What Is Plateau Pressure in Mechanical Ventilation?

Plateau pressure is the pressure measured inside the airways at the end of a breath, once gas flow has completely stopped and the system is momentarily at rest. In mechanical ventilation, it serves as a bedside estimate of how much stretching force the lungs and chest wall are experiencing during each breath. Keeping this number in check is one of the central goals of lung-protective ventilation, because excessive pressure is strongly linked to lung damage and higher mortality in critically ill patients.

How Plateau Pressure Is Measured

Measuring plateau pressure requires a brief pause at the end of the ventilator’s inspiratory cycle, called an end-inspiratory hold. During this pause, the ventilator’s valve closes so no more air flows in or out, and the pressure inside the circuit and lungs equilibrates. The number displayed on the ventilator screen at that moment of zero flow is the plateau pressure. In practice, clinicians typically perform this measurement while the patient is breathing passively, meaning the patient’s own respiratory muscles are not actively pushing or pulling against the ventilator.

The measurement itself takes only a fraction of a second of extra hold time, but the conditions matter. If the patient is making active breathing efforts, the muscles can distort the pressure reading, making the plateau unreliable. A multicentre analysis found that when patients were triggering breaths themselves, unreadable plateau measurements were common and tended to occur in patients with stronger inspiratory effort.

Another hidden confounder is trapped air. In patients with obstructive lung disease, air can get caught in the lungs before the next breath begins, creating an additional pressure called auto-PEEP. If auto-PEEP is present but unrecognized, the plateau pressure reading can overestimate the true stretching force being applied to the lungs, and it can also interfere with how the ventilator delivers pressure-targeted breaths.¹

Why the 30 cmH₂O Threshold Matters

For decades, critical care guidelines have recommended keeping plateau pressure at or below 30 cmH₂O in patients with acute lung injury or acute respiratory distress syndrome (ARDS). This threshold has become one of the anchoring numbers of lung-protective ventilation. A systematic review of studies spanning from 2000 onward found that the routinely recorded plateau pressure was already below 30 cmH₂O in the vast majority of published trials and observational studies, with mean values typically hovering in the mid-to-upper 20s.²

But what happens when the number climbs above 30? A meta-regression analysis looking at ARDS patients found that short-term mortality was significantly higher when plateau pressures exceeded 32 cmH₂O during the first three days of ICU admission. Pressures below 27 or 30 cmH₂O, by contrast, were not associated with a measurable reduction in death risk compared to slightly higher values.³ That suggests the relationship between plateau pressure and harm is not perfectly linear; the danger seems to ramp up most clearly once pressures push past the low 30s, especially in the early days of illness.

Some ventilation protocols have used even higher ceilings in specific circumstances. One randomized trial testing a strategy of high PEEP with recruitment maneuvers included a protocol for reducing PEEP when plateau pressure exceeded 40 cmH₂O, treating that as a hard stop for barotrauma risk.⁴ In most routine care, though, 30 cmH₂O remains the working upper limit.

Driving Pressure and Why Plateau Pressure Alone Is Not Enough

Plateau pressure tells you the total pressure the respiratory system is experiencing at end-inspiration, but it lumps together two different components: the pressure needed to inflate the lungs and the pressure needed to push against the chest wall and abdomen. To get a better sense of how much the lungs themselves are being stretched, clinicians subtract the PEEP (the baseline pressure maintained between breaths) from the plateau pressure. The result is called driving pressure.

Driving pressure has emerged as a stronger predictor of survival in ARDS than plateau pressure alone. A large analysis of individual patient data from multiple randomized trials found that for every increase of about 7 cmH₂O in driving pressure, the risk of death rose by roughly 40 percent. Strikingly, this held true even in patients whose plateau pressures and tidal volumes were already within the “protective” range.⁵ In other words, a patient could have a plateau pressure comfortably below 30, yet still be at elevated risk if their driving pressure was too high, typically because the amount of functional lung available to receive the breath was unusually small.

A separate analysis of two randomized controlled trials confirmed the pattern: nonsurvivors had a higher average driving pressure than survivors, and driving pressure independently predicted 90-day mortality even after accounting for other ventilator settings.⁶ The implication is that plateau pressure is a useful safety guardrail, but driving pressure gives a more granular view of whether each breath is proportional to the lung’s capacity to handle it.

