Acute respiratory failure: advanced ventilation modes (APRV and oscillator) · ICU Topics

Cinematic ICU scene of a ventilator in APRV mode showing a high continuous positive airway pressure with brief release phases, a severely hypoxaemic ARDS patient prone on the bed, an ECMO cannulation set standing by, clinical-blue lighting, medical educational, no faces, no text — FigureAPRV (airway pressure release ventilation) maintains a continuous high pressure (P-high) to keep the lung recruited with brief releases (T-low) to a low pressure (P-low) for ventilation — an open-lung, spontaneous-breathing strategy for refractory hypoxaemia. HFOV delivers very small tidal volumes at high frequency; trials (OSCILLATE, OSCAR) showed no outcome benefit and possible harm, so it is now reserved for exceptional cases. Both are rescue therapies, not routine.

APRV vs HFOV vs conventional ventilation

Feature	Conventional (VC/PC)	APRV	HFOV (Oscillator)
Mechanism	Volume or pressure controlled, cycled	Continuous high pressure + brief releases	Oscillating diaphragm at 3-15 Hz
Tidal volume	4-6 mL/kg (protective)	Variable (release-dependent)	1-3 mL/kg (tiny)
PEEP	Moderate (from PEEP/FiO2 table)	HIGH (P-high 25-35 cmH2O)	Mean airway pressure (high)
Spontaneous breathing	Possible (if mode allows)	YES (throughout — key feature)	NO (paralysis usually needed)
Sedation	Moderate-deep	Less (spontaneous)	Deep + paralysis
Oxygenation	Good	Often IMPROVED	Often improved
Mortality benefit	YES (low Vt — ARMA)	NO clear benefit	NO (OSCILLATE: trend to harm)
Current role	FIRST-LINE	Selected (salvage)	REFRACTORY only

[1]

When to consider advanced ventilation modes

Optimise conventional ventilation FIRST — low tidal volume (4-6 mL/kg), plateau <30, PEEP/FiO2 table, POCUS-guided PEEP
Try PRONE POSITIONING — PROSEVA: for severe ARDS (PaO2/FiO2 <150), >16h/day
Consider neuromuscular blockade — early severe ARDS (cisatracurium 48h — ACURASYS, but ROSE was negative)
If still refractory (PaO2/FiO2 <80):
- APRV: if patient can breathe spontaneously, may improve oxygenation and reduce sedation. Setting: P-high 25-30, P-low 0-5, T-high 4-6s, T-low 0.5-0.8s
- ECMO: for refractory hypoxaemia (VV-ECMO) — stronger evidence than HFOV
HFOV: LAST RESORT — only if APRV and ECMO unavailable/failed. Oscillator at mean Paw 25-30, frequency 4-6 Hz, delta pressure 60-90

[1]

Exam practice

SAQ — Choosing and setting APRV in refractory ARDS (vs HFOV)

10 minutes · 10 marks

A 40-year-old man (height 175 cm, PBW ~71 kg) with severe ARDS from H1N1 influenza is on day 2 of mechanical ventilation. Optimised conventional ventilation: Vt 420 mL (6 mL/kg PBW), RR 30, PEEP 16, FiO2 0.95. Plateau pressure 30 cmH2O, driving pressure 14 cmH2O. Despite 18 hours of prone positioning, ABG: pH 7.29, PaCO2 52, PaO2 59, HCO3 24 (P/F = 62). Noradrenaline 0.15 mcg/kg/min, RASS -1, triggering the ventilator. You are considering an advanced ventilation mode.

[1]

SAQ — Liberation from APRV after recovery

10 minutes · 10 marks

A 58-year-old woman has been on APRV for 6 days for severe ARDS from aspiration pneumonitis. Current settings: P-high 26 cmH2O, P-low 0, T-high 5 s, T-low 0.7 s, FiO2 0.30. ABG: pH 7.42, PaCO2 40, PaO2 92, HCO3 26. She is triggering comfortably at a spontaneous rate of 20/min, RASS -1, on no vasopressors, with a productive cough and improving compliance. P/F ratio is now 307. You plan to liberate her from APRV.

Clinical pearls

High-yield advanced ventilation points for CICM/FFICM exam

HFOV does NOT improve mortality — OSCILLATE trial showed trend to HARM. Ferguson 2013 (NEJM): HFOV vs conventional in early moderate-severe ARDS. Trial STOPPED EARLY — HFOV had HIGHER mortality (47% vs 35%, p=0.005), more sedation, more vasopressors. HFOV now RESERVED for refractory cases (not first-line). OSCAR trial (UK): no mortality benefit.[1] }
APRV allows SPONTANEOUS breathing throughout the cycle. This is the KEY feature. Unlike conventional ventilation (where patient breathes between machine breaths), APRV maintains continuous positive pressure (P-high) with brief releases (to P-low). Patient can breathe ANY time (during P-high and P-low). Benefits: less sedation, preserved diaphragm function, improved venous return, less muscle atrophy.[3] }
APRV settings: P-high, P-low, T-high, T-low. P-high: 25-35 cmH2O (high continuous pressure — maintains recruitment). P-low: 0-5 cmH2O (release pressure — allows exhalation). T-high: 4-6 seconds (most of cycle at high pressure). T-low: 0.5-0.8 seconds (brief release — typically terminated at 50-75% of peak expiratory flow to avoid derecruitment). Typical I:E ratio: 8:1 to 10:1 (inverse).[5] }
APRV theoretical benefits (not proven). (1) Continuous recruitment (high PEEP throughout). (2) Less derecruitment (brief releases only). (3) Spontaneous breathing → improved V/Q (diaphragm contraction → dependent lung recruitment). (4) Less sedation needed (patient comfortable, can breathe). (5) Cardiac: spontaneous breathing → negative intrathoracic pressure → improved venous return → better cardiac output. BUT: no RCT shows mortality benefit.[3] }
Standard ARDS ventilation remains the FOUNDATION. Low tidal volume (4-6 mL/kg — ARMA trial), plateau pressure <30 cmH2O, PEEP from PEEP/FiO2 table, driving pressure <15. These have PROVEN mortality benefit. APRV and HFOV are ADJUNCTS for refractory cases — not replacements for lung-protective ventilation.[4] }
Prone positioning has BETTER evidence than APRV or HFOV. PROSEVA trial (2013): prone >16h/day for severe ARDS reduced mortality from 33% to 16% (NNT 6). This is PROVEN, SIMPLE, and AVAILABLE everywhere. Always try prone positioning BEFORE considering APRV or HFOV.[4] }
ECMO is the strongest salvage therapy for refractory ARDS. VV-ECMO: provides oxygenation while lungs rest. EOLIA trial (2018): trend toward benefit (not statistically significant, but crossed futility boundary early). CESAtrial (ongoing). ECMO is NOW preferred over HFOV for refractory ARDS (better evidence, lower complications).[4] }
HFOV physiology — very different from conventional. (1) Oscillating diaphragm creates sinusoidal pressure wave (positive AND negative). (2) Frequency: 3-15 Hz (180-900 breaths/min). (3) Tidal volume: ~1-3 mL/kg (MUCH smaller than conventional). (4) Gas exchange: by CONVECTION + DIFFUSION (not bulk flow). (5) Mean airway pressure: high (25-30 cmH2O) — maintains recruitment. (6) Theory: lung protection (tiny Vt → less volutrauma) + recruitment (high mean Paw).[6] }
Why HFOV failed (OSCILLATE trial). Possible explanations: (1) HFOV requires HEAVY SEDATION + PARALYSIS → complications (weakness, VAP, prolonged ventilation). (2) HIGH mean airway pressure → reduced venous return → haemodynamic instability (more vasopressors). (3) Theoretical lung protection NOT realised in practice. (4) May cause RIGHT HEART FAILURE (high intrathoracic pressure → high pulmonary vascular resistance). (5) Delay in starting conventional lung-protective ventilation.[1] }
APRV may be useful for DYSSYNCHRONY. If patient is fighting the ventilator (patient-ventilator dyssynchrony) on conventional modes → consider APRV (allows spontaneous breathing at any time). May reduce: sedation requirements, agitation, dyssynchrony. Useful in patients who WANT to breathe but conventional mode fights them.[5] }
Transitioning from conventional to APRV. (1) Start P-high at plateau pressure from conventional mode (e.g., if plateau was 28, start P-high 28-30). (2) P-low: 0 (allow full release). (3) T-high: 4-6 seconds. (4) T-low: 0.5-0.8 seconds (titrate to terminate at 50-75% of peak expiratory flow rate). (5) FiO2: same as conventional. (6) Monitor: oxygenation, CO2 (may increase initially), haemodynamics, patient comfort.[5] }
Transitioning from APRV back to conventional. (1) Gradually REDUCE P-high (by 2 cmH2O every 2-4h). (2) Gradually SHORTEN T-high (toward conventional I:E). (3) When P-high ~15-20 and T-high ~1-2s → switch to PSV/CPAP (weaning mode). (4) Extubate if tolerating spontaneous breathing trial.[5] }
Complications of APRV. (1) AUTO-PEEP/air trapping (if T-low too long — incomplete release). (2) DYSSYNCHRONY (if release timing doesn't match patient's breathing). (3) HYPERCAPNIA (if T-high too long — insufficient ventilation). (4) HAEMODYNAMIC instability (high intrathoracic pressure → reduced venous return — though spontaneous breathing mitigates this). (5) Requires EXPERIENCED operator (less familiar than conventional modes).[3] }
Both APRV and HFOV should be used in SPECIALIST centres. These advanced modes require: (1) Experienced clinicians (respiratory therapists, intensivists). (2) Specialised ventilators (especially oscillator — not available everywhere). (3) Close monitoring (blood gases, haemodynamics). (4) Protocols for transition to/from conventional. If no expertise: stick with conventional lung-protective ventilation + prone + ECMO.[4] }

Red flags

Prognosis

OSCILLATE trial (Ferguson 2013, NEJM) — HFOV in ARDS

RCT: 548 patients with moderate-severe ARDS (PaO2/FiO2 <200). HFOV vs conventional low-tidal-volume ventilation.

