Critical Care & Ventilator Management

Ventilation is often taught as a set of modes and numbers, which conceals the two ideas that determine outcome.

The ventilator does not treat the lung. It buys time. Nothing about positive pressure ventilation reverses pneumonia, sepsis or acute respiratory distress syndrome. It substitutes for failing gas exchange while the underlying cause is treated, and the treatment of the cause is what determines survival.

The ventilator itself injures the lung, and that injury is the central problem of the discipline. Overdistension, repeated opening and closing of alveoli, and the inflammatory response to both, produce ventilator-induced lung injury that can exceed the harm of the original disease.

Everything modern in ventilation follows from that second idea. Low tidal volumes, plateau pressure limits, positive end-expiratory pressure and prone positioning are all attempts to limit machine-induced harm rather than to improve any number on the screen.

A third principle governs the whole of critical care. Accepting an imperfect number to avoid a harmful intervention is frequently the right decision, which is why permissive hypercapnia, conservative oxygen targets and restrictive transfusion thresholds all exist.

1. When to Ventilate

The decision is clinical rather than numerical, and the reasoning divides into three.

Failure to oxygenate, meaning hypoxaemia despite maximal supplemental oxygen.

Failure to ventilate, meaning rising carbon dioxide with acidosis and exhaustion.

Failure to protect the airway, meaning reduced consciousness or absent protective reflexes, which is why a Glasgow Coma Scale of 8 or below is conventionally taken as an indication regardless of the gas exchange.

A patient who is tiring is ventilated before they arrest, not after. Rising respiratory rate with falling tidal volume, use of accessory muscles, inability to speak in sentences and a rising carbon dioxide in a previously tachypnoeic patient all indicate exhaustion.

A normalising carbon dioxide in a severe asthmatic is an ominous sign, not a reassuring one, because it means ventilation has fallen to a level that can no longer keep pace with a rising demand.

2. Modes Without the Jargon

Every mode answers two questions: what triggers a breath, and what ends it.

Volume control delivers a set tidal volume, so volume is guaranteed and pressure varies with compliance. A stiff lung generates high pressures.

Pressure control delivers a set pressure, so pressure is limited and volume varies with compliance. A stiffening lung receives progressively smaller volumes, which must be watched for.

Assist control delivers a full breath whether the ventilator or the patient triggers it.

Synchronised intermittent mandatory ventilation delivers a set number of mandatory breaths and allows spontaneous breaths between them.

Pressure support assists each spontaneous breath with a set pressure and is the usual weaning mode, since the patient controls rate, timing and volume.

Continuous positive airway pressure applies constant pressure without additional inspiratory support, which recruits alveoli and reduces work of breathing.

The mode matters far less than the settings. Tidal volume, plateau pressure, positive end-expiratory pressure and oxygen fraction determine both gas exchange and lung injury, and arguments about mode obscure this.

3. Ventilator-Induced Lung Injury

Four mechanisms are described, and they explain every protective strategy.

Volutrauma is injury from overdistension, and it is driven by volume rather than pressure, which is why tidal volume is limited.

Barotrauma is injury from high pressure, producing pneumothorax and pneumomediastinum.

Atelectrauma is injury from repeated opening and collapse of alveoli with each breath, and it is prevented by adequate positive end-expiratory pressure.

Biotrauma is the systemic inflammatory response released by the injured lung, which contributes to multi-organ failure and explains why lung injury kills through organs other than the lung.

Plateau pressure is the number that matters, measured during an inspiratory hold, because it reflects alveolar distending pressure rather than the resistance of the tube and airways. Keeping it below about 30 cm of water is a core target.

4. Acute Respiratory Distress Syndrome

Defined by acute onset within a week of a known insult, bilateral opacities not explained by effusion or collapse, respiratory failure not explained by cardiac failure or fluid overload, and impaired oxygenation graded by the ratio of arterial oxygen to inspired fraction.

The lung is not uniformly diseased, which is the key concept. Consolidated dependent regions receive no ventilation while aerated regions receive all of it, so the available lung is small. This is the baby lung, and it is why a normal tidal volume delivered to an adult with this syndrome overdistends the small aerated fraction.

