich edema fluid directly inactivates existing surfactant. The resulting increase in alveolar surface tension promotes end-expiratory alveolar collapse (atelectasis), worsening lung compliance and gas exchange.

3.4 Impaired Gas Exchange

Gas exchange impairment in ARDS is multifactorial. Ventilation–perfusion (V/Q) mismatch — blood flow to non-aerated or poorly aerated lung regions — is the dominant mechanism of hypoxemia. Increased dead-space ventilation (ventilation to non-perfused regions) contributes to impaired CO₂ elimination. Shunt physiology (perfusion of completely non-ventilated alveoli) produces hypoxemia refractory to supplemental oxygen.

3.5 Heterogeneous Lung Mechanics

A pathognomonic mechanical feature of ARDS is the marked heterogeneity of lung involvement. The lung is not uniformly diseased but rather comprises a patchwork of atelectatic/consolidated regions (dependent zones), edematous tissue, and relatively normal or overdistended regions (non-dependent zones). This creates the "baby lung" concept — the functional lung size may be dramatically reduced, often equivalent to the lung volume of a small child.

This mechanical heterogeneity predisposes to ventilator-induced lung injury (VILI) through:

• Volutrauma: Regional overdistension of aerated lung units.
• Atelectrauma: Cyclic recruitment and derecruitment of unstable alveoli.
• Biotrauma: Release of inflammatory mediators from mechanotransduction.

3.6 Temporal Phases

ARDS pathophysiology progresses through overlapping temporal phases:

• Exudative phase (days 1–7): Characterized by alveolar–capillary barrier disruption, protein-rich edema, neutrophil infiltration, and hyaline membrane formation.
• Fibroproliferative phase (days 7–21): Tissue remodeling with fibroblast proliferation, collagen deposition, and potential progression to fibrosis.

---

4. Direct Pulmonary versus Indirect Extrapulmonary Causes

ARDS is classically dichotomized by the pathway of injury:

Direct (Pulmonary) Causes

Direct injury results from an insult that reaches the lung epithelium via the airways:

Cause	Frequency	Mechanism
Pneumonia (bacterial, viral, fungal)	Most common pulmonary cause	Direct epithelial infection and damage
Aspiration of gastric contents	Common	Chemical pneumonitis + bacterial superinfection
Inhalation injury	Less common	Toxic gas or smoke的直接 epithelial injury
Pulmonary contusion	Traumatic	Direct mechanical disruption of alveolar architecture
Near-drowning	Rare	Osmotic and chemical injury to epithelium

Indirect (Extrapulmonary) Causes

Indirect injury results from a systemic inflammatory insult that reaches the lung via the bloodstream, primarily damaging the pulmonary endothelium:

Cause	Frequency	Mechanism
Sepsis (non-pulmonary)	Most common overall cause	Systemic inflammation; endothelial activation and injury
Severe trauma with shock/multiple transfusions	Common	Ischemia–reperfusion; transfusion-related acute lung injury (TRALI)
Acute pancreatitis	Common	Systemic release of pancreatic enzymes and inflammatory mediators
Burns	Less common	Systemic inflammatory response syndrome
Drug overdose/toxicity	Less common	Direct or immune-mediated injury
Cardiopulmonary bypass	Rare	Systemic inflammation from bypass circuit

The distinction has pathophysiological relevance: direct injury produces more prominent alveolar epithelial damage, alveolar collapse, and thicker hyaline membranes, while indirect injury is associated with more diffuse endothelial injury and systemic inflammation. In practice, many patients have overlapping direct and indirect causes.

---

5. Epidemiology and Risk Factors

5.1 Incidence

The LUNG-SAFE study remains the largest epidemiological investigation, reporting that ARDS constitutes 10.4% of ICU admissions and 23.4% of mechanically ventilated patients. Population-based incidence estimates range from 3.65 to 78.9 per 100,000 person-years, varying substantially by region and diagnostic criteria. The COVID-19 pandemic caused a dramatic surge, with some cohorts reporting ARDS incidences of 50–80% among critically ill patients.