In-hospital mortality: HFOV 47% vs conventional 35% (RR 1.33, p=0.005) — HFOV WORSE
Trial STOPPED EARLY: for safety (HFOV clearly harmful)
Vasopressor use: MORE with HFOV
CONCLUSION: HFOV is HARMFUL in early ARDS. Should NOT be used routinely. Reserve for refractory cases only. [1]

OSCAR trial (Young 2013, NEJM): HFOV vs conventional in UK ARDS. No significant difference in 30-day mortality (41% vs 39%). No benefit. APRV systematic reviews: improved oxygenation in some studies, no clear mortality benefit. Insufficient evidence for routine use.

[1]

APRV in depth — mechanics, settings and tuning

APRV waveform showing P-high, T-high, P-low, T-low and brief release phases with spontaneous breathing on the high pressure — FigureAPRV four variables — P-high/T-high maintain recruitment; brief T-low releases allow CO2 clearance while limiting derecruitment. Spontaneous breathing on P-high is intentional.

Airway pressure release ventilation (APRV) is best understood as two levels of continuous positive airway pressure (CPAP) with a time-cycled release. The ventilator holds a high pressure (P-high) for most of the respiratory cycle and briefly releases to a low pressure (P-low). The patient may breathe spontaneously at any point — during P-high and during the release — and these spontaneous breaths are superimposed on the pressure waveform. The defining feature is therefore not the release itself but the near-continuous mean airway pressure with preserved spontaneous effort (an "open-lung" strategy).[3] }

The four APRV variables

APRV settings — the four variables and how to set them

Variable	Typical range	What it controls	How to titrate
P-high (high pressure)	25-35 cmH2O	Mean airway pressure → recruitment & oxygenation	Start at plateau pressure from conventional mode (or 25-30). Limit by driving pressure / plateau <30 and haemodynamics
P-low (release pressure)	0-5 cmH2O (often 0, i.e. atmospheric)	How much the lung is allowed to derecruit on release	Usually set to 0 (full release). Raising P-low reduces Vt and CO2 clearance but may stabilise haemodynamics
T-high (release interval)	4-6 s (long)	Time at recruitment pressure; the longer T-high, the higher the mean Paw	Increase to improve oxygenation; reduce if hypercapnia/ventilation inadequate
T-low (release time)	0.2-0.8 s (brief)	Time allowed for exhalation — must be SHORT to avoid derecruitment	Titrate to terminate at 50-75% of peak expiratory flow (T-PEF), NOT a fixed time

[1]

The inverse I:E ratio is a consequence, not a setting: because T-high is long and T-low is brief, the ratio is typically 8:1 to 10:1 (sometimes 5:1 to 15:1). This keeps the alveolus open for most of the cycle, minimising tidal opening-and-closing (atelectrauma). Oxygenation is governed mainly by the mean airway pressure (dominated by P-high and the long T-high); CO2 clearance is governed by the release volume (the bigger the swing between P-high and P-low during T-low, the more gas exchanged) plus the patient's own spontaneous breaths.[5] }

The T-low concept — termination of peak expiratory flow (T-PEF)

This is the single most important and most misunderstood APRV concept for the exam. During the brief release, expiratory flow spikes then decays as the lung empties toward P-low. If T-low is too long, the lung fully empties → derecruitment (loss of the recruitment you spent T-high building). If T-low is too short, exhalation is incomplete → air trapping and intrinsic PEEP (auto-PEEP). The standard approach is to end the release when expiratory flow has fallen to 50-75% of its peak — this preserves an "open-lung" with a small residual alveolar pressure while still clearing CO2. In practice T-low is therefore 0.2-0.8 s in most adults, adjusted by watching the expiratory flow waveform.[3] }

Why spontaneous breathing matters

In conventional controlled ventilation the diaphragm is largely inactive; dependent (dorsal) lung — the most oedematous/collapsed region in ARDS — receives little ventilation and is poorly recruited. With active diaphragmatic contraction, pleural pressure falls preferentially in dependent zones, recruiting dorsal alveoli and improving V/Q matching. Spontaneous effort also generates negative intrathoracic pressure, which augments venous return and reduces the right-ventricular afterload increase seen with positive-pressure ventilation. This is the physiological rationale for preferring APRV over deep sedation + paralysis when the patient can breathe.[3] }

"True APRV" vs BiLevel / Biphasic / BiVent

Vendor names confuse the literature. APRV classically means a long T-high with one or two brief releases per minute (extreme inverse ratio). BiLevel / BiVent / DuoPAP use the same two-CPAP architecture but with more comparable times (e.g. P-high 4 s, P-low 2 s) and are often programmed like a pressure-targeted mode with pressure support added on top. For the exam the unifying concept is: all are pressure-targeted, time-cycled, two-level modes that allow spontaneous breathing throughout; the difference is the time ratio. The clinical claim of APRV (extreme inverse ratio, very brief releases) is the most recruitment-favourable.[3] }

Starting and titrating APRV — a practical protocol

Set P-high at the plateau pressure of the preceding conventional mode (or 25-30 cmH2O if de novo). Respect a driving pressure <15 cmH2O and avoid P-high that drops the SpO2-corrected mean arterial pressure.
Set P-low to 0 cmH2O (atmospheric) — full release maximises CO2 clearance. Add 3-5 cmH2O only if haemodynamically unstable or if release volume is excessive.
Set T-high to 4-6 s (start 4 s). Increase if oxygenation poor; this raises mean Paw.
Set T-low to 0.5-0.8 s, then titrate to terminate at 50-75% of peak expiratory flow on the flow waveform. Watch for auto-PEEP.
Set FiO2 the same as on conventional; reduce as oxygenation improves.
Monitor: arterial/venous blood gases at 30-60 min (expect a transient rise in PaCO2), mean arterial pressure, urine output, spontaneous effort (sedation to a Richmond score of −1 to −2 — the patient should be breathing).
Re-tune: failing oxygenation → ↑ P-high (by 2 cmH2O) or ↑ T-high; failing ventilation → ↑ release gradient (lower P-low or lengthen T-low cautiously) or accept permissive hypercapnia (keep pH >7.20).
Daily: lung-protective limits still apply (release Vt ~4-6 mL/kg PBW, plateau/driving pressure bounded); plan transition to conventional/PSV as lung injury resolves.

[1]

Weaning from APRV back to conventional ventilation

Lower P-high by 2 cmH2O every 2-4 h as oxygenation and compliance improve (target P-high ~20-22).
Shorten T-high toward a more conventional I:E ratio (e.g. 4 s → 2 s → 1 s) — the mode becomes BiLevel-like.
When P-high is ~15-20 cmH2O with a near-symmetric ratio, switch to pressure support (PSV/CPAP) with PS 8-12 cmH2O and PEEP 5-8.
Perform a spontaneous breathing trial; extubate if tolerated (rapid shallow breathing index <105, stable ABG, intact cough/mental status).
Avoid prolonged APRV in a recovering patient — the very feature that recruits in early ARDS (high mean pressure) can delay liberation once compliance improves.

[1]

HFOV in depth — mechanism, tuning and why the trials failed

Physiology of open-lung APRV and oscillatory gas exchange concepts for refractory hypoxaemia — FigureOpen-lung physiology — prolonged high mean airway pressure recruits dorsal units; release duration is short enough to limit full exhalation and keep end-expiratory lung volume high.

HFOV concept and evidence summary: high mean airway pressure, small tidal volumes, OSCILLATE increased mortality, OSCAR no benefit — FigureHFOV — rescue only. OSCILLATE showed increased mortality; OSCAR showed no benefit. Not first-line ARDS therapy.

How the oscillator works

High-frequency oscillatory ventilation (HFOV) uses an oscillating piston/diaphragm (driven by an electromagnet in adult machines such as the SensorMedics 3100B) to generate a sinusoidal pressure wave around a high mean airway pressure. Crucially, both inspiration and expiration are active — the diaphragm actively pulls back during expiration, which differentiates HFOV from high-frequency jet or percussive ventilation where expiration is passive. A continuous bias flow washes CO2 from the circuit.[5] }

HFOV variables

HFOV settings — what each knob does

Variable	Typical adult range	Effect of increasing it
Mean airway pressure (mPaw)	25-30 cmH2O (above the conventional PEEP)	↑ Oxygenation (recruitment) — but ↑ intrathoracic pressure → ↓ venous return, ↑ PVR
Frequency	3-6 Hz (range 3-15 Hz; higher for small lungs)	↑ Frequency → ↓ tidal volume → ↑ CO2 (counter-intuitive!)
ΔP (power/amplitude)	60-90 cmH2O	↑ ΔP → ↑ tidal volume → ↑ CO2 clearance (the "chest wiggle" should be visible to the iliac crests)
Inspiratory time %	~33%	Minor effect on Vt
Bias flow	20-40 L/min	Provides fresh gas; ↑ bias flow aids CO2 washout
FiO2	As required	Oxygenation

[1]

