Mechanical Ventilation: Modes, Waveforms & Lung Protection

One question the ventilator asks

Mode names are marketing. "AC/VC," "SIMV/PC," "PRVC," "VC+," "Autoflow," "APV" — different manufacturers give different names to nearly identical behavior, and the names have caused more confusion in ICU education than almost anything else.

Strip the names away and there is exactly one question underneath.

Volume control: you set the volume, the pressure floats

You order a tidal volume; the ventilator delivers exactly that volume every breath, at a set flow, whatever pressure it takes.

• Guaranteed: minute ventilation, and therefore CO₂ clearance.
• Floats: airway pressure. If the lungs get stiffer (pulmonary edema, ARDS progression, abdominal distension) or the airways narrow (bronchospasm, secretions, a kinked tube), the pressure rises to push the same volume in.
• The risk: barotrauma. The ventilator will keep pushing until it hits the high-pressure alarm.
• What you watch: peak and plateau pressures. A rising pressure in volume control is your early-warning system.

Pressure control: you set the pressure, the volume floats

You order an inspiratory pressure; the ventilator applies exactly that pressure for a set inspiratory time, and whatever volume goes in, goes in.

• Guaranteed: peak alveolar pressure — a built-in ceiling on barotrauma.
• Floats: tidal volume, and therefore minute ventilation. If compliance worsens, the volume falls.
• The risk: hypoventilation. A patient whose lungs stiffen can silently drop their tidal volume and retain CO₂ with no pressure alarm at all.
• What you watch: exhaled tidal volume and minute ventilation. A falling volume in pressure control is your early-warning system.

The comparison

Dual-control: having it both ways, imperfectly

PRVC, VC+, Autoflow, APV, AutoFlow — all of these are pressure-controlled breaths whose pressure the ventilator adjusts, breath to breath, to hit a target volume. The waveform looks like pressure control (decelerating flow); the guarantee looks like volume control (a volume target).

The catch is worth knowing. If the patient's own effort increases — because they are agitated, air-hungry, or in metabolic acidosis — they pull in a bigger volume, the ventilator sees the target exceeded and lowers the pressure support on the next breath. The patient works harder; the ventilator helps less. In a distressed patient this can feed a cycle of increasing work of breathing that looks, on the monitor, like the ventilator is doing fine.

Trigger, target, cycle: the anatomy of a breath

Every breath, in every mode, on every ventilator ever made, has four phases and three decisions. Once you can name them, unfamiliar modes stop being mysterious.¹

The three decisions

Describe any mode with those three words and you have described it completely:

• Volume assist-control = patient- or time-triggered, volume-targeted, volume-cycled.
• Pressure control = patient- or time-triggered, pressure-targeted, time-cycled.
• Pressure support = patient-triggered only, pressure-targeted, flow-cycled.

That last one is the important distinction. Pressure control and pressure support look nearly identical on a pressure–time scalar. The difference is the cycle variable: pressure control ends inspiration when the clock says so; pressure support ends when the patient's inspiratory flow falls to a set fraction of its peak — typically 25%. Pressure support lets the patient decide how long the breath lasts. Pressure control does not. That single difference is why pressure support is a weaning mode and pressure control is not.

Trigger, in detail

• Time trigger. The set rate fires the breath. A fully passive patient gets only time-triggered breaths.
• Flow trigger. The ventilator runs a small bias flow through the circuit and watches for the patient to divert some of it. Typically set at 1–3 L/min. Generally more sensitive and lower-work than pressure triggering.
• Pressure trigger. The patient must generate a negative pressure — typically −0.5 to −2 cmH₂O — to open the valve.

Cycle, in detail

• Volume cycled: ends when the set volume has been delivered.
• Time cycled: ends when the set inspiratory time expires.
• Flow cycled: ends when inspiratory flow decays to a set percentage of peak (the "expiratory sensitivity" or "cycle-off" setting, usually 25%).

The flow-cycle threshold is adjustable and clinically meaningful. In a COPD patient with high resistance, flow decays slowly, so inspiration drags on past the patient's own neural inspiratory time — the patient starts exhaling against a still-inflating ventilator. Raising the cycle-off to 40–50% shortens inspiration and fixes it. In restrictive disease flow decays fast and inspiration may end prematurely; lowering the threshold lengthens it.