5.2 Risk Factors

Established risk factors include:

• Demographics: Male sex; age >65 years (most patients are in their sixth or seventh decade)
• Comorbidities: Diabetes mellitus, obesity, chronic obstructive pulmonary disease, chronic heart failure, chronic liver disease, alcohol use disorder
• Acute risk factors: Sepsis (pulmonary or non-pulmonary), severe trauma, aspiration, pancreatitis, multiple transfusions, burn injury, drug overdose
• Genetic factors: Polymorphisms in genes encoding angiotensin-converting enzyme, surfactant proteins, and inflammatory cytokines have been associated with differential susceptibility, though none are clinically actionable

In low- and middle-income countries, infectious pathogens such as Plasmodium species, dengue virus, and Leptospira contribute additional risk, and limited diagnostic resources likely lead to substantial underdetection.

5.3 Mortality

Pre-pandemic ARDS mortality from the LUNG-SAFE study was 34.9% (mild), 40.3% (moderate), and 46.1% (severe), with an overall mortality of approximately 40%. Some studies suggest a gradual mortality decline over recent decades attributable to improved ventilatory management, while others find that mortality has remained relatively static at 30–50% depending on case mix and severity.

---

6. Clinical Presentation and Diagnostic Workup

6.1 Clinical Presentation

Patients typically present with acute-onset dyspnea and hypoxemia that worsen progressively within 6–72 hours of an inciting event. Physical examination reveals tachypnea, increased work of breathing (accessory muscle use, intercostal retractions), and diffuse crackles (rales) on auscultation, frequently bibasilar. Tachycardia, altered mental status from hypoxemia, and central or peripheral cyanosis may be present.

6.2 Diagnostic Workup

The diagnosis of ARDS is based on the Berlin/Global Definition criteria, requiring:

Imaging:

• Chest radiography or CT: Bilateral opacities not fully explained by effusions, atelectasis, or nodules.
• Lung ultrasound (per Global Definition): Bilateral B-lines or consolidations in a pattern consistent with non-cardiogenic pulmonary edema.
• CT may further characterize disease distribution and identify complications (pneumothorax, pleural effusion, barotrauma).

Oxygenation assessment:

• PaO₂/FiO₂ ratio (requiring arterial blood gas) with minimum PEEP/CPAP of 5 cm H₂O.
• SpO₂/FiO₂ ratio ≤315 (when SpO₂ ≤97%) as alternative per Global Definition.

Cardiac evaluation to exclude cardiogenic edema:

• Echocardiography (preferred noninvasive method) to assess left ventricular systolic and diastolic function.
• B-type natriuretic peptide (BNP) or N-terminal proBNP — though levels overlap substantially between ARDS and cardiogenic edema, particularly when right heart strain or sepsis elevates BNP in ARDS.
• Pulmonary artery catheterization is generally not recommended for routine diagnosis.

Laboratory evaluation:

• Complete blood count with differential, comprehensive metabolic panel, coagulation profile, lactate, troponin, blood cultures, and respiratory viral/bacterial testing.
• Bronchoscopy with bronchoalveolar lavage may be indicated when infection is suspected and noninvasive sampling is negative.

---

7. Differential Diagnosis

Differentiating ARDS from cardiogenic pulmonary edema is the most critical diagnostic challenge, and the two conditions may coexist ("mixed edema"), particularly in elderly patients with heart disease and sepsis. Key distinguishing features:

Feature	ARDS	Cardiogenic Pulmonary Edema
Onset	Within 7 days of insult	Often acute with cardiac event
Risk factors	Sepsis, pneumonia, aspiration, trauma, pancreatitis	Hypertension, coronary disease, valvular disease, cardiomyopathy
Imaging	Patchy, peripheral, asymmetrical opacities	Central/"bat wing" distribution, Kerley B lines, pleural effusions
Chest ultrasound	Spared areas, subpleural consolidation, pleural line abnormalities	Homogeneous B-lines
BNP/NT-proBNP	Variable (may be elevated due to right heart strain)	Markedly elevated
Echocardiography	Normal LV function (usually)	LV systolic or diastolic dysfunction
Response to diuretics	Poor	Often rapid improvement
Pulmonary artery occlusion pressure	≤18 mm Hg (if measured)	>18 mm Hg

Other important differential diagnoses include bilateral pneumonia, diffuse alveolar hemorrhage, acute interstitial pneumonia, acute eosinophilic pneumonia, cryptogenic organizing pneumonia, exacerbation of interstitial lung disease, pulmonary vasculitis, and disseminated malignancy.