Gas exchange below dead space — the core paradox

HFOV delivers tidal volumes of ~1-3 mL/kg — less than anatomical dead space. Bulk convective flow therefore cannot explain gas exchange. Instead, gas transport in HFOV relies on: (1) direct alveolar ventilation of proximal, short-pathway units; (2) convective dispersion / Taylor dispersion (the oscillating axial flow smears gas along the bronchial tree); (3) pendelluft — asynchronous filling between lung units with different time constants, so gas moves between alveoli rather than in and out; (4) molecular diffusion in the alveolar region; and (5) cardiogenic mixing. The net effect is alveolar ventilation achieved without bulk tidal flow, keeping the lung open and minimising volutrauma and atelectrauma in theory.[5] }

Tuning logic — oxygenation vs ventilation are decoupled

On HFOV the two failure modes are managed with different knobs, unlike conventional ventilation:

Hypoxaemia → increase mPaw (recruitment) and FiO2.
Hypercapnia → decrease frequency (counter-intuitive — slower = bigger Vt) and/or increase ΔP (bigger swing = bigger Vt), or increase bias flow. [1]

Why OSCILLATE and OSCAR changed practice

Before 2013 HFOV was widely used in severe ARDS on the basis of physiological plausibility and oxygenation improvement. Two large RCTs published simultaneously in the NEJM in 2013 reversed this:

OSCILLATE (Ferguson): stopped early for harm — in-hospital mortality 47% vs 35% (RR 1.33), more vasopressors, more sedation.[1] }
OSCAR (Young): no mortality difference at 30 days.[2] }

The 2016 Cochrane review (Sud) concluded HFOV may increase in-hospital mortality and provides no survival benefit. HFOV is therefore now reserved for rescue in refractory hypoxaemia when all other options (lung-protective ventilation, prone, ECMO) have been exhausted or are unavailable.[12] }

Plausible mechanisms of harm

(1) Heavy sedation and prolonged paralysis required for HFOV → weakness, ICU-acquired weakness, prolonged ventilation, VAP. (2) High mean airway pressure → reduced venous return and increased pulmonary vascular resistance → haemodynamic instability and acute cor pulmonale / right heart failure. (3) The theoretical lung protection was never realised because mean airway pressure was often set above the upper inflection point of the pressure-volume curve, causing overdistension rather than protection. (4) Delay in instituting proven conventional lung-protective ventilation. (5) Inadequate humidification and airway complications in older oscillator circuits.[1] }

Conventional and dual-control modes — PCV, PSV, SIMV, PRVC

Advanced mode selection is built on three control variables (pressure, volume, flow) and three phase variables (trigger, limit, cycle). A mode is fully described by stating what triggers the breath, what limits it, and what cycles it off, plus whether spontaneous breaths are permitted. This "rational framework" is the CICM/FFICM-friendly way to compare modes.[6] }

PCV vs PSV vs SIMV vs dual-control (PRVC/VC+) — the four mode families

Feature	PCV (pressure control)	PSV (pressure support)	SIMV (synchronised intermittent mandatory)	PRVC / VC+ (dual control, volume-targeted)
Trigger	Time (or patient)	Patient (flow/pressure)	Time for mandatory + patient for spontaneous	Time (or patient)
Limit	Set pressure (constant)	Set pressure support	Set pressure/volume for mandatory	Pressure, but auto-adjusted breath-to-breath
Cycle	Time (set I-time)	Flow (cycles at ~25% of peak flow)	Time (mandatory) / flow (spontaneous)	Time (decelerating flow to target volume)
Flow pattern	Decelerating	Decelerating	Variable	Decelerating
Vt delivered	Variable (depends on compliance/resistance)	Variable (patient + effort-dependent)	Variable for mandatory; variable for spontaneous	Guaranteed (auto-adjusts pressure to hit set Vt)
Key strength	Limits pressure; good leak tolerance; uniform gas distribution	Patient-ventilator synchrony; for weaning	Backup rate guaranteed	Volume guarantee with decelerating-flow benefit
Key risk	Hypoventilation if compliance falls (Vt drops silently)	Ineffective triggering; auto-cycling; apnoea if drive lost	Prolongs ventilation (superseded); stacking	Can still overdistend if set Vt too high
Typical use	ARDS (pressure-limited); severe dyssynchrony	Liberation/weaning; spontaneous breathing patients	Largely historical — avoid for weaning	Default volume-targeted mode on many ICU ventilators

[1]

Why SIMV has fallen out of favour

Synchronised intermittent mandatory ventilation delivers a set number of mandatory, synchronised breaths and allows unsupported (or pressure-supported) spontaneous breaths in between. Two landmark RCTs — Brochard (1994) and Esteban (1995) — showed that SIMV-based weaning prolongs the duration of mechanical ventilation compared with pressure-support or T-piece weaning, likely because the mandatory rate buffers the work of breathing unevenly and the patient "fights" the mandatory breaths. SIMV is therefore now considered inferior for liberation and is largely retained only for initial ventilation or as a backup pattern.[6] }

Dual-control / closed-loop volume targeting

PRVC (Pressure Regulated Volume Control; vendor names: AutoFlow, VC+, APV) combines the decelerating flow and pressure-limiting benefits of pressure control with a volume guarantee. The ventilator delivers a "test" breath, measures compliance, and then breath-to-breath adjusts the inspiratory pressure (within a safety cap) to hit the set tidal volume. The clinician sets Vt, PEEP, FiO2, I-time/rate; the machine manages pressure. This is now the default volume-targeted mode on most modern ICU ventilators because it tolerates small leaks, compensates for changing compliance (e.g. worsening oedema), and still bounds peak pressure. It does not by itself guarantee lung protection — the clinician must still ensure Vt 4-6 mL/kg PBW and plateau/driving pressure limits.[6] }

Adaptive and closed-loop modes — NAVA and PAV+

Conventional and dual-control modes decide how much support to give based on set pressure or volume. The next generation decides based on the patient's own instantaneous effort — so-called "patient-driven" or adaptive modes. The unifying principle: support is proportional to demand, which improves synchrony, avoids both under- and over-assist, and aims to protect the diaphragm from disuse atrophy as well as the lung from overdistension.[4] }

NAVA vs PAV+ — patient-driven adaptive modes

Feature	NAVA (neurally adjusted ventilatory assist)	PAV+ (proportional assist ventilation, load-adjustable)
Signal that drives support	Electrical activity of the diaphragm (EAdi) via a specialised nasogastric catheter with electrodes at the oesophago-gastric junction	Patient's inspiratory flow and volume; resistance & elastance estimated intermittently by the ventilator
How support scales	Pressure = NAVA level (cmH2O/µV) × EAdi — linear with neural drive	Pressure = (gain × elastance × volume) + (gain × resistance × flow) — proportional to instantaneous effort
Trigger	Neural (EAdi) — bypasses pneumatic trigger	Flow/pressure (pneumatic)
Cycle-off	Neural (when EAdi falls to 70% of peak)	Flow/volume (patient effort)
Strengths	Eliminates ineffective triggering; works despite auto-PEEP (COPD); prevents over-assist; patient fully controls rate, Vt, Ti	Amplifies patient effort in proportion to demand; auto-tunes gain to compliance; lung- and diaphragm-protective
Limitations	Catheter placement; EAdi affected by oesophageal position, electrocautery, sedation; cost	Needs reliable effort; runaway risk if gain > ~90% of elastance; not for profoundly weak drive
Backup	PSV backup if apnoea / EAdi lost	Apnoea backup (PSV/PCV)
Best evidence	Improved synchrony, reduced over-assist; mortality benefit not proven	Improved synchrony; trials ongoing for earlier liberation

[1]

NAVA in detail

NAVA delivers pressure in direct proportion to the electrical activity of the diaphragm (EAdi), captured by an array of electrodes on a specialised feeding tube positioned at the level of the crural diaphragm. The breath is therefore triggered, limited, and cycled by the patient's neural output rather than by airway pressure or flow. Because the trigger is neural, it is immune to auto-PEEP — a major advantage in COPD/asthma where pneumatic triggering is delayed by intrinsic PEEP and most "ineffective efforts" occur. The clinician sets a NAVA level (cmH2O per µV of EAdi), which determines how aggressively neural drive is amplified. An intrinsic safety feature: if the patient stops breathing (deep sedation, brainstem injury), a backup pressure-support mode takes over. Randomised data show improved patient-ventilator synchrony and less over-assist; no mortality benefit has yet been demonstrated.[13] }

PAV+ in detail

PAV+ delivers pressure proportional to the patient's instantaneous inspiratory effort — it amplifies the work the patient is already doing ("you press, I push with you"). To do this the ventilator must know the patient's resistance and elastance, which PAV+ estimates by intermittently applying brief end-inspiratory and end-expiratory pauses (the "load-adjustable gain factors" method). The gain is then automatically tuned to a fraction (≤ ~90%) of the measured load. Because the patient controls rate, tidal volume, flow and inspiratory time entirely, PAV+ provides variable support that tracks changing demand (fever, agitation, sepsis) and may both protect the lung (less overdistension) and the diaphragm (less disuse atrophy). The principal hazard is runaway: if the gain approaches or exceeds 100% of the patient's own elastance/resistance, the system becomes self-exciting and generates very large tidal volumes — modern PAV+ limits the gain to prevent this and requires a preserved respiratory drive.[14] }

Adaptive Support Ventilation (ASV) — closed-loop targeting

ASV is a closed-loop mode that targets a clinician-set minute volume and uses the minimum work-of-breathing principle: it continuously adjusts rate, tidal volume, pressure support and I:E ratio toward the calculated least-effort pattern (a derivative of Otis' equation). If the patient is apnoeic it delivers pressure-controlled mandatory breaths; as the patient's effort increases, mandatory support falls away. ASV has been shown to reduce the number of manual ventilator adjustments and may shorten time to liberation, though — like NAVA and PAV+ — without a proven mortality benefit. It is best understood as an automated, lung- and diaphragm-aware version of SIMV-plus-PSV.[6] }