Baseline: PEEP

The fourth phase is exhalation, and its variable is PEEP — positive end-expiratory pressure, the pressure the ventilator maintains at end-exhalation.

PEEP does four things: it holds recruitable alveoli open, preventing the repeated open–close cycle that causes atelectrauma; it improves oxygenation by increasing functional residual capacity and reducing shunt; it redistributes lung water in pulmonary edema; and it counterbalances intrinsic PEEP in obstructive disease, reducing trigger work.

It also has costs: it raises intrathoracic pressure and can reduce venous return and cardiac output, it can overdistend already-open alveoli, and in unilateral disease it can divert blood flow toward the diseased lung and worsen shunt.

The modes, decoded

Assist-control (AC) — volume or pressure

Every breath is a full ventilator breath. The set rate is a floor, not a ceiling: if the patient triggers above it, they get a fully supported breath at the set volume or pressure, not a partial one.

• AC/VC (volume assist-control): every breath delivers the set tidal volume. The most common initial mode in the ICU and the mode of the ARDSNet protocol.
• AC/PC (pressure assist-control): every breath delivers the set inspiratory pressure for the set inspiratory time.

The classic failure mode: a tachypneic patient in AC/VC receives a full tidal volume with every triggered breath. At a rate of 35 with a 450 mL tidal volume, that is 15.75 L/min of minute ventilation — profound respiratory alkalosis, and, in an obstructive patient, no time to exhale, producing auto-PEEP. Recognizing "the patient is over-breathing the vent in AC" and understanding why it matters is high-yield.

SIMV — synchronized intermittent mandatory ventilation

A set number of mandatory breaths (volume or pressure), synchronized to the patient's effort. Breaths taken above that rate are the patient's own spontaneous breaths, supported only by whatever pressure support has been added.

The result is a mixed pattern: some fully supported breaths and some minimally supported ones. Historically SIMV was used to wean by decreasing the mandatory rate — an approach that has been shown to prolong weaning compared with daily spontaneous breathing trials or pressure support weaning, largely because the unsupported breaths impose real work of breathing.

Pressure support ventilation (PSV) — a spontaneous mode

Every breath is patient-triggered and flow-cycled. There is no set rate — if the patient stops breathing, they get nothing but the apnea backup. This is the primary weaning mode and the usual mode for a spontaneous breathing trial.

• PS level determines the size of the breath.
• PEEP is set independently.
• Cycle-off % determines when inspiration ends and is adjustable, as above.

PRVC / VC+ / Autoflow / APV — dual control

Pressure-controlled breaths with a volume target, adjusted breath to breath. Covered in the previous section — remember the feedback loop that reduces support as patient effort rises.

APRV — airway pressure release ventilation

Detailed in the advanced modes section. In brief: a high pressure held for a long time (P-high, T-high), with brief releases to a low pressure (P-low, T-low) for CO₂ clearance, and spontaneous breathing permitted throughout. Conceptually it is inverse-ratio pressure control that allows spontaneous effort at the high pressure.

Backup modes and safety

• Apnea ventilation: the ventilator reverts to a mandatory mode if no breath is detected within a set interval. Verify these settings are appropriate on every spontaneous-mode patient — this is a real safety check, not a formality.
• High-pressure limit: in volume modes, the pressure at which the ventilator terminates inspiration and dumps the breath. If a patient's peak pressure hits this limit, they are not receiving the set tidal volume.
• Low minute ventilation alarm: the single most important alarm in a pressure-controlled or spontaneous patient.

Choosing initial settings — a defensible starting point

For a typical adult without severe obstructive disease:^2,3

Reading the scalars

Four curves tell you nearly everything: pressure–time, flow–time, volume–time, and the pressure–volume loop. Two of them do most of the work.

Pressure–time: where the pressure went

In a volume-controlled breath with square flow, the pressure curve has a characteristic shape:

• A rapid initial rise as flow starts against airway resistance.
• A slower linear climb as volume accumulates against compliance.
• A sharp corner at peak inspiratory pressure (PIP) — the highest pressure in the circuit.
• If you perform an inspiratory hold, flow stops and the pressure falls to a plateau: plateau pressure (Pplat) — the pressure distending the alveoli once no gas is moving.