---

8. Evidence-Based Management

The management of ARDS is organized around (1) lung-protective mechanical ventilation, (2) adjunctive therapies targeting specific pathophysiological derangements, (3) treatment of the underlying cause, and (4) rescue therapies for refractory cases.

8.1 Lung-Protective Mechanical Ventilation

Low tidal volume ventilation is the single most important intervention. The landmark ARMA trial (ARDS Network, 2000) randomized 861 patients to 6 mL/kg versus 12 mL/kg tidal volume (predicted body weight), with plateau pressure ≤30 cm H₂O in the low-volume group. The trial was stopped early when 28-day mortality was 31.0% in the low-volume group versus 39.8% in the traditional group (P = 0.007), and ventilator-free days were significantly increased. This trial established 6 mL/kg predicted body weight and plateau pressure ≤30 cm H₂O as the standard of care, a recommendation reaffirmed as a strong recommendation by subsequent ATS/ESICM/SCCM guidelines.

Driving pressure (ΔP = tidal volume / respiratory system compliance) has emerged as a critical prognostic marker. A secondary analysis of multiple ARDS Network trials demonstrated that ΔP was the ventilation variable most strongly associated with survival: reductions in ΔP mediated 75% of the treatment effect in tidal volume trials and 45% in PEEP trials. A 1-SD increase in ΔP (approximately 7 cm H₂O) was associated with increased mortality even among patients receiving "protective" tidal volumes and plateau pressures. This finding suggests that ΔP, which reflects the functional "baby lung" size, may be a more useful target than tidal volume normalized to predicted body weight alone.

8.2 Positive End-Expiratory Pressure (PEEP)

The optimal PEEP strategy remains controversial. Higher PEEP can recruit collapsed lung units, improve oxygenation, and reduce atelectrauma, but may also cause overdistension and hemodynamic compromise.

Key trials:

• The EXPRESS trial compared high PEEP (titrated to plateau pressure 28–30 cm H₂O) versus low PEEP and found no difference in hospital mortality, though the high-PEEP group had improved oxygenation and more organ-failure–free days.
• The ART trial (2017) tested a lung recruitment maneuver (prolonged, high-pressure inflation) plus PEEP titration against a conventional low-PEEP strategy. The experimental strategy significantly increased 28-day mortality (55.3% vs. 49.3%; HR 1.20) and 6-month mortality, with more barotrauma (5.6% vs. 1.6%).

Current guidelines:

• Conditional recommendation for higher PEEP without lung recruitment maneuvers in moderate-to-severe ARDS (low-to-moderate certainty of evidence).
• Strong recommendation against prolonged lung recruitment maneuvers (airway pressure >35 cm H₂O for >60 seconds) in moderate-to-severe ARDS due to adverse hemodynamic effects.

Network meta-analysis suggests that higher PEEP without recruitment maneuvers is associated with lower mortality than both lower PEEP and higher PEEP with prolonged recruitment maneuvers.

8.3 Prone Positioning

The PROSEVA trial (2013) provided definitive evidence for prone positioning in severe ARDS. In this multicenter trial, patients with PaO₂/FiO₂ <150 mm Hg who were placed prone for ≥16 hours daily had dramatically lower mortality: 28-day mortality 16.0% in the prone group versus 32.8% in the supine group (P < 0.001), with a hazard ratio for death of 0.39 (95% CI 0.25–0.63). The benefit was attributed to improved V/Q matching, more homogeneous ventilation distribution, and enhanced secretion clearance.

Guidelines make a strong recommendation for prone positioning for >12 hours daily in patients with severe ARDS (PaO₂/FiO₂ <150 mm Hg). Despite this evidence, the LUNG-SAFE study found that prone positioning was used in only 16.3% of severe ARDS patients, highlighting a persistent evidence-to-practice gap.

8.4 Fluid Management

The FACTT trial (ARDS Network, 2006) randomized 1,000 patients to a conservative versus liberal fluid management strategy for seven days. While 60-day mortality did not differ between groups, the conservative strategy significantly improved lung function, shortened duration of mechanical ventilation, and reduced ICU length of stay without increasing non-pulmonary organ failure.

Current practice favors a conservative fluid strategy (targeting net even to negative fluid balance) once shock has resolved, guided by dynamic measures of fluid responsiveness rather than static pressure targets. Pulmonary artery catheterization is not recommended for routine management, as FACTT found no benefit over central venous catheter guidance and more complications with PAC use.