High-yield clinical pearls — advanced and adaptive modes

Advanced ventilation pearls for CICM/FFICM/EDIC — beyond the basics

Describe a mode by trigger–limit–cycle. PCV = time/pressure/time; PSV = patient/pressure/flow; PRVC = time/pressure(auto)/time-to-volume; NAVA = neural/effort-proportional/neural. If you can name all three phase variables you understand the mode.[6] }
APRV is two CPAP levels with a time-cycled release — NOT a mode that "gives a breath". There is no set tidal volume; the release generates the bulk of ventilation, augmented by spontaneous breaths. The mean airway pressure is dominated by the long T-high and is what maintains recruitment.[3] }
T-low is set by watching expiratory flow, not by a clock. The release should end at 50-75% of peak expiratory flow (T-PEF). End too late → derecruitment; end too early → air-trapping/auto-PEEP. This single waveform skill distinguishes a competent APRV operator.[3] }
On HFOV, raising the frequency REDUCES ventilation and raises CO2 — the most counter-intuitive fact in advanced ventilation. To clear CO2 you lower frequency and/or raise ΔP (power). To improve oxygenation you raise mean airway pressure.[5] }
HFOV tidal volumes (1-3 mL/kg) are below anatomical dead space. Gas exchange therefore occurs by convective dispersion, pendelluft and molecular diffusion — NOT bulk flow. This is why HFOV can ventilate at all despite "impossibly small" Vt.[5] }
HFOV actively expires. Unlike high-frequency jet or percussive ventilation (passive expiration), the oscillator's diaphragm pulls back on expiration — this is what prevents gas trapping and is the defining mechanical feature of "oscillation".[5] }
The OSCILLATE harm signal is the single most testable fact in this topic. Stopped early; in-hospital mortality 47% vs 35% (RR 1.33); more vasopressors and sedation. HFOV is therefore rescue only, behind lung-protective ventilation, prone and ECMO.[1] }
PSV is flow-cycled — it ends when inspiratory flow falls to ~25% of peak. This means the patient, not the machine, ends the breath, which is why PSV is the weaning mode. Trouble: in COPD the high internal resistance delays flow decay → prolonged inspiration and dynamic hyperinflation; cycle the ventilator earlier (raise flow cycle-off %) or add expiratory trigger adjustment.[6] }
PCV silently drops tidal volume if compliance falls. Because pressure is fixed, a worsening lung (oedema, pleural effusion, mainstem intubation) reduces Vt without any alarm until minute ventilation collapses. Always watch exhaled Vt and driving pressure on PCV.[6] }
PRVC/AutoFlow guarantees volume by auto-adjusting pressure breath-to-breath. It is the modern default volume-targeted mode — but it does NOT prevent overdistension. You must still cap Vt at 4-6 mL/kg PBW and watch driving pressure <15.[6] }
SIMV prolongs ventilation — do not wean with it. Brochard (1994) and Esteban (1995) both showed SIMV-based weaning takes longer than PSV or T-piece. SIMV is now historical for liberation; retain it only for backup rate.[6] }
NAVA triggers off the diaphragm (EAdi) and is immune to auto-PEEP. This makes it the synchrony mode of choice in severe COPD/asthma where ineffective triggering dominates. Backup PSV covers apnoea. Gain is set as a NAVA level (cmH2O/µV).[13] }
PAV+ amplifies patient effort and can RUNAWAY. If the gain approaches the patient's own elastance/resistance (>~90%), the loop self-excites → huge Vt. Modern PAV+ caps the gain and auto-measures resistance/elastance (the "load-adjustable gain factors" of Xirouchaki 2008).[14] }
The common thread of NAVA and PAV+ is "support proportional to demand". Both aim to avoid over-assist (which causes diaphragm disuse atrophy) and under-assist (high work of breathing). Neither has proven a mortality benefit — improved synchrony and lung/diaphragm protection are the demonstrated gains.[4] }
Never forget the proven bundle before reaching for advanced modes. Low tidal volume (4-6 mL/kg, ARMA), plateau <30 cmH2O, driving pressure <15, prone ≥16 h/day for PaO2/FiO2 <150 (PROSEVA), and VV-ECMO for refractory hypoxaemia (EOLIA) all have better evidence than APRV, HFOV, NAVA or PAV+.[7] }
Prone positioning (PROSEVA) reduces mortality from ~33% to ~16% in severe ARDS — NNT ~6. It is cheap, available everywhere and proven; always deploy it before APRV/HFOV.[8] }
ECMO is now preferred over HFOV for refractory ARDS. EOLIA (2018) showed a strong trend to benefit (crossed the futility boundary early); a Bayesian re-analysis suggested probable benefit. ECMO rests the lung without the haemodynamic cost of high mean airway pressure.[9] }
Cisatracurium for 48 h (ACURASYS) reduced mortality in early severe ARDS — but ROSE (2019) did NOT confirm it. Current practice: short-course paralysis is reasonable for severe dyssynchrony/dangerous transpulmonary pressures, but routine 48-h paralysis is not standard.[10] }
The diaphragm is now a target organ. Both over-assist (disuse atrophy) and under-assist (load-induced injury) cause diaphragm dysfunction and prolong ventilation. This is the rationale for NAVA/PAV+ and for monitoring EAdi and respiratory effort even on conventional modes.[4] }
APRV is pointless in a paralysed patient. Its whole advantage is preserved spontaneous breathing. If you must paralyse (e.g. for severe dyssynchrony on HFOV), APRV collapses into an inverse-ratio pressure-controlled mode with no proven benefit.[3] }
Mean airway pressure, not PEEP, is the recruitment variable on APRV and HFOV. On APRV raise it by lengthening T-high or raising P-high; on HFOV by raising mPaw directly. Watch the haemodynamic and RV cost.[5] }
All advanced modes belong in specialist centres with protocols for transition. The evidence for any of them is weaker than the proven bundle; the risk of harm (auto-PEEP, RV failure, runaway, prolonged ventilation) is real. If the expertise is absent, stick with conventional lung-protective ventilation + prone + ECMO referral.[5] }

Additional trials — the evidence base

OSCAR trial (Young 2013, NEJM) — HFOV in UK ARDS

RCT: 795 patients with ARDS (PaO2/FiO2 <200) across 37 UK ICUs. HFOV vs conventional ventilation.

30-day mortality: HFOV 41.3% vs conventional 41.6% — no significant difference
No difference in duration of ventilation, ICU or hospital stay
CONCLUSION: HFOV provides no benefit in moderate-severe ARDS. Combined with OSCILLATE's harm signal, HFOV was demoted from routine to rescue.

[1]

ARMA trial (ARDSNet 2000, NEJM) — low vs traditional tidal volume

RCT: 861 patients with ALI/ARDS. Vt 6 mL/kg PBW, plateau ≤30 vs Vt 12 mL/kg, plateau ≤50.

Mortality: 31% (low Vt) vs 40% (traditional) — absolute risk reduction 9%, NNT 11
Stopped early for benefit
CONCLUSION: Low tidal volume lung-protective ventilation is the foundation of ARDS care — the single best-evidenced ventilator intervention, against which all advanced modes are mere adjuncts.[7] }

PROSEVA trial (Guérin 2013, NEJM) — prone positioning in severe ARDS

RCT: 466 patients with severe ARDS (PaO2/FiO2 <150, FiO2 ≥60%, PEEP ≥5). Prone ≥16 h/day vs supine.

28-day mortality: 16% (prone) vs 33% (supine) — NNT 6
90-day mortality: 24% vs 41%
CONCLUSION: Prolonged prone positioning reduces mortality dramatically in severe ARDS. Cheapest, most available and most effective intervention to try before any advanced mode.[8] }

EOLIA trial (Combes 2018, NEJM) — VV-ECMO in very severe ARDS

RCT: 249 patients with very severe ARDS (PaO2/FiO2 <50) — stopped early for futility at interim analysis.