The interventions that reduce mortality

Low tidal volume ventilation, at around 6 mL per kilogram of predicted body weight, reduces mortality, and this is the single most important finding in the field.

Predicted body weight is calculated from height, not measured weight, because lung size correlates with height and an obese patient does not have larger lungs.

Prone positioning reduces mortality in severe disease, by improving the distribution of ventilation and perfusion and reducing overdistension of the non-dependent lung.

Permissive hypercapnia is accepted as the price of low tidal volumes, since the harm from ventilator-induced injury exceeds the harm from a raised carbon dioxide within reasonable limits.

Higher positive end-expiratory pressure recruits collapsed alveoli and prevents atelectrauma, and it is titrated rather than fixed. Its cost is reduced venous return and the risk of overdistending already aerated lung, which is why it is set against oxygenation, compliance and haemodynamics rather than pushed to a maximum.

5. Weaning and Its Failures

Weaning begins on the day ventilation begins, because prolonged ventilation causes diaphragmatic atrophy, ventilator-associated pneumonia and delirium.

Daily assessment of readiness combined with a spontaneous breathing trial is the most effective approach, and it outperforms gradual reduction of support.

Paired daily sedation interruption and spontaneous breathing trials shorten ventilation and intensive care stay, because oversedation is the commonest reason a patient who could breathe does not.

The commonest causes of failed weaning are excessive respiratory load, respiratory muscle weakness, cardiac failure unmasked by the switch from positive to negative intrathoracic pressure, and delirium or oversedation.

The cardiac cause is worth understanding. Positive pressure ventilation reduces venous return and left ventricular afterload. Removing it increases both, which can precipitate pulmonary oedema in a patient with limited cardiac reserve, so failure to wean is sometimes a cardiac problem presenting as a respiratory one.

6. Shock and Oxygen Delivery

Shock is inadequate tissue oxygen delivery, not a blood pressure number. A patient can be normotensive and shocked, and lactate and base deficit are better markers than pressure.

Noradrenaline is the first-line vasopressor in septic shock, and early lactate measurement, cultures before antibiotics, and antibiotics within the first hour are the interventions that change outcome.

Source control is the intervention most often delayed. No antibiotic regimen compensates for an undrained abscess, an obstructed urinary tract or dead bowel, and identifying and correcting the source is frequently the difference between response and deterioration.

Fluid resuscitation has become more conservative, since excessive fluid worsens oedema, oxygenation and outcome, and balanced crystalloids are preferred over normal saline, which causes hyperchloraemic acidosis in volume.

Fluid responsiveness is assessed rather than assumed. A passive leg raise, or the variation in stroke volume with the respiratory cycle in a ventilated patient, predicts whether more fluid will increase cardiac output, and roughly half of haemodynamically unstable patients are not fluid responsive at all.

7. Oxygen Therapy and Non-Invasive Support

Escalation runs from simple oxygen through non-invasive support to intubation, and choosing correctly avoids both under-treatment and unnecessary intubation.

Delivery devices

Low-flow devices such as nasal cannulae and simple masks deliver a variable inspired oxygen fraction, because the patient entrains room air around the device and the proportion depends on their inspiratory flow.

Venturi masks deliver a fixed fraction by entraining air in a fixed ratio at high total flow, which is why they are used where a precise concentration matters, as in a patient at risk of carbon dioxide retention.

High-flow nasal oxygen delivers heated humidified gas at flows exceeding inspiratory demand, providing a reliable fraction, washing out anatomical dead space and generating a small positive pressure. It is better tolerated than a mask and has become first-line in many patients with hypoxaemic respiratory failure.

Non-invasive ventilation

Its strongest evidence is in two conditions. Exacerbations of chronic obstructive pulmonary disease with hypercapnic acidosis, where it reduces intubation and mortality, and cardiogenic pulmonary oedema, where positive pressure reduces preload and afterload and improves oxygenation rapidly.

