I now have all the information needed for a comprehensive answer. Let me compile it.
ARDS: Definition, Management, and Prone Ventilation
Chest Radiograph
Frontal CXR in ARDS: extensive bilateral opacities without features of elevated pressure edema (no interlobular septal thickening, no pleural effusions) — Murray & Nadel's
1. Evolution of ARDS Definitions
1967 — Ashbaugh (Original Description)
12 patients with acute respiratory distress, bilateral opacities on CXR, low compliance, and hypoxemia responding to PEEP. Initially called adult respiratory distress syndrome; later renamed acute to include children.
1988 — Murray Lung Injury Score
Three-part system: (1) acuity, (2) associated risk factors, (3) Lung Injury Score using hypoxemia + PEEP + compliance + CXR. Score >2.5 = ARDS. No exclusion of cardiogenic edema.
1994 — AECC Definition
| Variable | Acute Lung Injury (ALI) | ARDS |
|---|
| Onset | Acute | Acute |
| PaO₂/FiO₂ | ≤300 mmHg | ≤200 mmHg |
| CXR | Bilateral infiltrates | Bilateral infiltrates |
| Exclusion | PAWP ≤18 mmHg or no clinical LV failure | PAWP ≤18 mmHg or no clinical LV failure |
Limitations of AECC: no defined time frame for "acute"; CXR interpretation variability (kappa only 0.55); PaO₂/FiO₂ fluctuates with PEEP level; PAWP criterion problematic.
2012 — Berlin Definition (Current Standard)
Developed by an international expert panel to address AECC limitations. ARDS is defined by all four criteria:
- Timing: Onset within 1 week of a known clinical insult
- Imaging: Bilateral opacities on CXR/CT — not fully explained by effusions, atelectasis, or nodules
- Origin: Respiratory failure not fully explained by cardiac failure or fluid overload (echocardiography recommended if no ARDS risk factor)
- Oxygenation: PaO₂/FiO₂ ≤300 mmHg with PEEP or CPAP ≥5 cmH₂O
Berlin Severity Classification
| Severity | PaO₂/FiO₂ (with PEEP ≥5 cmH₂O) | Mortality |
|---|
| Mild | 200 < PaO₂/FiO₂ ≤ 300 | ~27% |
| Moderate | 100 < PaO₂/FiO₂ ≤ 200 | ~32% |
| Severe | PaO₂/FiO₂ ≤ 100 | ~45% |
The term ALI is now obsolete — those patients are reclassified as mild ARDS under Berlin.
Key improvements over AECC: time frame now explicit (1 week); PAWP criterion removed (PA catheters rarely used); PEEP requirement standardizes oxygenation assessment; recognizes concomitant ARDS + cardiogenic edema is possible.
Remaining limitations: Only marginally better mortality prediction than AECC; interobserver variability remains (κ ≈ 0.5, 67% due to CXR interpretation). — Fishman's Pulmonary Diseases, Murray & Nadel's
2023 — Global Definition of ARDS (Matthay et al.)
The most important update since Berlin. Three categories now exist:
Universal Criteria (apply to all)
- Acute onset within 1 week of insult
- Bilateral opacities not fully explained by effusions/collapse/nodules
- Hypoxemia not fully explained by cardiac failure/fluid overload
Category-Specific Criteria
| Category | Oxygenation Criterion | Setting |
|---|
| Non-intubated ARDS | PaO₂/FiO₂ ≤300 or SpO₂/FiO₂ ≤315 (if SpO₂ ≤97%) on HFNO ≥30 L/min or NIV/CPAP ≥5 cmH₂O | Standard |
| Intubated ARDS | Same Berlin P/F thresholds (mild/moderate/severe) or S/F equivalents | Standard |
| Resource-limited | SpO₂/FiO₂ ≤315 (SpO₂ ≤97%) — no PEEP requirement, no ABG, no CXR required; lung ultrasound accepted | LMICs |
Key changes from Berlin:
- HFNO patients now qualify (captures indolent presentations, e.g., COVID-19)
- SpO₂/FiO₂ (S/F ratio) accepted as substitute for PaO₂/FiO₂ — eliminates need for arterial blood gas
- Lung ultrasound (bilateral B-lines/consolidations) accepted as substitute for CXR in resource-limited settings
- Importantly: does not change management for severe ARDS — prone positioning, lung-protective ventilation, and ECMO remain the core interventions
2. Management of ARDS
Step 1 — Treat the Underlying Cause
Identify and treat the precipitating insult (sepsis, pneumonia, aspiration, trauma, pancreatitis, etc.). This is the most critical step.
Step 2 — Oxygenation Targets
- SpO₂ 88–95% acceptable (permissive hypoxemia to avoid O₂ toxicity)
- FiO₂ <0.7 preferred if achievable with PEEP
- High-flow nasal cannula (HFNO): reasonable first step in non-intubated mild/moderate ARDS (FLORALI trial: lower 90-day mortality vs. NIV face mask)
- NIV via helmet: shown to reduce intubation rate vs. face mask in a single-center RCT (n=83)
Step 3 — Lung-Protective Mechanical Ventilation (ARDSNet Protocol)
| Parameter | Target |
|---|
| Tidal volume (VT) | 6 mL/kg predicted body weight |
| Plateau pressure | ≤30 cmH₂O |
| PEEP | Titrated (higher PEEP strategy for moderate–severe) |
| FiO₂ | Minimize to <0.7 when possible |
| Driving pressure | Target <15 cmH₂O |
| Permissive hypercapnia | pH ≥7.20 acceptable |
The baby lung concept explains the rationale: ARDS lung is heterogeneous, and delivered VT is concentrated in small regions of aerated lung — conventional volumes cause volutrauma and atelectrauma (VILI).