60-day mortality: 35% (ECMO) vs 46% (conventional) — not statistically significant (RR 0.76, p=0.09)
High crossover (28% control patients received rescue ECMO)
Bayesian re-analysis suggested 77-96% probability of benefit
CONCLUSION: VV-ECMO is the salvage modality of choice for refractory hypoxaemia — preferred over HFOV.[9] }

ACURASYS (Papazian 2010) vs ROSE (PETAL 2019) — early cisatracurium in ARDS

ACURASYS: 340 patients, severe ARDS (PaO2/FiO2 <150). Cisatracurium 48 h reduced adjusted 90-day mortality (HR 0.44) and reduced barotrauma — benefit.
ROSE: 1006 patients, moderate-severe ARDS (PaO2/FiO2 <150, PEEP ≥8). Stopped early for futility — no mortality difference (42.5% vs 42.8%), more ICU-acquired weakness with placebo-receiver handling.
CONCLUSION: Routine 48-h cisatracurium is not standard. Short-course paralysis remains reasonable for dangerous transpulmonary pressures or severe dyssynchrony. HFOV requires paralysis — another mark against it.[10] }[11] }

Cochrane review (Sud 2016) — HFOV vs conventional ventilation in ARDS

9 RCTs, 1629 patients
HFOV may increase in-hospital mortality (RR 1.15) and provides no 30-day mortality benefit
No consistent improvement in barotrauma, oxygenation failure or ventilation duration
CONCLUSION: HFOV should not be used routinely; at best a rescue option when all proven therapies have failed.[12] }

Red flags — expanded

Don't-miss red flags in advanced ventilation

HFOV trend to HARM (OSCILLATE) and no benefit (OSCAR, Cochrane) — rescue only, behind proven therapy.[1] }[2] }[12] }
APRV gives NO proven mortality benefit — salvage, not a first-line lung-protective strategy.[3] }
APRV in a paralysed patient = inverse-ratio PCV with no advantage — don't use it if the patient cannot breathe.
APRV T-low too long → derecruitment; too short → auto-PEEP — titrate to 50-75% T-PEF, never a fixed clock time.[3] }
On HFOV, raising frequency raises PaCO2 — to ventilate, lower frequency and/or raise ΔP.[5] }
PCV silently drops Vt when compliance falls — always watch exhaled Vt and driving pressure.[6] }
PAV+ can RUNAWAY if gain approaches the patient's own elastance/resistance — cap gain ≤90% and confirm preserved drive.[14] }
NAVA catheter misplacement gives false EAdi — confirm position by waveform and X-ray; backup PSV covers apnoea.[13] }
Right heart failure on HFOV/APRV: high mean airway pressure raises PVR and RV afterload — check for acute cor pulmonale, especially in severe ARDS.[5] }
SIMV weaning prolongs ventilation — use PSV/T-piece for liberation.[6] }
Never abandon the proven bundle: low Vt (ARMA), plateau <30, driving pressure <15, prone ≥16 h (PROSEVA), VV-ECMO for refractory (EOLIA). Advanced modes are adjuncts, not substitutes.[7] }[8] }[9] }

Pitfalls — exam one-liners

Common pitfalls and one-line exam answers

"APRV is a recruitment mode." Yes — through near-continuous mean airway pressure (long T-high) with brief releases, allowing spontaneous breathing.
"Does APRV improve survival?" No proven mortality benefit; improved oxygenation in selected series. Salvage/adjunct only.
"Does HFOV improve survival?" No — OSCILLATE showed trend to HARM; OSCAR and Cochrane showed no benefit. Rescue only.
"HFOV frequency and CO2?" Inverse relationship — higher frequency → smaller Vt → higher CO2. Lower frequency or higher ΔP to ventilate.
"Which modes allow spontaneous breathing throughout?" APRV, BiLevel, PSV, NAVA, PAV+, ASV — NOT volume/pressure control with paralysis.
"What cycles PSV off?" Flow — at ~25% of peak inspiratory flow (patient-controlled).
"Best trigger in severe COPD?" NAVA (neural trigger) — immune to auto-PEEP and ineffective efforts.
"What guarantees Vt with decelerating flow?" PRVC/AutoFlow/VC+ (dual-control, volume-targeted).
"Order of escalating ARDS therapy?" Lung-protective ventilation (ARMA) → optimise PEEP → prone ≥16 h (PROSEVA) → short paralysis if dyssynchronous → VV-ECMO (EOLIA) → APRV/HFOV only as rescue where ECMO unavailable.

[1]

References

[1]Ferguson ND, Cook DJ, Guyatt GH, et al.; OSCILLATE Trial Investigators; Canadian Critical Care Trials Group. High-frequency oscillation in early acute respiratory distress syndrome N Engl J Med, 2013.PMID 23339639
[2]Young D, Lamb SE, Shah S, et al.; OSCAR Study Group. High-frequency oscillation for acute respiratory distress syndrome N Engl J Med, 2013.PMID 23339638
[3]Facchin F, Fan E. Airway Pressure Release Ventilation and High-Frequency Oscillatory Ventilation: Potential Strategies to Treat Severe Hypoxemia and Prevent Ventilator-Induced Lung Injury Respir Care, 2015.PMID 26405188
[4]Vaporidi K, Akoumianaki E, Telias I, et al. NAVA and PAV+ for lung and diaphragm protection Curr Opin Crit Care, 2020.PMID 31738231
[5]Fan E, Needham DM, Stewart TE. New modalities of mechanical ventilation: high-frequency oscillatory ventilation and airway pressure release ventilation Clin Chest Med, 2006.PMID 17085250
[6]Mireles-Cabodevila E, Hatipoglu U, Chatterjee A, et al. A rational framework for selecting modes of ventilation Respir Care, 2013.PMID 22710796
[7]The Acute Respiratory Distress Syndrome Network (ARDSNet). Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome N Engl J Med, 2000.PMID 10793162
[8]Guérin C, Reignier J, Richard JC, et al.; PROSEVA Trial Investigators. Prone positioning in severe acute respiratory distress syndrome N Engl J Med, 2013.PMID 23688302
[9]Combes A, Hajage D, Capellier G, et al.; EOLIA Trial Investigators. Extracorporeal Membrane Oxygenation for Severe Acute Respiratory Distress Syndrome N Engl J Med, 2018.PMID 29791822
[10]Papazian L, Forel JM, Gacouin A, et al.; ACURASYS Study Investigators. Neuromuscular blockers in early acute respiratory distress syndrome N Engl J Med, 2010.PMID 20843245
[11]National Heart, Lung, and Blood Institute PETAL Clinical Trials Network; Moss M, Huang DT, Brower RG, et al. Early Neuromuscular Blockade in the Acute Respiratory Distress Syndrome N Engl J Med, 2019.PMID 31112383
[12]Sud S, Friedrich JO, Taccone P, et al. High-frequency oscillatory ventilation versus conventional ventilation for acute respiratory distress syndrome Cochrane Database Syst Rev, 2016.PMID 27043185
[13]Sinderby C, Navalesi P, Beck J, et al. Neural control of mechanical ventilation in respiratory failure Nat Med, 1999.PMID 10581089
[14]Xirouchaki N, Kondili E, Vaporidi K, et al. Proportional assist ventilation with load-adjustable gain factors in critically ill patients: comparison with pressure support Intensive Care Med, 2008.PMID 18607562

[1]

APRV vs HFOV vs conventional ventilation

Feature	Conventional (VC/PC)	APRV	HFOV (Oscillator)
Mechanism	Volume or pressure controlled, cycled	Continuous high pressure + brief releases	Oscillating diaphragm at 3-15 Hz
Tidal volume	4-6 mL/kg (protective)	Variable (release-dependent)	1-3 mL/kg (tiny)
PEEP	Moderate (from PEEP/FiO2 table)	HIGH (P-high 25-35 cmH2O)	Mean airway pressure (high)
Spontaneous breathing	Possible (if mode allows)	YES (throughout — key feature)	NO (paralysis usually needed)
Sedation	Moderate-deep	Less (spontaneous)	Deep + paralysis
Oxygenation	Good	Often IMPROVED	Often improved
Mortality benefit	YES (low Vt — ARMA)	NO clear benefit	NO (OSCILLATE: trend to harm)
Current role	FIRST-LINE	Selected (salvage)	REFRACTORY only

[1]

When to consider advanced ventilation modes

Optimise conventional ventilation FIRST — low tidal volume (4-6 mL/kg), plateau <30, PEEP/FiO2 table, POCUS-guided PEEP
Try PRONE POSITIONING — PROSEVA: for severe ARDS (PaO2/FiO2 <150), >16h/day
Consider neuromuscular blockade — early severe ARDS (cisatracurium 48h — ACURASYS, but ROSE was negative)
If still refractory (PaO2/FiO2 <80):
- APRV: if patient can breathe spontaneously, may improve oxygenation and reduce sedation. Setting: P-high 25-30, P-low 0-5, T-high 4-6s, T-low 0.5-0.8s
- ECMO: for refractory hypoxaemia (VV-ECMO) — stronger evidence than HFOV
HFOV: LAST RESORT — only if APRV and ECMO unavailable/failed. Oscillator at mean Paw 25-30, frequency 4-6 Hz, delta pressure 60-90

[1]

Exam practice

SAQ — Choosing and setting APRV in refractory ARDS (vs HFOV)

10 minutes · 10 marks

[1]