It is used cautiously in hypoxaemic failure from pneumonia or acute respiratory distress syndrome, because delayed intubation while a patient generates large tidal volumes against a stiff lung can itself injure the lung, a phenomenon described as patient self-inflicted lung injury.

A trial of non-invasive support requires a clear time limit and a reassessment, since the harm lies in persisting with it in a patient who is not improving rather than in trying it.

Contraindications include impaired consciousness, inability to protect the airway, facial trauma, vomiting, haemodynamic instability and untreated pneumothorax.

8. Sedation, Delirium and the Complications of Intensive Care

Sedation is titrated to a target

Deep sedation prolongs ventilation, lengthens stay and increases delirium and mortality, so the modern aim is the lightest level compatible with safety and comfort, assessed against a validated scale.

Analgesia comes first. Treating pain adequately reduces the sedative requirement substantially, and an agitated patient is assessed for pain, hypoxia, hypercapnia, full bladder and delirium before the sedation is increased.

Daily sedation interruption allows assessment of neurological state and readiness to wean, and paired with a spontaneous breathing trial it shortens ventilation and stay.

Delirium

Delirium affects a large proportion of ventilated patients and is independently associated with mortality and long-term cognitive impairment, so it is a diagnosis to be screened for rather than a nuisance to be sedated.

Hypoactive delirium is commoner than hyperactive and is routinely missed, because a quiet patient causes no trouble. Screening tools designed for the intensive care unit detect it.

Management is non-pharmacological first: early mobilisation, sleep hygiene, orientation, glasses and hearing aids, family presence, and minimising deliriogenic drugs including benzodiazepines and anticholinergics.

The complications that accumulate

Ventilator-associated pneumonia is reduced by head-up positioning, oral care, avoiding unnecessary sedation and daily weaning assessment.

Venous thromboembolism, stress ulceration, pressure injury and critical illness neuromyopathy all follow from immobility and severity of illness, and each has a specific preventive measure.

Early mobilisation is an intervention with evidence behind it, reducing weakness and delirium, and it is possible in ventilated patients rather than reserved for those already extubated.

9. Worked Examples

Example 1. A 70 kg man of 170 cm with acute respiratory distress syndrome is ventilated with a tidal volume of 700 mL. What is wrong and what should it be?

The tidal volume is roughly double what it should be, and it is calculated from the wrong weight.

Low tidal volume ventilation at around 6 mL per kilogram of predicted body weight reduces mortality, and predicted body weight is derived from height rather than measured weight, because lung volume correlates with height and a heavier patient does not have larger lungs.

For a man of 170 cm, predicted body weight is around 66 kg, giving a target tidal volume of roughly 400 mL rather than 700 mL.

The reason it matters is the baby lung concept. In this syndrome the lung is not uniformly affected, so consolidated dependent regions receive nothing and the small aerated fraction receives the entire tidal volume, which overdistends it and produces volutrauma and biotrauma.

Plateau pressure should be measured on an inspiratory hold and kept below about 30 cm of water, and any resulting hypercapnia is accepted as the price of protection.

Example 2. A severe asthmatic has been tachypnoeic for two hours. The carbon dioxide, previously 28, is now 42 and the patient is quieter. Interpret.

This is deterioration, not improvement, and it is one of the most dangerous misreadings in emergency practice.

A patient with acute severe asthma hyperventilates, so the expected carbon dioxide is low. A normal value means that alveolar ventilation has fallen to a level that only just matches a carbon dioxide production that is itself raised by the work of breathing, and a rising value means ventilation is now inadequate.

The clinical picture supports it. Becoming quieter reflects reduced air movement rather than reduced bronchospasm, which is the same reasoning behind the silent chest, and exhaustion follows a period of maximal effort.

This patient requires urgent senior help, preparation for intubation, continued maximal bronchodilator therapy, magnesium and steroid, and consideration of intensive care.

If ventilated, the strategy is deliberately unconventional: low respiratory rate with prolonged expiratory time to allow complete exhalation, accepting hypercapnia, because dynamic hyperinflation and breath stacking cause circulatory collapse.

Example 3. A ventilated patient becomes suddenly hypotensive with high airway pressures and reduced breath sounds on one side. What is the diagnosis and immediate action?