Pplat is only valid in a passive patient during an end-inspiratory hold of about 0.5 seconds. A patient who is actively breathing contaminates the measurement.

In pressure control, the pressure curve is a square-topped plateau by design — you set it. There is no PIP-versus-Pplat comparison available without a hold, which is one reason volume control is preferred when you want to track mechanics closely.

Flow–time: the curve that finds air trapping

The inspiratory limb is above the line, the expiratory limb below.

• Square inspiratory flow = volume control at a set constant flow.
• Decelerating inspiratory flow = pressure control, pressure support, or dual control.
• The expiratory limb should always return to the zero line before the next breath begins.

Volume–time: leaks and trapping

Inspired and expired volumes should be equal. A curve that fails to return to baseline means volume is going somewhere: a cuff leak, a bronchopleural fistula, a chest tube air leak, or a circuit disconnection. A curve that returns but progressively creeps upward at end-expiration over successive breaths is air trapping.

The pressure–volume loop

Plotted with volume on the vertical axis and pressure on the horizontal, one breath makes a loop.

• Lower inflection point — the pressure at which collapsed alveoli begin to open; historically used to set PEEP just above it.
• Upper inflection point / "beaking" — where the loop flattens at the top: additional pressure produces little additional volume. That is overdistension, and it is a signal to reduce tidal volume or plateau pressure.
• Slope of the loop = compliance. A loop that rotates clockwise (flatter) over hours means the lung is getting stiffer.
• A wide loop (large area between the limbs) means high resistance — hysteresis, the energy lost overcoming resistance.

Flow–volume loop

Mostly used to spot obstruction and secretions. A sawtooth pattern on the expiratory limb is the classic sign of secretions in the large airways — and it is a genuinely useful, evidence-based indication to suction, rather than suctioning on a schedule.

The numbers: compliance, resistance, driving pressure, mechanical power

Five numbers, each computed from things already on your ventilator screen.

Static compliance

C_stat = V_T ÷ (Pplat − PEEP)

Normal in a ventilated adult is roughly 50–100 mL/cmH₂O; in ARDS it commonly falls to 20–40. It answers: how much volume do I get per unit of distending pressure? A falling compliance is a lung that is getting stiffer — worsening edema, consolidation, atelectasis, or extrapulmonary restriction from a distended abdomen.

Airway resistance

R_aw = (PIP − Pplat) ÷ Flow (flow in L/sec)

Normal is roughly 4–8 cmH₂O/L/sec through an endotracheal tube. Rising resistance means bronchospasm, secretions, a partially obstructed or too-small tube, or a patient biting.

Driving pressure — the number that changed practice

ΔP = Pplat − PEEP, which is also V_T ÷ C_stat

Driving pressure is the pressure swing actually applied to the aerated lung. Its importance is that it normalizes tidal volume to the size of the functional lung rather than to the patient's height — and in ARDS, where much of the lung is consolidated, the "baby lung" that is still open may be a fraction of the anatomic lung.

Mechanical power — energy, not pressure

Driving pressure is a pressure. But injury is done by energy delivered over time, and the same driving pressure at a rate of 35 delivers far more energy than at a rate of 12. Mechanical power unifies tidal volume, driving pressure, flow, respiratory rate and PEEP into a single number in joules per minute.⁵

A commonly used simplified form for volume-controlled ventilation:

Power (J/min) ≈ 0.098 × RR × V_T × (P_peak − ½ × ΔP)

with V_T in liters and pressures in cmH₂O.

Values above roughly 17–22 J/min have been associated with lung injury and worse outcomes across several studies; there is no single validated threshold, and the exact number depends on the equation used. Its practical value at the bedside is comparative: it makes visible that increasing the respiratory rate to improve CO₂ clearance is not free, and it shows you which of several possible changes reduces the energy load most.

Oxygenation indices

• P/F ratio = PaO₂ ÷ FiO₂ (as a decimal). Mild ARDS 200–300, moderate 100–200, severe ≤100 — at PEEP ≥5.
• SpO₂/FiO₂ (S/F) ratio — the new global definition of ARDS (2024) permits an S/F ratio ≤315 (with SpO₂ ≤97%) in place of a P/F ratio, and extends the definition to patients on high-flow nasal oxygen at ≥30 L/min. This matters at the bedside because it means ARDS can now be diagnosed without an arterial line and without intubation.⁶
• Oxygenation index = (FiO₂ × mean airway pressure × 100) ÷ PaO₂ — used mostly in pediatrics and ECMO candidacy.