Where Plateau Pressure Can Be Misleading

One of the trickiest aspects of plateau pressure is that it reflects the entire respiratory system, not just the lungs. The chest wall, the diaphragm, and the abdomen all contribute their own elastic resistance. In a patient with a normal body habitus lying flat in bed, the chest wall’s contribution is relatively modest and predictable. But in certain situations, the chest wall accounts for a large share of the total pressure, and that makes the plateau number misleading.

Obesity is the most common example. In mechanically ventilated patients with severe obesity, the weight of the abdominal contents presses upward against the diaphragm, and the chest wall itself is less compliant. A study of morbidly obese patients (average BMI around 44.5) found that their mean plateau pressures averaged about 40 cmH₂O, far above the conventional 30 cmH₂O threshold.⁷ Yet that high plateau pressure did not necessarily mean the lungs themselves were being over-stretched; much of the pressure was being consumed pushing against the heavy, stiff chest wall. The study found a significant relationship between transpulmonary pressure (the pressure actually across the lung) and plateau pressure, but plateau pressure alone gave an incomplete picture.

Similar issues arise in patients with tense abdominal distension from any cause, large pleural effusions, or chest wall edema. In all these cases, a plateau pressure above 30 might trigger alarm bells that lead clinicians to reduce tidal volumes or PEEP, even though the lungs themselves are not being excessively stressed. This paradox is a major reason the field has moved toward more refined measurements.

Transpulmonary Pressure and Esophageal Monitoring

If plateau pressure overstates lung stress in patients with stiff chest walls and potentially understates it in patients with very compliant chest walls, the natural question is: can we measure the lung component alone? The answer is yes, using a technique called esophageal pressure monitoring. A thin balloon catheter placed in the esophagus estimates the pressure in the pleural space, which is the space between the lung and the chest wall. Subtracting that pleural pressure estimate from the airway pressure gives the transpulmonary pressure, which reflects the actual stress on lung tissue.

Esophageal manometry has proven useful for estimating both lung stress and the intensity of a patient’s own breathing effort, particularly in ARDS.⁸ It allows clinicians to separate what the chest wall is doing from what the lung is experiencing.⁹ A research group that directly studied whether plateau pressure adequately reflects lung stress concluded that plateau pressure and tidal volume are, by themselves, inadequate surrogates for the true stress and strain within the lung.¹⁰

Despite these advantages, esophageal monitoring is not routinely used in every ICU. The catheter placement requires some expertise, positioning can affect accuracy, and interpreting the numbers in the context of a real patient (who may be on vasopressors, lying prone, or retaining fluid) adds complexity. Many ICU teams still rely on plateau pressure as the primary monitoring tool and reserve esophageal manometry for difficult cases where the standard numbers don’t seem to fit the clinical picture.

Strategies That Bring Plateau Pressure Down

When plateau pressure creeps too high, clinicians have several levers to pull, and understanding what each lever actually does helps clarify why plateau pressure is such a central monitoring target.

The most direct intervention is reducing tidal volume, the amount of air delivered with each breath. Smaller breaths mean less stretching force. This is the cornerstone of lung-protective ventilation in ARDS, where tidal volumes are typically set relative to the patient’s predicted body weight rather than their actual weight. Reducing tidal volume is effective at lowering both plateau pressure and driving pressure, but it comes with trade-offs: less air per breath means higher carbon dioxide levels unless the breathing rate goes up, and some patients tolerate the resulting higher CO₂ levels better than others.

Adjusting PEEP can also affect plateau pressure, though the relationship is less straightforward. Raising PEEP can recruit collapsed lung regions and actually improve compliance, which might keep plateau pressure from rising as much as expected. But in a lung that’s already well-recruited, adding more PEEP without reducing tidal volume will push plateau pressure higher with little benefit.

Neuromuscular blocking agents offer another route. By paralyzing the respiratory muscles, these drugs eliminate any patient effort that might contribute to chaotic pressure swings and allow the chest wall to relax fully. In patients with moderate to severe ARDS, neuromuscular blockade has been shown to decrease both plateau pressure and total PEEP, while also improving oxygenation.¹¹ The downside is that prolonged paralysis carries its own risks, including ICU-acquired weakness, so it is usually reserved for the sickest patients and used for limited periods.