SAQ — Liberation from APRV after recovery

10 minutes · 10 marks

Clinical pearls

High-yield advanced ventilation points for CICM/FFICM exam

HFOV does NOT improve mortality — OSCILLATE trial showed trend to HARM. Ferguson 2013 (NEJM): HFOV vs conventional in early moderate-severe ARDS. Trial STOPPED EARLY — HFOV had HIGHER mortality (47% vs 35%, p=0.005), more sedation, more vasopressors. HFOV now RESERVED for refractory cases (not first-line). OSCAR trial (UK): no mortality benefit.[1] }
APRV allows SPONTANEOUS breathing throughout the cycle. This is the KEY feature. Unlike conventional ventilation (where patient breathes between machine breaths), APRV maintains continuous positive pressure (P-high) with brief releases (to P-low). Patient can breathe ANY time (during P-high and P-low). Benefits: less sedation, preserved diaphragm function, improved venous return, less muscle atrophy.[3] }
APRV settings: P-high, P-low, T-high, T-low. P-high: 25-35 cmH2O (high continuous pressure — maintains recruitment). P-low: 0-5 cmH2O (release pressure — allows exhalation). T-high: 4-6 seconds (most of cycle at high pressure). T-low: 0.5-0.8 seconds (brief release — typically terminated at 50-75% of peak expiratory flow to avoid derecruitment). Typical I:E ratio: 8:1 to 10:1 (inverse).[5] }
APRV theoretical benefits (not proven). (1) Continuous recruitment (high PEEP throughout). (2) Less derecruitment (brief releases only). (3) Spontaneous breathing → improved V/Q (diaphragm contraction → dependent lung recruitment). (4) Less sedation needed (patient comfortable, can breathe). (5) Cardiac: spontaneous breathing → negative intrathoracic pressure → improved venous return → better cardiac output. BUT: no RCT shows mortality benefit.[3] }
Standard ARDS ventilation remains the FOUNDATION. Low tidal volume (4-6 mL/kg — ARMA trial), plateau pressure <30 cmH2O, PEEP from PEEP/FiO2 table, driving pressure <15. These have PROVEN mortality benefit. APRV and HFOV are ADJUNCTS for refractory cases — not replacements for lung-protective ventilation.[4] }
Prone positioning has BETTER evidence than APRV or HFOV. PROSEVA trial (2013): prone >16h/day for severe ARDS reduced mortality from 33% to 16% (NNT 6). This is PROVEN, SIMPLE, and AVAILABLE everywhere. Always try prone positioning BEFORE considering APRV or HFOV.[4] }
ECMO is the strongest salvage therapy for refractory ARDS. VV-ECMO: provides oxygenation while lungs rest. EOLIA trial (2018): trend toward benefit (not statistically significant, but crossed futility boundary early). CESAtrial (ongoing). ECMO is NOW preferred over HFOV for refractory ARDS (better evidence, lower complications).[4] }
HFOV physiology — very different from conventional. (1) Oscillating diaphragm creates sinusoidal pressure wave (positive AND negative). (2) Frequency: 3-15 Hz (180-900 breaths/min). (3) Tidal volume: ~1-3 mL/kg (MUCH smaller than conventional). (4) Gas exchange: by CONVECTION + DIFFUSION (not bulk flow). (5) Mean airway pressure: high (25-30 cmH2O) — maintains recruitment. (6) Theory: lung protection (tiny Vt → less volutrauma) + recruitment (high mean Paw).[6] }
Why HFOV failed (OSCILLATE trial). Possible explanations: (1) HFOV requires HEAVY SEDATION + PARALYSIS → complications (weakness, VAP, prolonged ventilation). (2) HIGH mean airway pressure → reduced venous return → haemodynamic instability (more vasopressors). (3) Theoretical lung protection NOT realised in practice. (4) May cause RIGHT HEART FAILURE (high intrathoracic pressure → high pulmonary vascular resistance). (5) Delay in starting conventional lung-protective ventilation.[1] }
APRV may be useful for DYSSYNCHRONY. If patient is fighting the ventilator (patient-ventilator dyssynchrony) on conventional modes → consider APRV (allows spontaneous breathing at any time). May reduce: sedation requirements, agitation, dyssynchrony. Useful in patients who WANT to breathe but conventional mode fights them.[5] }
Transitioning from conventional to APRV. (1) Start P-high at plateau pressure from conventional mode (e.g., if plateau was 28, start P-high 28-30). (2) P-low: 0 (allow full release). (3) T-high: 4-6 seconds. (4) T-low: 0.5-0.8 seconds (titrate to terminate at 50-75% of peak expiratory flow rate). (5) FiO2: same as conventional. (6) Monitor: oxygenation, CO2 (may increase initially), haemodynamics, patient comfort.[5] }
Transitioning from APRV back to conventional. (1) Gradually REDUCE P-high (by 2 cmH2O every 2-4h). (2) Gradually SHORTEN T-high (toward conventional I:E). (3) When P-high ~15-20 and T-high ~1-2s → switch to PSV/CPAP (weaning mode). (4) Extubate if tolerating spontaneous breathing trial.[5] }
Complications of APRV. (1) AUTO-PEEP/air trapping (if T-low too long — incomplete release). (2) DYSSYNCHRONY (if release timing doesn't match patient's breathing). (3) HYPERCAPNIA (if T-high too long — insufficient ventilation). (4) HAEMODYNAMIC instability (high intrathoracic pressure → reduced venous return — though spontaneous breathing mitigates this). (5) Requires EXPERIENCED operator (less familiar than conventional modes).[3] }
Both APRV and HFOV should be used in SPECIALIST centres. These advanced modes require: (1) Experienced clinicians (respiratory therapists, intensivists). (2) Specialised ventilators (especially oscillator — not available everywhere). (3) Close monitoring (blood gases, haemodynamics). (4) Protocols for transition to/from conventional. If no expertise: stick with conventional lung-protective ventilation + prone + ECMO.[4] }

Red flags

Prognosis

OSCILLATE trial (Ferguson 2013, NEJM) — HFOV in ARDS

RCT: 548 patients with moderate-severe ARDS (PaO2/FiO2 <200). HFOV vs conventional low-tidal-volume ventilation.

In-hospital mortality: HFOV 47% vs conventional 35% (RR 1.33, p=0.005) — HFOV WORSE
Trial STOPPED EARLY: for safety (HFOV clearly harmful)
Vasopressor use: MORE with HFOV
CONCLUSION: HFOV is HARMFUL in early ARDS. Should NOT be used routinely. Reserve for refractory cases only. [1]

[1]

APRV in depth — mechanics, settings and tuning

The four APRV variables

APRV settings — the four variables and how to set them

Variable	Typical range	What it controls	How to titrate
P-high (high pressure)	25-35 cmH2O	Mean airway pressure → recruitment & oxygenation	Start at plateau pressure from conventional mode (or 25-30). Limit by driving pressure / plateau <30 and haemodynamics
P-low (release pressure)	0-5 cmH2O (often 0, i.e. atmospheric)	How much the lung is allowed to derecruit on release	Usually set to 0 (full release). Raising P-low reduces Vt and CO2 clearance but may stabilise haemodynamics
T-high (release interval)	4-6 s (long)	Time at recruitment pressure; the longer T-high, the higher the mean Paw	Increase to improve oxygenation; reduce if hypercapnia/ventilation inadequate
T-low (release time)	0.2-0.8 s (brief)	Time allowed for exhalation — must be SHORT to avoid derecruitment	Titrate to terminate at 50-75% of peak expiratory flow (T-PEF), NOT a fixed time

[1]

The T-low concept — termination of peak expiratory flow (T-PEF)

Why spontaneous breathing matters

"True APRV" vs BiLevel / Biphasic / BiVent

Starting and titrating APRV — a practical protocol

Set P-high at the plateau pressure of the preceding conventional mode (or 25-30 cmH2O if de novo). Respect a driving pressure <15 cmH2O and avoid P-high that drops the SpO2-corrected mean arterial pressure.
Set P-low to 0 cmH2O (atmospheric) — full release maximises CO2 clearance. Add 3-5 cmH2O only if haemodynamically unstable or if release volume is excessive.
Set T-high to 4-6 s (start 4 s). Increase if oxygenation poor; this raises mean Paw.
Set T-low to 0.5-0.8 s, then titrate to terminate at 50-75% of peak expiratory flow on the flow waveform. Watch for auto-PEEP.
Set FiO2 the same as on conventional; reduce as oxygenation improves.
Monitor: arterial/venous blood gases at 30-60 min (expect a transient rise in PaCO2), mean arterial pressure, urine output, spontaneous effort (sedation to a Richmond score of −1 to −2 — the patient should be breathing).
Re-tune: failing oxygenation → ↑ P-high (by 2 cmH2O) or ↑ T-high; failing ventilation → ↑ release gradient (lower P-low or lengthen T-low cautiously) or accept permissive hypercapnia (keep pH >7.20).
Daily: lung-protective limits still apply (release Vt ~4-6 mL/kg PBW, plateau/driving pressure bounded); plan transition to conventional/PSV as lung injury resolves.

[1]

Weaning from APRV back to conventional ventilation

Lower P-high by 2 cmH2O every 2-4 h as oxygenation and compliance improve (target P-high ~20-22).
Shorten T-high toward a more conventional I:E ratio (e.g. 4 s → 2 s → 1 s) — the mode becomes BiLevel-like.
When P-high is ~15-20 cmH2O with a near-symmetric ratio, switch to pressure support (PSV/CPAP) with PS 8-12 cmH2O and PEEP 5-8.
Perform a spontaneous breathing trial; extubate if tolerated (rapid shallow breathing index <105, stable ABG, intact cough/mental status).
Avoid prolonged APRV in a recovering patient — the very feature that recruits in early ARDS (high mean pressure) can delay liberation once compliance improves.