Putting them together: reading a deterioration

A patient on AC/VC 6 mL/kg, PEEP 10:

Lung protection and VILI

Mechanical ventilation is a life-saving therapy that injures lungs. Four mechanisms, and they overlap.

The four mechanisms

Volutrauma — overdistension injury. The primary insult is excessive volume, not pressure per se: alveolar stretch beyond physiologic limits disrupts the epithelial–endothelial barrier, causing edema and hyaline membranes. Animal work showing that chest strapping to limit volume prevents injury even at high pressure is the reason we call it volutrauma rather than barotrauma.

Barotrauma — the macroscopic version: air escaping into places it should not be. Pneumothorax, pneumomediastinum, subcutaneous emphysema, pneumoperitoneum.

Atelectrauma — shear injury from alveoli repeatedly opening and collapsing with each breath. The tissue stress at the interface between open and collapsed lung is enormous. This is what adequate PEEP prevents.

Biotrauma — the systemic consequence. Mechanical stress triggers alveolar release of IL-6, IL-8 and TNF-α; injurious ventilation increases circulating cytokines and contributes to distant organ failure. It is one reason ARDS patients die of multi-organ failure rather than hypoxemia.

The protective bundle

1. Low tidal volume. ARMA (2000) randomized 861 patients to 6 versus 12 mL/kg PBW and found mortality of 31.0% versus 39.8% — an absolute reduction of nearly 9%, one of the largest effects in critical care.² 6 mL/kg PBW is the standard, reducible to 4 mL/kg if plateau pressure demands it.

2. Plateau pressure ≤30 cmH₂O. The ARDSNet ceiling.

3. Driving pressure ≤15 cmH₂O. The modern refinement.⁴

4. Adequate PEEP. The ESICM 2023 guidelines suggest higher PEEP over lower PEEP in moderate–severe ARDS, but against routine aggressive recruitment maneuvers — the ART trial found higher mortality with an aggressive recruitment-plus-titrated-PEEP strategy.^3,8

5. Permissive hypercapnia. Accepting a pH down to roughly 7.20 (some tolerate lower) is preferable to injurious volumes. Not appropriate with raised intracranial pressure or severe pulmonary hypertension.

6. Prone positioning. PROSEVA randomized 466 patients with severe ARDS (P/F <150) to prone sessions of at least 16 consecutive hours; 28-day mortality was 16.0% versus 32.8% — a halving.⁹ Prone positioning is strongly recommended for moderate–severe ARDS.³ It works by recruiting dorsal lung, making transpulmonary pressure more uniform, improving V/Q matching, and relieving the weight of the heart on the left lower lobe.

7. Neuromuscular blockade — selectively. ACURASYS (2010) suggested a mortality benefit from 48 hours of cisatracurium in early severe ARDS;¹⁰ ROSE (2019), conducted with lighter sedation and a higher-PEEP strategy, found no benefit.¹¹ Current practice reserves it for severe ARDS with refractory hypoxemia, severe dyssynchrony, or during proning — not routinely.

8. ECMO for refractory cases. EOLIA did not meet its primary endpoint but showed a strong trend and very high crossover; combined with subsequent analyses, VV-ECMO is offered at experienced centers for the most severe refractory hypoxemia.¹²

The rescue ladder for refractory hypoxemia

Roughly in order:

1. Optimize the basics — sedation and synchrony, secretions, tube position, confirm PEEP and FiO₂ are what you think they are
2. PEEP titration
3. Prone positioning ≥16 hours
4. Neuromuscular blockade
5. Inhaled pulmonary vasodilators (nitric oxide, epoprostenol) — improve oxygenation but have not shown a mortality benefit
6. VV-ECMO referral

Patient–ventilator asynchrony

Asynchrony is common — present in a meaningful fraction of breaths in many ventilated patients — and a high asynchrony index has been associated with longer ventilation and higher mortality.¹³ It is also, unlike much of critical care, something you can often fix from the bedside by changing a setting rather than adding a drug.