Prone positioning, the practice of turning a patient face-down, does not necessarily lower plateau pressure in a straightforward way, but it changes the distribution of pressure across the lung. In the supine position, the weight of the heart and abdominal organs compresses the lung regions underneath, causing them to collapse while the upper regions get over-inflated. Turning the patient prone redistributes these forces, creating a more even spread of ventilation, stress, and strain throughout the lung.¹² The mortality benefit of prone positioning in severe ARDS appears to stem primarily from reduced overdistension in the non-dependent regions and less cyclic opening and closing of collapsed areas in the dependent regions.¹³ In effect, the same plateau pressure may be far less harmful in the prone position because the force is no longer concentrated in a few overstretched zones.

Compliance and What It Tells You About the Lung

You cannot fully understand plateau pressure without understanding compliance, the lung’s willingness to expand for a given amount of pressure. A highly compliant lung inflates easily; a stiff lung requires more pressure to accept the same volume of air. Compliance is calculated by dividing the delivered tidal volume by the driving pressure (plateau pressure minus PEEP).¹⁴

In healthy lungs, compliance is relatively high, meaning you can deliver a reasonable breath without generating much pressure. In ARDS, large portions of the lung are flooded or collapsed, so the functional lung that remains is much smaller. The same volume of air now gets crammed into a smaller space, compliance drops, and plateau pressure shoots up. This is why the same tidal volume that is perfectly safe for one patient can be dangerous for another: it depends on how much working lung is available.

Tracking compliance over time also helps clinicians gauge whether the lung is improving or worsening. If plateau pressure is climbing but tidal volume and PEEP haven’t changed, compliance is falling, which often signals worsening lung injury, fluid overload, or new atelectasis. If plateau pressure drops while settings stay the same, compliance is improving, potentially a sign that treatment is working.

When “Safe” Numbers Still Cause Harm

One of the harder lessons the field has absorbed over the past two decades is that staying below the 30 cmH₂O plateau pressure target does not guarantee safety. The driving pressure research showed this clearly: patients whose plateau pressures were within protective limits still had higher mortality if their driving pressure was elevated.⁵ Several attempts to find reliable clinical surrogates for lung stress and strain, including plateau pressure, CT-based density changes, and inflammatory markers, have all been found to have limitations.¹⁵

The reason is partly conceptual. Plateau pressure is a global measurement. It averages the pressure across the entire respiratory system, healthy lung regions and diseased lung regions alike. In a heterogeneous ARDS lung where some areas are open and others are consolidated or flooded, the local stress on the open areas can be far higher than what the single plateau number implies. Two patients can have identical plateau pressures yet wildly different levels of regional lung strain, because the distribution of damage is different. This heterogeneity is part of what makes ARDS so challenging to ventilate safely.

The practical takeaway for clinicians, and for families trying to understand a loved one’s ventilator settings, is that plateau pressure is a necessary monitoring tool but not a sufficient one. It works best when interpreted alongside driving pressure, compliance trends, and when available, transpulmonary pressure. No single number captures the full picture.

Plateau Pressure in Spontaneous and Assisted Breathing Modes

Most of the research on plateau pressure comes from patients who are deeply sedated or paralyzed, breathing passively on the ventilator. But many ICU patients are on assisted modes of ventilation, where the machine helps each breath but the patient initiates and contributes effort. Measuring plateau pressure in these patients is trickier.

During an inspiratory hold in assisted ventilation, the patient’s own muscles may still be contracting, which can either inflate the reading (if the muscles are pushing air in) or deflate it (if the muscles relax unevenly). An analysis of a multicentre database found that muscular activity frequently made plateau measurements unreliable during patient-triggered assisted ventilation, and that unreadable plateaus were more common in patients who were making stronger inspiratory efforts.¹⁶

This creates a practical gap. The patients on assisted modes are often the ones transitioning toward recovery, and understanding their lung mechanics during that transition matters for deciding when to reduce support. Some ICUs use brief periods of passive breathing (via a short-acting sedative or a transient switch in ventilator mode) to obtain a clean plateau reading. Others accept the limitation and rely more heavily on other markers like airway occlusion pressure or respiratory rate variability to guide weaning decisions. The field does not have a consensus solution yet, which is one of the open frontiers in ventilator monitoring.