8.5 Neuromuscular Blockade (NMB)

The role of neuromuscular blockade has been one of the most contested areas in ARDS management:

ACURASYS trial (2010): Early, 48-hour infusion of cisatracurium in patients with severe ARDS (PaO₂/FiO₂ <150) was associated with lower adjusted 90-day mortality (HR 0.68; 95% CI 0.48–0.98; P = 0.04) and more ventilator-free days.

ROSE trial (2019): In a contemporary setting with protocolized lung-protective ventilation and lighter sedation targets, early cisatracurium showed no benefit — 90-day mortality was 42.5% in the intervention group versus 42.8% in controls (P = 0.93), and the trial was stopped early for futility. The NMB group had more serious cardiovascular adverse events.

Current guidelines provide a conditional recommendation for NMB use in patients with early (<48 hours) severe ARDS (PaO₂/FiO₂ ≤100 mm Hg), emphasizing that NMB should be used only after optimizing sedation and mechanical ventilation, for a limited duration (≤48 hours when possible). In practice, NMB is now positioned as a selective therapy for ventilator dyssynchrony or refractory hypoxemia rather than for routine use.

8.6 Corticosteroids

The DEXA-ARDS trial (2020) randomized patients with moderate-to-severe ARDS to dexamethasone (20 mg daily for 5 days, then 10 mg for 5 days) or placebo. Dexamethasone significantly increased ventilator-free days (12.3 vs. 7.5) and reduced 60-day mortality (21% vs. 36%).

Pooled analysis of 19 RCTs by the 2023 ATS guideline update indicates that corticosteroids likely decrease mortality and reduce duration of mechanical ventilation and hospital stay, though with increased risk of serious hyperglycemia. The SCCM/ESICM guidelines also recommend steroids for ARDS with PaO₂/FiO₂ <200. The choice of specific corticosteroid appears less important than the decision to administer treatment.

8.7 Extracorporeal Membrane Oxygenation (ECMO)

The EOLIA trial (2018) randomized patients with very severe ARDS (PaO₂/FiO₂ <50 for >3 hours, or <80 for >6 hours, or pH <7.25 with PaCO₂ ≥60) to immediate venovenous ECMO versus continued conventional management (with ECMO permitted as rescue for refractory hypoxemia). At 60 days, 35% of the ECMO group died versus 46% of controls (relative risk 0.76; 95% CI 0.55–1.04; P = 0.09) — a 9% absolute risk reduction that did not reach statistical significance. However, 28% of control patients crossed over to ECMO as rescue, complicating interpretation.

The 2023 ATS guideline provides a conditional recommendation for VV-ECMO in selected patients with severe ARDS (low certainty of evidence), noting that it may decrease mortality and increase organ-failure–free days, but increases hemorrhage risk. ECMO is recommended for patients with severe hypoxemia or hypercapnia despite optimal conventional management, preferably within the first 7 days of ARDS, at experienced centers.

8.8 Rescue Therapies for Refractory Hypoxemia

Inhaled pulmonary vasodilators (inhaled nitric oxide, inhaled epoprostenol) improve oxygenation by selectively dilating vessels in ventilated lung regions, reducing shunt fraction. However, meta-analyses demonstrate no improvement in survival, and inhaled nitric oxide is associated with increased risk of renal replacement therapy. Current guidelines recommend against routine use; these agents are reserved as a temporary bridge to ECMO.

8.9 Treatment of Underlying Cause

Definitive management requires identification and treatment of the precipitating condition: appropriate antimicrobial therapy for sepsis/pneumonia, source control for intra-abdominal infection, supportive care for pancreatitis, and management of shock. Time-sensitive antibiotic administration in septic patients directly influences outcomes.

---

9. ARDS Subphenotypes and Precision Medicine

Latent class analysis of clinical and biomarker data has consistently identified two biological subphenotypes:

• Hyperinflammatory phenotype: Higher plasma IL-6, IL-8, lower serum bicarbonate, higher vasopressor requirements; associated with higher mortality.
• Hypoinflammatory phenotype: Lower inflammatory cytokines, better metabolic profile; lower mortality.

Post-hoc analyses of RCTs suggest differential treatment responses: the hyperinflammatory phenotype may derive greater benefit from higher PEEP, conservative fluid management, and simvastatin, while the hypoinflammatory phenotype does not.