The reflex response — "the patient is bucking the vent, give more sedation" — treats the alarm rather than the cause, adds delirium and ventilator days, and frequently doesn't work because the asynchrony is mechanical.

Flow starvation (flow asynchrony)

What it looks like: in a volume-controlled breath, the pressure–time curve is scooped or concave rather than convex — the patient is pulling harder than the set flow delivers, dragging the pressure down mid-inspiration.

Why: set inspiratory flow is lower than the patient's demand.

Fixes: increase inspiratory flow rate; shorten inspiratory time; switch to a decelerating flow pattern or to a pressure-targeted mode where flow is delivered on demand; and treat the drive itself (pain, anxiety, acidosis, hypoxemia, fever).

Double triggering (breath stacking)

What it looks like: two ventilator breaths back to back with little or no exhalation between them, giving a stacked tidal volume that can be twice the set volume.

Why: the patient's neural inspiratory time is longer than the ventilator's set inspiratory time. Inspiration ends while the patient is still inhaling; that continued effort triggers a second breath immediately.

Why it matters: the delivered volume is double what is charted. This is a direct volutrauma mechanism, and it is invisible unless you look at the waveform or the exhaled volume.

Fixes: lengthen inspiratory time; increase tidal volume slightly (within protective limits); reduce flow; sometimes switch to pressure control where inspiratory time is fixed; address the high drive.

Reverse triggering

What it looks like: a ventilator-initiated breath is followed by a patient effort that appears entrained to it — an effort locked in rhythm with the machine, sometimes producing a second stacked breath.

Why: in deeply sedated patients, the mechanical breath itself stimulates a diaphragmatic contraction via a respiratory-entrainment reflex. Notably, it happens because of sedation depth, not despite it — so deepening sedation is not reliably the answer.

Why it matters: it can cause breath stacking and generates unrecognized transpulmonary pressure swings. It is under-recognized and easy to mistake for the patient "waking up."¹⁴

Ineffective triggering (missed efforts)

What it looks like: a visible dip in the pressure curve and/or a notch in the expiratory flow curve, with no delivered breath. The patient's rate on the monitor is lower than the rate you can see them breathing at the bedside.

Why: usually auto-PEEP — the effort is spent overcoming trapped pressure. Also trigger sensitivity set too low, or diaphragm weakness.

Fixes: treat the auto-PEEP (longer expiratory time, bronchodilators, suction, lower rate), add external PEEP up to ~80% of measured intrinsic PEEP, increase trigger sensitivity, reduce sedation if weakness from oversedation is contributing.

Auto-triggering

What it looks like: delivered breaths with no patient effort — a rate on the ventilator higher than the patient is actually breathing.

Why: trigger set too sensitive; condensation in the circuit; a circuit or cuff leak; cardiac oscillations transmitted to the airway in a hyperdynamic patient.

Fixes: drain the circuit, find the leak, reduce trigger sensitivity.

Cycling asynchrony

Premature cycling: the ventilator ends inspiration before the patient is done — often produces a double trigger. Delayed cycling: the ventilator keeps inflating after the patient has begun to exhale, seen as a pressure spike at the end of inspiration; classic in COPD on pressure support because high resistance makes flow decay slowly. Adjust the flow cycle-off percentage (raise it for COPD, lower it for restrictive disease).

APRV, HFNC and noninvasive support

High-flow nasal cannula (HFNC)

Heated, humidified oxygen at 30–60 L/min through a wide-bore nasal interface. Its benefits are more mechanical than most people assume:

• Washes out anatomic dead space in the nasopharynx, so more of each breath is fresh gas.
• Delivers a reliable, high FiO₂ because the flow exceeds the patient's inspiratory flow — a standard nasal cannula gets diluted by entrained room air.
• Generates a small amount of positive pressure (roughly 2–5 cmH₂O with the mouth closed) — a modest PEEP effect.
• Reduces work of breathing and is far better tolerated than a mask, allowing patients to talk, eat and cooperate.