Step 4 — Adjuncts to Mechanical Ventilation
Conservative Fluid Management
- Restrict fluids once haemodynamically stable (FACTT trial: conservative fluid strategy = more ventilator-free days, no mortality benefit)
Neuromuscular Blockade (NMB)
- ACURASYS trial (2010): Cisatracurium ×48h reduced 90-day mortality vs. placebo (31.6% vs. 40.7%) in heavily sedated patients with PaO₂/FiO₂ <150
- ROSE trial (2019, n=1006): NMB showed NO benefit over usual care with light sedation (42.5% vs. 42.8%)
- Current recommendation: Not routine for all moderate–severe ARDS; may be beneficial in severe ventilator dyssynchrony or to facilitate prone positioning
Corticosteroids
- Theoretical benefit: anti-inflammatory, antifibrotic
- No high-dose benefit early; LaSRS (2006) — moderate-dose methylprednisolone in persistent ARDS (days 7–21) showed no mortality benefit and increased re-intubation
- Initiating after day 14 was associated with increased mortality
- Growing evidence for use in overlapping conditions (severe CAP, septic shock)
- Not recommended routinely
ECMO (VV-ECMO)
- Increasingly used at specialized centres for most severe ARDS
- Prone positioning can still be performed in patients on VV-ECMO, though it requires additional expertise
3. ★ PRONE VENTILATION — Special Focus
Physiologic Rationale
Placing a patient prone improves ARDS outcomes through:
- V/Q matching — blood flow redistributes from dorsal (now non-dependent) regions; ventilation becomes more uniform
- Recruitment of dorsal alveoli — previously compressed by the heart and abdominal contents
- Reduced VILI — CT studies show more homogeneous ventilation prone vs. supine, reducing overdistension anteriorly and atelectasis posteriorly
- Heart repositioning — removes cardiac compression from posterior lungs, improving aeration and reducing airway closure
Key Trials
| Trial | Year | Population | Intervention | Result |
|---|
| Earlier trials | Pre-2013 | All ARDS | <12h prone/day | No mortality benefit |
| PROSEVA | 2013 | PaO₂/FiO₂ <150, ≤36h from MV | Prone ≥16h/day + lung-protective ventilation | 28-day mortality: 16% prone vs. 32.8% supine (p<0.001) |
| Meta-analysis (6 RCTs) | — | Prone + protective ventilation | ≥16h prone | RR of death 0.74 (95% CI 0.59–0.95); attenuated without protective ventilation |
PROSEVA Trial — Key Details
- 466 patients, PaO₂/FiO₂ <150 mmHg
- Enrolled within 36 hours of MV initiation
- All patients received low tidal volume ventilation
- Heavy sedation + NMB used in majority of both arms
- 28-day mortality: 16.0% vs. 32.8% — highly significant
- Approximately 4 additional ventilator-free days in prone group
Current Recommendations
Prone positioning is strongly recommended for PaO₂/FiO₂ <150 mmHg after application of adequate PEEP — minimum 16 hours per day — initiated early (within 36h). — Murray & Nadel's, Fishman's
Despite this strong evidence, real-world implementation remains poor: only 3–16% of eligible patients in recent trials received prone positioning, and even trials targeting very severe ARDS (P/F <80) used prone in only 60% before randomisation.
Practical Considerations
| Issue | Detail |
|---|
| Pressure injuries | Nose, face, ears — careful padding required |
| Airway security | ETT, CVCs, arterial lines must be secured before turning |
| Eye care | Pressure on eyes → retinal ischaemia, especially in hypotensive patients |
| Haemodynamic instability | Arrhythmias and haemodynamic shifts possible during turning |
| Cardiac output | Prone may reduce cardiac index up to 25% (especially jackknife position/IVC compression); prior MI or ischaemic heart disease increases risk |
| Contraindications (relative) | Unstable spine, open abdomen, raised ICP, haemodynamic instability |
Awake/Non-Intubated Prone Positioning
During COVID-19, awake prone positioning in spontaneously breathing HFNO patients was widely adopted.
Systematic reviews (PMID 38597483, 37796372) showed improvement in oxygenation and reduced intubation rates in COVID-19 ARDS, though evidence in non-COVID ARDS remains limited.
Prolonged Prone Positioning (≥24h) — 2025 Update
A
2025 meta-analysis (PMID 41199320) of 9 studies (n=1,045) found
no significant mortality benefit for sessions ≥24h vs. standard ≥16h (HR 0.72; 95% CI 0.41–1.25). Evidence was low certainty.
PPP outside clinical trials is not currently supported.
Summary Table
| Feature | AECC 1994 | Berlin 2012 | Global 2023 |
|---|
| Time frame | Not defined | ≤1 week | ≤1 week |
| Imaging | CXR bilateral | CXR/CT | CXR/CT/lung USS |
| Oxygenation | P/F ≤300 or ≤200 | P/F with PEEP ≥5 | P/F or S/F; HFNO patients included |
| PEEP requirement | No | Yes (≥5 cmH₂O) | Yes (or HFNO ≥30 L/min; waived in RLS) |
| Cardiac exclusion | PAWP ≤18 | Echo if no risk factor | Echo if no risk factor |
| Non-intubated | Not included | Not included | Included (new) |
| Resource-limited | Not addressed | Not addressed | Addressed (new) |
| Severity | ALI vs ARDS | Mild/Moderate/Severe | Same (Berlin preserved) |
Sources: Fishman's Pulmonary Diseases and Disorders; Murray & Nadel's Textbook of Respiratory Medicine; Barash Clinical Anesthesia, 9e; Matthay et al. 2023 Global ARDS Definition; PROSEVA trial; ROSE trial; Jung et al. Critical Care 2025 [PMID 41199320]; Tasaka et al. ARDS CPG 2022 [PMID 35753956].