[1]

HFOV in depth — mechanism, tuning and why the trials failed

How the oscillator works

HFOV variables

HFOV settings — what each knob does

Variable	Typical adult range	Effect of increasing it
Mean airway pressure (mPaw)	25-30 cmH2O (above the conventional PEEP)	↑ Oxygenation (recruitment) — but ↑ intrathoracic pressure → ↓ venous return, ↑ PVR
Frequency	3-6 Hz (range 3-15 Hz; higher for small lungs)	↑ Frequency → ↓ tidal volume → ↑ CO2 (counter-intuitive!)
ΔP (power/amplitude)	60-90 cmH2O	↑ ΔP → ↑ tidal volume → ↑ CO2 clearance (the "chest wiggle" should be visible to the iliac crests)
Inspiratory time %	~33%	Minor effect on Vt
Bias flow	20-40 L/min	Provides fresh gas; ↑ bias flow aids CO2 washout
FiO2	As required	Oxygenation

[1]

Gas exchange below dead space — the core paradox

Tuning logic — oxygenation vs ventilation are decoupled

On HFOV the two failure modes are managed with different knobs, unlike conventional ventilation:

Hypoxaemia → increase mPaw (recruitment) and FiO2.
Hypercapnia → decrease frequency (counter-intuitive — slower = bigger Vt) and/or increase ΔP (bigger swing = bigger Vt), or increase bias flow. [1]

Why OSCILLATE and OSCAR changed practice

Before 2013 HFOV was widely used in severe ARDS on the basis of physiological plausibility and oxygenation improvement. Two large RCTs published simultaneously in the NEJM in 2013 reversed this:

OSCILLATE (Ferguson): stopped early for harm — in-hospital mortality 47% vs 35% (RR 1.33), more vasopressors, more sedation.[1] }
OSCAR (Young): no mortality difference at 30 days.[2] }

Plausible mechanisms of harm

Conventional and dual-control modes — PCV, PSV, SIMV, PRVC

PCV vs PSV vs SIMV vs dual-control (PRVC/VC+) — the four mode families

Feature	PCV (pressure control)	PSV (pressure support)	SIMV (synchronised intermittent mandatory)	PRVC / VC+ (dual control, volume-targeted)
Trigger	Time (or patient)	Patient (flow/pressure)	Time for mandatory + patient for spontaneous	Time (or patient)
Limit	Set pressure (constant)	Set pressure support	Set pressure/volume for mandatory	Pressure, but auto-adjusted breath-to-breath
Cycle	Time (set I-time)	Flow (cycles at ~25% of peak flow)	Time (mandatory) / flow (spontaneous)	Time (decelerating flow to target volume)
Flow pattern	Decelerating	Decelerating	Variable	Decelerating
Vt delivered	Variable (depends on compliance/resistance)	Variable (patient + effort-dependent)	Variable for mandatory; variable for spontaneous	Guaranteed (auto-adjusts pressure to hit set Vt)
Key strength	Limits pressure; good leak tolerance; uniform gas distribution	Patient-ventilator synchrony; for weaning	Backup rate guaranteed	Volume guarantee with decelerating-flow benefit
Key risk	Hypoventilation if compliance falls (Vt drops silently)	Ineffective triggering; auto-cycling; apnoea if drive lost	Prolongs ventilation (superseded); stacking	Can still overdistend if set Vt too high
Typical use	ARDS (pressure-limited); severe dyssynchrony	Liberation/weaning; spontaneous breathing patients	Largely historical — avoid for weaning	Default volume-targeted mode on many ICU ventilators

[1]

Why SIMV has fallen out of favour

Dual-control / closed-loop volume targeting

Adaptive and closed-loop modes — NAVA and PAV+

NAVA vs PAV+ — patient-driven adaptive modes

Feature	NAVA (neurally adjusted ventilatory assist)	PAV+ (proportional assist ventilation, load-adjustable)
Signal that drives support	Electrical activity of the diaphragm (EAdi) via a specialised nasogastric catheter with electrodes at the oesophago-gastric junction	Patient's inspiratory flow and volume; resistance & elastance estimated intermittently by the ventilator
How support scales	Pressure = NAVA level (cmH2O/µV) × EAdi — linear with neural drive	Pressure = (gain × elastance × volume) + (gain × resistance × flow) — proportional to instantaneous effort
Trigger	Neural (EAdi) — bypasses pneumatic trigger	Flow/pressure (pneumatic)
Cycle-off	Neural (when EAdi falls to 70% of peak)	Flow/volume (patient effort)
Strengths	Eliminates ineffective triggering; works despite auto-PEEP (COPD); prevents over-assist; patient fully controls rate, Vt, Ti	Amplifies patient effort in proportion to demand; auto-tunes gain to compliance; lung- and diaphragm-protective
Limitations	Catheter placement; EAdi affected by oesophageal position, electrocautery, sedation; cost	Needs reliable effort; runaway risk if gain > ~90% of elastance; not for profoundly weak drive
Backup	PSV backup if apnoea / EAdi lost	Apnoea backup (PSV/PCV)
Best evidence	Improved synchrony, reduced over-assist; mortality benefit not proven	Improved synchrony; trials ongoing for earlier liberation

[1]

NAVA in detail

PAV+ in detail

Adaptive Support Ventilation (ASV) — closed-loop targeting

High-yield clinical pearls — advanced and adaptive modes

Advanced ventilation pearls for CICM/FFICM/EDIC — beyond the basics

Describe a mode by trigger–limit–cycle. PCV = time/pressure/time; PSV = patient/pressure/flow; PRVC = time/pressure(auto)/time-to-volume; NAVA = neural/effort-proportional/neural. If you can name all three phase variables you understand the mode.[6] }
APRV is two CPAP levels with a time-cycled release — NOT a mode that "gives a breath". There is no set tidal volume; the release generates the bulk of ventilation, augmented by spontaneous breaths. The mean airway pressure is dominated by the long T-high and is what maintains recruitment.[3] }
T-low is set by watching expiratory flow, not by a clock. The release should end at 50-75% of peak expiratory flow (T-PEF). End too late → derecruitment; end too early → air-trapping/auto-PEEP. This single waveform skill distinguishes a competent APRV operator.[3] }
On HFOV, raising the frequency REDUCES ventilation and raises CO2 — the most counter-intuitive fact in advanced ventilation. To clear CO2 you lower frequency and/or raise ΔP (power). To improve oxygenation you raise mean airway pressure.[5] }
HFOV tidal volumes (1-3 mL/kg) are below anatomical dead space. Gas exchange therefore occurs by convective dispersion, pendelluft and molecular diffusion — NOT bulk flow. This is why HFOV can ventilate at all despite "impossibly small" Vt.[5] }
HFOV actively expires. Unlike high-frequency jet or percussive ventilation (passive expiration), the oscillator's diaphragm pulls back on expiration — this is what prevents gas trapping and is the defining mechanical feature of "oscillation".[5] }
The OSCILLATE harm signal is the single most testable fact in this topic. Stopped early; in-hospital mortality 47% vs 35% (RR 1.33); more vasopressors and sedation. HFOV is therefore rescue only, behind lung-protective ventilation, prone and ECMO.[1] }
PSV is flow-cycled — it ends when inspiratory flow falls to ~25% of peak. This means the patient, not the machine, ends the breath, which is why PSV is the weaning mode. Trouble: in COPD the high internal resistance delays flow decay → prolonged inspiration and dynamic hyperinflation; cycle the ventilator earlier (raise flow cycle-off %) or add expiratory trigger adjustment.[6] }
PCV silently drops tidal volume if compliance falls. Because pressure is fixed, a worsening lung (oedema, pleural effusion, mainstem intubation) reduces Vt without any alarm until minute ventilation collapses. Always watch exhaled Vt and driving pressure on PCV.[6] }
PRVC/AutoFlow guarantees volume by auto-adjusting pressure breath-to-breath. It is the modern default volume-targeted mode — but it does NOT prevent overdistension. You must still cap Vt at 4-6 mL/kg PBW and watch driving pressure <15.[6] }
SIMV prolongs ventilation — do not wean with it. Brochard (1994) and Esteban (1995) both showed SIMV-based weaning takes longer than PSV or T-piece. SIMV is now historical for liberation; retain it only for backup rate.[6] }
NAVA triggers off the diaphragm (EAdi) and is immune to auto-PEEP. This makes it the synchrony mode of choice in severe COPD/asthma where ineffective triggering dominates. Backup PSV covers apnoea. Gain is set as a NAVA level (cmH2O/µV).[13] }
PAV+ amplifies patient effort and can RUNAWAY. If the gain approaches the patient's own elastance/resistance (>~90%), the loop self-excites → huge Vt. Modern PAV+ caps the gain and auto-measures resistance/elastance (the "load-adjustable gain factors" of Xirouchaki 2008).[14] }
The common thread of NAVA and PAV+ is "support proportional to demand". Both aim to avoid over-assist (which causes diaphragm disuse atrophy) and under-assist (high work of breathing). Neither has proven a mortality benefit — improved synchrony and lung/diaphragm protection are the demonstrated gains.[4] }
Never forget the proven bundle before reaching for advanced modes. Low tidal volume (4-6 mL/kg, ARMA), plateau <30 cmH2O, driving pressure <15, prone ≥16 h/day for PaO2/FiO2 <150 (PROSEVA), and VV-ECMO for refractory hypoxaemia (EOLIA) all have better evidence than APRV, HFOV, NAVA or PAV+.[7] }
Prone positioning (PROSEVA) reduces mortality from ~33% to ~16% in severe ARDS — NNT ~6. It is cheap, available everywhere and proven; always deploy it before APRV/HFOV.[8] }
ECMO is now preferred over HFOV for refractory ARDS. EOLIA (2018) showed a strong trend to benefit (crossed the futility boundary early); a Bayesian re-analysis suggested probable benefit. ECMO rests the lung without the haemodynamic cost of high mean airway pressure.[9] }
Cisatracurium for 48 h (ACURASYS) reduced mortality in early severe ARDS — but ROSE (2019) did NOT confirm it. Current practice: short-course paralysis is reasonable for severe dyssynchrony/dangerous transpulmonary pressures, but routine 48-h paralysis is not standard.[10] }
The diaphragm is now a target organ. Both over-assist (disuse atrophy) and under-assist (load-induced injury) cause diaphragm dysfunction and prolong ventilation. This is the rationale for NAVA/PAV+ and for monitoring EAdi and respiratory effort even on conventional modes.[4] }
APRV is pointless in a paralysed patient. Its whole advantage is preserved spontaneous breathing. If you must paralyse (e.g. for severe dyssynchrony on HFOV), APRV collapses into an inverse-ratio pressure-controlled mode with no proven benefit.[3] }
Mean airway pressure, not PEEP, is the recruitment variable on APRV and HFOV. On APRV raise it by lengthening T-high or raising P-high; on HFOV by raising mPaw directly. Watch the haemodynamic and RV cost.[5] }
All advanced modes belong in specialist centres with protocols for transition. The evidence for any of them is weaker than the proven bundle; the risk of harm (auto-PEEP, RV failure, runaway, prolonged ventilation) is real. If the expertise is absent, stick with conventional lung-protective ventilation + prone + ECMO referral.[5] }

Additional trials — the evidence base

OSCAR trial (Young 2013, NEJM) — HFOV in UK ARDS

RCT: 795 patients with ARDS (PaO2/FiO2 <200) across 37 UK ICUs. HFOV vs conventional ventilation.