Researchers have developed parsimonious classifiers using as few as 3–4 variables (e.g., IL-8, bicarbonate, vasopressor use) to identify subphenotypes at the bedside, but point-of-capable testing is not yet clinically available. Prospective identification of subphenotypes for predictive enrichment in clinical trials represents a major frontier.

---

10. Outcomes and Long-Term Sequelae

10.1 In-Hospital Mortality

Pre-pandemic mortality rates from LUNG-SAFE: 34.9% (mild), 40.3% (moderate), 46.1% (severe). Mortality is influenced by the underlying cause (higher with sepsis), severity of illness, number of organ failures, and age.

10.2 Pulmonary Recovery

Most survivors return to normal or near-normal pulmonary function within 6–12 months, though a mild restrictive defect and reduced diffusion capacity (DLCO) commonly persist. Pulmonary fibrosis may develop during the fibroproliferative phase and contribute to persistent respiratory symptoms.

10.3 Physical Recovery

Persistent functional limitations are common. At 5-year follow-up, ARDS survivors demonstrated median 6-minute walk distances approximately 76% of predicted, with ICU-acquired weakness as a major determinant. Physical impairments include muscle wasting, joint contractures, tracheal stenosis, entrapment neuropathies, and cosmetic concerns from surgical scars.

10.4 Cognitive Impairment

Cognitive deficits are present in 70–100% of survivors at hospital discharge, 46–80% at 1 year, and approximately 20% at 5 years. Deficits predominantly affect executive function, short-term memory, learning, and psychomotor speed. Delirium during critical illness is the primary predictor of long-term cognitive impairment, with contributions from systemic inflammation, sedative exposure (particularly benzodiazepines), and hypoxemia.

10.5 Psychological Sequelae

Prolonged psychiatric symptoms are highly prevalent. A 5-year longitudinal study found that 52% of survivors experienced continuous or recurring symptoms: anxiety (38%), depression (32%), and post-traumatic stress disorder (23%). Co-occurrence of symptoms across domains is common. Worse pre-ARDS mental health is a strong predictor of long-term psychiatric morbidity.

10.6 Quality of Life

Health-related quality of life (measured by SF-36) remains significantly below population norms at 5 years, particularly in physical domains. These deficits are more strongly linked to muscle weakness and systemic sequelae than to residual pulmonary dysfunction.

---

11. Health Inequities in ARDS

Growing evidence documents substantial disparities in ARDS outcomes:

Racial and ethnic disparities: Black patients have higher ARDS incidence than White patients (OR 1.16 in trauma-associated ARDS). Non-Hispanic Black and American Indian/Alaska Native populations experience the highest age-adjusted mortality rates from respiratory failure. During the COVID-19 pandemic, racial disparities widened further.

Mechanisms of inequity:

• Undetected hypoxemia: Pulse oximetry is less accurate in patients with darker skin tones, leading to occult hypoxemia and delayed recognition of deterioration.
• Differential sedation: Hispanic patients with ARDS have approximately 5 times the odds of receiving deep sedation compared to non-Hispanic White patients, which is associated with worse outcomes.
• Structural factors: Socioeconomic barriers to healthcare access, environmental exposures, multigenerational housing, occupational risk, insurance status, and systemic underinvestment in public health — particularly in the Southern United States — contribute to higher exposure and worse outcomes.
• Hospital factors: Admission to "safety-net" hospitals is associated with increased mortality and longer lengths of stay. Underinsured patients are more likely to receive non-lung-protective ventilation.

---

12. COVID-19-Associated ARDS

The pandemic provided both a natural experiment and a challenge to conventional ARDS management.

12.1 Distinctive Pathobiology

COVID-19-associated ARDS (CARDS) shares core features with classic ARDS but exhibits distinctive vascular pathology:

• Severe endothelialitis: Direct infection of endothelial cells by SARS-CoV-2, with upregulation of adhesion molecules (VCAM1, ICAM1) and downregulation of VE-cadherin.
• Immunothrombosis: Extensive micro- and macro-thrombi, with microthrombi up to 9 times more frequent than in influenza ARDS. NET-mediated platelet activation and suppressed fibrinolysis (elevated plasminogen activator inhibitor-1) contribute.
• L-type and H-type phenotypes: Early CARDS often presents with profound hypoxemia but near-normal respiratory compliance ("L-type"), challenging conventional assumptions that severe hypoxemia implies recruitable, poorly compliant lungs. Later progression to "H-type" with decreased compliance and higher recruitability may occur.