FLORALI (2015) compared HFNC, standard oxygen and noninvasive ventilation in acute hypoxemic respiratory failure. Intubation rates did not differ significantly overall, but 90-day mortality was lower with HFNC, and in the subgroup with P/F ≤200 the intubation rate was lower too.¹⁵

Noninvasive positive pressure ventilation (NIPPV)

CPAP provides continuous pressure with no inspiratory boost — pure recruitment and preload/afterload effects. BiPAP adds an inspiratory pressure above the expiratory pressure; the difference between IPAP and EPAP is the pressure support, and that is what does the ventilating.

The evidence is strongest where the pathophysiology matches the therapy:

• COPD exacerbation with respiratory acidosis — the best-supported indication; NIPPV reduces intubation and mortality by unloading the respiratory muscles and counterbalancing auto-PEEP.
• Cardiogenic pulmonary edema — positive intrathoracic pressure reduces preload and LV afterload and pushes fluid out of the alveoli.
• Immunocompromised patients — avoiding intubation avoids ventilator-associated pneumonia.

It is much weaker, and potentially harmful through delayed intubation, in de novo hypoxemic respiratory failure, ARDS, and altered mental status.

APRV — airway pressure release ventilation

Four settings:

• P-high — the sustained inflation pressure (commonly 20–30 cmH₂O)
• T-high — how long it is held (commonly 4–6 seconds)
• P-low — the release pressure (usually set to 0)
• T-low — the brief release (commonly 0.4–0.8 seconds, titrated so that expiratory flow terminates at roughly 50–75% of peak expiratory flow)

The concept: hold the lung open at a high mean airway pressure for a long time (continuous recruitment), release briefly to dump CO₂, and let the patient breathe spontaneously on top of P-high throughout. Spontaneous effort recruits dependent lung — the diaphragm pulls dorsally, where ARDS is worst.

Advantages: high mean airway pressure with a lower peak; preserved spontaneous breathing; often less sedation. Cautions: the very short T-low means auto-PEEP is intentional and must be managed deliberately; tidal volumes during spontaneous efforts are not directly controlled and can be injurious; hemodynamic effects of a sustained high mean pressure; and it requires a team fluent in the mode. Evidence is mixed — a single-center randomized trial found shorter ventilation with early APRV in ARDS, but larger confirmatory data are lacking.¹⁷

Adjuncts

• Inhaled pulmonary vasodilators (nitric oxide, epoprostenol): dilate vessels only in ventilated lung units, improving V/Q matching. Improve oxygenation; no demonstrated mortality benefit; used as a bridge.
• Recruitment maneuvers: sustained inflation to open collapsed lung. The ART trial showed harm with an aggressive strategy, and guidelines advise against routine use.⁸
• Esophageal balloon manometry: estimates pleural pressure so you can calculate transpulmonary pressure — most useful in obesity, ascites, and chest wall restriction, where a high plateau pressure may reflect the chest wall rather than the lung.

Complications and alarms

The alarms, and what each one means

Sudden deterioration: DOPE

• Displacement — the tube has migrated into the right mainstem or out of the trachea. Check depth at the lip, listen bilaterally, check waveform capnography.
• Obstruction — mucus plug, blood clot, the patient biting. Pass a suction catheter; if it will not pass, the tube is obstructed.
• Pneumothorax — sudden hypotension, unilateral absent breath sounds, rising peak pressure with a rising plateau, tracheal deviation late.
• Equipment — circuit disconnection, ventilator failure, oxygen supply failure.

Bag with 100% oxygen while working through it, and feel the bag: a stiff bag suggests obstruction or pneumothorax, an easy bag with no chest rise suggests displacement or a leak.

Ventilator-associated events

VAP prevention bundle — head of bed 30–45°, daily sedation interruption and readiness-to-extubate assessment, oral care with attention to the evidence on chlorhexidine (which has become more contested and varies by institution), subglottic secretion drainage where available, VTE and stress-ulcer prophylaxis per protocol, and early mobility. Nurses execute nearly all of it.

Cuff management — 20–30 cmH₂O. Too low permits microaspiration and VAP; too high causes tracheal mucosal ischemia, and over days, tracheal stenosis or a tracheoesophageal fistula. Measure rather than estimate; the "minimal leak" and palpation techniques are unreliable.