30-day mortality: HFOV 41.3% vs conventional 41.6% — no significant difference
No difference in duration of ventilation, ICU or hospital stay
CONCLUSION: HFOV provides no benefit in moderate-severe ARDS. Combined with OSCILLATE's harm signal, HFOV was demoted from routine to rescue.

[1]

ARMA trial (ARDSNet 2000, NEJM) — low vs traditional tidal volume

RCT: 861 patients with ALI/ARDS. Vt 6 mL/kg PBW, plateau ≤30 vs Vt 12 mL/kg, plateau ≤50.

Mortality: 31% (low Vt) vs 40% (traditional) — absolute risk reduction 9%, NNT 11
Stopped early for benefit
CONCLUSION: Low tidal volume lung-protective ventilation is the foundation of ARDS care — the single best-evidenced ventilator intervention, against which all advanced modes are mere adjuncts.[7] }

PROSEVA trial (Guérin 2013, NEJM) — prone positioning in severe ARDS

RCT: 466 patients with severe ARDS (PaO2/FiO2 <150, FiO2 ≥60%, PEEP ≥5). Prone ≥16 h/day vs supine.

28-day mortality: 16% (prone) vs 33% (supine) — NNT 6
90-day mortality: 24% vs 41%
CONCLUSION: Prolonged prone positioning reduces mortality dramatically in severe ARDS. Cheapest, most available and most effective intervention to try before any advanced mode.[8] }

EOLIA trial (Combes 2018, NEJM) — VV-ECMO in very severe ARDS

RCT: 249 patients with very severe ARDS (PaO2/FiO2 <50) — stopped early for futility at interim analysis.

60-day mortality: 35% (ECMO) vs 46% (conventional) — not statistically significant (RR 0.76, p=0.09)
High crossover (28% control patients received rescue ECMO)
Bayesian re-analysis suggested 77-96% probability of benefit
CONCLUSION: VV-ECMO is the salvage modality of choice for refractory hypoxaemia — preferred over HFOV.[9] }

ACURASYS (Papazian 2010) vs ROSE (PETAL 2019) — early cisatracurium in ARDS

ACURASYS: 340 patients, severe ARDS (PaO2/FiO2 <150). Cisatracurium 48 h reduced adjusted 90-day mortality (HR 0.44) and reduced barotrauma — benefit.
ROSE: 1006 patients, moderate-severe ARDS (PaO2/FiO2 <150, PEEP ≥8). Stopped early for futility — no mortality difference (42.5% vs 42.8%), more ICU-acquired weakness with placebo-receiver handling.
CONCLUSION: Routine 48-h cisatracurium is not standard. Short-course paralysis remains reasonable for dangerous transpulmonary pressures or severe dyssynchrony. HFOV requires paralysis — another mark against it.[10] }[11] }

Cochrane review (Sud 2016) — HFOV vs conventional ventilation in ARDS

9 RCTs, 1629 patients
HFOV may increase in-hospital mortality (RR 1.15) and provides no 30-day mortality benefit
No consistent improvement in barotrauma, oxygenation failure or ventilation duration
CONCLUSION: HFOV should not be used routinely; at best a rescue option when all proven therapies have failed.[12] }

Red flags — expanded

Don't-miss red flags in advanced ventilation

HFOV trend to HARM (OSCILLATE) and no benefit (OSCAR, Cochrane) — rescue only, behind proven therapy.[1] }[2] }[12] }
APRV gives NO proven mortality benefit — salvage, not a first-line lung-protective strategy.[3] }
APRV in a paralysed patient = inverse-ratio PCV with no advantage — don't use it if the patient cannot breathe.
APRV T-low too long → derecruitment; too short → auto-PEEP — titrate to 50-75% T-PEF, never a fixed clock time.[3] }
On HFOV, raising frequency raises PaCO2 — to ventilate, lower frequency and/or raise ΔP.[5] }
PCV silently drops Vt when compliance falls — always watch exhaled Vt and driving pressure.[6] }
PAV+ can RUNAWAY if gain approaches the patient's own elastance/resistance — cap gain ≤90% and confirm preserved drive.[14] }
NAVA catheter misplacement gives false EAdi — confirm position by waveform and X-ray; backup PSV covers apnoea.[13] }
Right heart failure on HFOV/APRV: high mean airway pressure raises PVR and RV afterload — check for acute cor pulmonale, especially in severe ARDS.[5] }
SIMV weaning prolongs ventilation — use PSV/T-piece for liberation.[6] }
Never abandon the proven bundle: low Vt (ARMA), plateau <30, driving pressure <15, prone ≥16 h (PROSEVA), VV-ECMO for refractory (EOLIA). Advanced modes are adjuncts, not substitutes.[7] }[8] }[9] }

Pitfalls — exam one-liners

Common pitfalls and one-line exam answers

"APRV is a recruitment mode." Yes — through near-continuous mean airway pressure (long T-high) with brief releases, allowing spontaneous breathing.
"Does APRV improve survival?" No proven mortality benefit; improved oxygenation in selected series. Salvage/adjunct only.
"Does HFOV improve survival?" No — OSCILLATE showed trend to HARM; OSCAR and Cochrane showed no benefit. Rescue only.
"HFOV frequency and CO2?" Inverse relationship — higher frequency → smaller Vt → higher CO2. Lower frequency or higher ΔP to ventilate.
"Which modes allow spontaneous breathing throughout?" APRV, BiLevel, PSV, NAVA, PAV+, ASV — NOT volume/pressure control with paralysis.
"What cycles PSV off?" Flow — at ~25% of peak inspiratory flow (patient-controlled).
"Best trigger in severe COPD?" NAVA (neural trigger) — immune to auto-PEEP and ineffective efforts.
"What guarantees Vt with decelerating flow?" PRVC/AutoFlow/VC+ (dual-control, volume-targeted).
"Order of escalating ARDS therapy?" Lung-protective ventilation (ARMA) → optimise PEEP → prone ≥16 h (PROSEVA) → short paralysis if dyssynchronous → VV-ECMO (EOLIA) → APRV/HFOV only as rescue where ECMO unavailable.

[1]

References

[1]Ferguson ND, Cook DJ, Guyatt GH, et al.; OSCILLATE Trial Investigators; Canadian Critical Care Trials Group. High-frequency oscillation in early acute respiratory distress syndrome N Engl J Med, 2013.PMID 23339639
[2]Young D, Lamb SE, Shah S, et al.; OSCAR Study Group. High-frequency oscillation for acute respiratory distress syndrome N Engl J Med, 2013.PMID 23339638
[3]Facchin F, Fan E. Airway Pressure Release Ventilation and High-Frequency Oscillatory Ventilation: Potential Strategies to Treat Severe Hypoxemia and Prevent Ventilator-Induced Lung Injury Respir Care, 2015.PMID 26405188
[4]Vaporidi K, Akoumianaki E, Telias I, et al. NAVA and PAV+ for lung and diaphragm protection Curr Opin Crit Care, 2020.PMID 31738231
[5]Fan E, Needham DM, Stewart TE. New modalities of mechanical ventilation: high-frequency oscillatory ventilation and airway pressure release ventilation Clin Chest Med, 2006.PMID 17085250
[6]Mireles-Cabodevila E, Hatipoglu U, Chatterjee A, et al. A rational framework for selecting modes of ventilation Respir Care, 2013.PMID 22710796
[7]The Acute Respiratory Distress Syndrome Network (ARDSNet). Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome N Engl J Med, 2000.PMID 10793162
[8]Guérin C, Reignier J, Richard JC, et al.; PROSEVA Trial Investigators. Prone positioning in severe acute respiratory distress syndrome N Engl J Med, 2013.PMID 23688302
[9]Combes A, Hajage D, Capellier G, et al.; EOLIA Trial Investigators. Extracorporeal Membrane Oxygenation for Severe Acute Respiratory Distress Syndrome N Engl J Med, 2018.PMID 29791822
[10]Papazian L, Forel JM, Gacouin A, et al.; ACURASYS Study Investigators. Neuromuscular blockers in early acute respiratory distress syndrome N Engl J Med, 2010.PMID 20843245
[11]National Heart, Lung, and Blood Institute PETAL Clinical Trials Network; Moss M, Huang DT, Brower RG, et al. Early Neuromuscular Blockade in the Acute Respiratory Distress Syndrome N Engl J Med, 2019.PMID 31112383
[12]Sud S, Friedrich JO, Taccone P, et al. High-frequency oscillatory ventilation versus conventional ventilation for acute respiratory distress syndrome Cochrane Database Syst Rev, 2016.PMID 27043185
[13]Sinderby C, Navalesi P, Beck J, et al. Neural control of mechanical ventilation in respiratory failure Nat Med, 1999.PMID 10581089
[14]Xirouchaki N, Kondili E, Vaporidi K, et al. Proportional assist ventilation with load-adjustable gain factors in critically ill patients: comparison with pressure support Intensive Care Med, 2008.PMID 18607562