12.2 Management Lessons

• Standard lung-protective ventilation (low tidal volume, driving pressure limitation) remains foundational.
• High PEEP may be deleterious in the L-type phenotype due to overdistension of relatively normal lung regions and adverse hemodynamic effects. PEEP should be individualized based on recruitability assessment.
• Corticosteroids (dexamethasone) demonstrated clear mortality benefit in the RECOVERY trial and are standard of care.
• Anticoagulation: Despite extensive thrombosis, full therapeutic anticoagulation was not superior to prophylactic anticoagulation in randomized trials, emphasizing the inflammatory driver of thrombotic risk.
• The pandemic underscored the need for phenotype-based, personalized ventilatory strategies rather than a "one-size-fits-all" approach.

---

13. Limitations and Evidence Gaps

Definitional limitations: The 2023 Global Definition improves inclusivity but the SpO₂/FiO₂ criterion loses accuracy at high saturations and in patients with darker skin pigmentation. The clinical and prognostic equivalence of cases identified by the resource-limited definition versus the full definition requires further validation.

Pathophysiological gaps: While hyperinflammatory and hypoinflammatory subphenotypes are reproducible in research cohorts, there is no validated point-of-care platform for real-time phenotype identification. The transition from endothelial injury to fibroproliferation is incompletely understood, and no therapies directly target this transition.

Management evidence gaps:

• Driving pressure–guided ventilation has not been prospectively validated as a ventilation target in a randomized trial.
• ECMO: The EOLIA trial did not reach statistical significance; the role of ECMO versus continued conventional therapy is debated, particularly given the high crossover rate. Ongoing trials may clarify benefit in specific subgroups.
• Neuromuscular blockade: The discordance between ACURASYS and ROSE highlights uncertainty about the appropriate comparator (deep vs. light sedation), optimal agent, and duration. The role of NMB in specific phenotypes is unknown.
• Fluid management: The optimal timing, monitoring modality (dynamic vs. static indices), and role of albumin remain undefined.
• Corticosteroids: The optimal agent, dose, timing, and duration — and whether benefit is uniform across subphenotypes — remain uncertain.

Long-term outcome gaps: Mechanisms linking critical illness to persistent cognitive impairment are poorly understood. Effective post-ICU rehabilitation interventions (physical, cognitive, psychological) are lacking, and most survivors do not receive structured follow-up.

Health equity gaps: The mechanisms driving disparities in ARDS care and outcomes are incompletely characterized, and no interventions have been prospectively tested to reduce them.

---

14. TLDR

ARDS is an acute, diffuse, inflammatory lung injury causing non-cardiogenic pulmonary edema and hypoxemic respiratory failure. The diagnostic framework evolved from the 1994 AECC criteria through the 2012 Berlin Definition (mild/moderate/severe categories based on PaO₂/FiO₂) to the 2023 Global Definition, which now includes HFNO, SpO₂/FiO₂ criteria, lung ultrasound, and a resource-limited adaptation. Pathophysiology centers on alveolar–capillary barrier disruption, neutrophil-driven inflammation, surfactant inactivation, heterogeneous lung mechanics, and VILI risk. Direct causes (pneumonia, aspiration) damage the epithelium; indirect causes (sepsis, pancreatitis, trauma) damage the endothelium. Management is anchored by low tidal volume ventilation (6 mL/kg) and plateau pressure ≤30 cm H₂O, with adjunctive therapies including prone positioning (strong recommendation for severe ARDS), conservative fluid strategies once shock resolves, and conditional use of corticosteroids, higher PEEP, neuromuscular blockade (early severe ARDS only, ≤48 hours), and VV-ECMO (selected severe cases). Major trials (ARMA, PROSEVA, FACTT, ACURASYS, ROSE, DEXA-ARDS, EOLIA, ART) have defined and refined the evidence base. Mortality is 35–46%, and survivors face persistent physical, cognitive, and psychological impairments lasting years. COVID-19 ARDS revealed distinctive vascular immunothrombosis and highlighted the need for phenotype-specific care. Health inequities by race, ethnicity, and socioeconomic status are substantial and multifactorial. Major research gaps include real-time subphenotype identification, prospective validation of driving pressure targets, and effective post-ICU rehabilitation interventions.

---

Companion Files

No companion files accompany this report.