Other complications

• Ventilator-induced diaphragm dysfunction: the diaphragm atrophies measurably within days of full support — an argument for preserving spontaneous effort where safe.
• Hemodynamic effects: positive intrathoracic pressure reduces venous return and preload. A patient who becomes hypotensive immediately after intubation and initiation of positive pressure is usually volume-depleted, vasodilated from induction drugs, or air-trapping — not "having a cardiac event."
• Oxygen toxicity: prolonged high FiO₂ produces reactive oxygen species and absorption atelectasis. Wean FiO₂ toward a target of SpO₂ 88–95% rather than maximizing saturation.
• Barotrauma: pneumothorax, pneumomediastinum, subcutaneous emphysema. Suspect it with any acute rise in pressures plus hemodynamic change.
• PICS and post-ICU outcomes: delirium, ICU-acquired weakness, and PTSD are downstream complications of ventilation and sedation that appear in the Behavioral/Psychosocial section of the blueprint.

Liberation: SAT, SBT and extubation

Liberation from the ventilator is a nursing-driven process in most ICUs, and the exam treats it that way.

Daily paired trials

The ABC trial paired a daily spontaneous awakening trial (SAT) with a spontaneous breathing trial (SBT) and found more ventilator-free days, shorter ICU and hospital stays, and lower one-year mortality than the SBT-only strategy — a number needed to treat of about seven to save a life at one year.¹⁸ Pairing them is the standard of care.

SAT — hold sedation and assess. Safety screen first: no active seizures, no alcohol withdrawal, no agitation requiring escalating sedation, no neuromuscular blockade, no myocardial ischemia in the last 24 hours, no rising ICP. Failure criteria include sustained anxiety or agitation, respiratory rate over ~35 for several minutes, SpO₂ under 88%, respiratory distress, or an acute arrhythmia.

SBT — if the SAT passes, screen for the breathing trial.

SBT readiness screen

• The cause of respiratory failure is resolving or resolved — the single most important criterion
• Adequate oxygenation: P/F >150–200, or FiO₂ ≤0.4–0.5 with PEEP ≤5–8
• Hemodynamically stable: no vasopressors, or a low and stable dose
• Able to initiate a spontaneous breath
• pH >7.25; no significant metabolic derangement
• Adequate mentation and cough; a manageable secretion burden

Conducting the trial

30 minutes to 2 hours on low pressure support (typically 5–8 cmH₂O with PEEP 5), CPAP alone, or a T-piece. Longer trials do not improve outcomes and tire the patient.

Failure criteria — stop the trial and return to full support:

• Respiratory rate >35 sustained
• SpO₂ <90% or a substantial fall
• Heart rate >140 or a >20% change; new arrhythmia
• Systolic BP >180 or <90
• Marked accessory muscle use, paradoxical abdominal motion, diaphoresis, agitation or anxiety

Extubation is a separate decision

Passing an SBT means the patient can breathe. Extubation asks a different question: can they protect their airway?

• Mentation and cooperation — can they follow commands?
• Cough strength — a weak cough predicts failure regardless of the SBT.
• Secretion burden — a patient needing suction every 1–2 hours will likely fail.
• Cuff leak test — deflate the cuff and check for an audible leak or a fall in exhaled volume. Absence of a leak suggests laryngeal edema; consider steroids before extubation in a high-risk patient. The test's predictive value is moderate, so interpret it alongside the risk profile (prolonged intubation, traumatic intubation, large tube, female sex).

Post-extubation support

Prophylactic support reduces reintubation in high-risk patients. HIGH-WEAN (2019) found that HFNC combined with noninvasive ventilation immediately after extubation reduced reintubation compared with HFNC alone in patients at high risk of failure.¹⁹ High risk usually means older age, ventilation longer than 7 days, chronic cardiac or lung disease, obesity, or a weak cough.

Watch for post-extubation stridor (laryngeal edema — racemic epinephrine, steroids, and readiness to reintubate) and for delayed failure, which typically appears at 24–72 hours from fatigue, secretions or fluid overload rather than immediately.

Tracheostomy

Considered for anticipated prolonged ventilation, typically after 7–14 days. Advantages: less sedation, easier oral care and mobility, less dead space and resistance, improved comfort, easier weaning. Trials of early versus late tracheostomy have not shown a consistent mortality benefit, so timing is individualized.