Tension pneumothorax, and the diagnosis is made clinically without waiting for imaging.

Positive pressure ventilation drives air through a breach in the visceral pleura with each breath and prevents its escape, so intrapleural pressure rises progressively. That collapses the lung, shifts the mediastinum, kinks the great veins and obstructs venous return, producing hypotension.

The combination of high airway pressures, unilateral reduced breath sounds and hypotension in a ventilated patient is the pattern, and tracheal deviation is a late sign.

Immediate action is decompression followed by intercostal drain insertion. Waiting for a chest radiograph in a hypotensive ventilated patient is a recognised cause of avoidable death.

The differential for sudden deterioration in a ventilated patient is worth having as a list: displaced tube, obstructed tube, pneumothorax and equipment failure.

Example 4. A patient repeatedly fails spontaneous breathing trials despite resolved pneumonia and normal gases on support. What should be considered?

Several causes beyond the lungs, and the most commonly missed is cardiac.

Positive pressure ventilation reduces venous return and reduces left ventricular afterload by lowering transmural pressure. Removing it reverses both effects, increasing preload and afterload simultaneously, which can precipitate pulmonary oedema in a patient with limited cardiac reserve. Weaning-induced cardiac failure therefore presents as a respiratory failure, and echocardiography during a trial can demonstrate it.

Oversedation is the next consideration, and paired daily sedation interruption with spontaneous breathing trials is the intervention that addresses it.

Diaphragmatic weakness develops within days of controlled ventilation and is compounded by critical illness neuromyopathy, malnutrition and electrolyte disturbance including low phosphate and magnesium.

Other contributors include delirium, pain, an unnecessarily narrow tube increasing resistance, and untreated intra-abdominal pressure or pleural effusion increasing load.

Example 5. Why is permissive hypercapnia accepted in acute respiratory distress syndrome when raised carbon dioxide causes acidosis?

Because the alternative causes more harm than the acidosis does.

Delivering enough tidal volume to normalise carbon dioxide in this syndrome requires volumes that overdistend the small aerated fraction of lung, since consolidated regions accept no ventilation. That overdistension produces volutrauma and triggers biotrauma, a systemic inflammatory response that contributes to multi-organ failure and to death.

The trial evidence is unambiguous: low tidal volume ventilation at around 6 mL per kilogram of predicted body weight reduces mortality compared with conventional volumes, despite producing higher carbon dioxide levels and lower pH.

Moderate hypercapnia is generally well tolerated. It causes vasodilatation, increased cerebral blood flow and sympathetic stimulation, and the acidosis is buffered over hours by renal bicarbonate retention.

The strategy is limited rather than unlimited. It is avoided or used cautiously where raised intracranial pressure, severe pulmonary hypertension or severe metabolic acidosis are present, and extremes of acidosis are corrected.

Summary

The ventilator buys time; treating the cause determines survival.

The ventilator injures the lung, and modern strategy limits that harm.

Accepting an imperfect number to avoid a harmful intervention is often correct.

Ventilate for failure to oxygenate, to ventilate, or to protect the airway.

Ventilate the exhausted patient before arrest, not after.

A normalising carbon dioxide in severe asthma is ominous.

Volume control guarantees volume; pressure control limits pressure.

Settings matter more than mode.

Volutrauma is overdistension; atelectrauma is repeated opening and closing.

Biotrauma explains why lung injury causes multi-organ failure.

Plateau pressure reflects alveolar distension and is kept below about 30 cm water.

In acute respiratory distress syndrome the lung is not uniformly diseased.

The baby lung concept explains why normal tidal volumes overdistend.

Low tidal volume at 6 mL/kg predicted body weight reduces mortality.

Predicted body weight is calculated from height, not measured weight.

Prone positioning reduces mortality in severe disease.

Permissive hypercapnia is the accepted price of lung protection.

Weaning begins on day one, using spontaneous breathing trials.

Failed weaning is often cardiac, from restored preload and afterload.

Shock is inadequate oxygen delivery, not a blood pressure number.