Ahpra registration PHY0002298174
How these guides are written and reviewed →
When someone is on a ventilator, a tube passes through the windpipe and their cough no longer works normally, so mucus can build up. Physiotherapists use two main techniques to deal with it. Hyperinflation means giving a bigger-than-usual breath — either by hand with a bag, or by adjusting the ventilator — taken in slowly, held, then released quickly, so the rush of air out carries mucus up towards the tube. Suction means passing a fine catheter down the tube to remove it. Both work, and both can upset breathing, heart rhythm and blood pressure, so they are done only when there is a clear reason — never simply because it is a scheduled time.
What these techniques are
Manual hyperinflation (MHI, and colloquially “bagging”) is the delivery of a breath larger than the ventilator's set tidal volume using a self-inflating or flow-inflating circuit connected to an artificial airway. Ventilator hyperinflation (VHI) achieves the same physiological aim by adjusting the ventilator itself, without disconnection. Endotracheal suction is the direct removal of secretions through a catheter passed into the artificial airway.
All three are physiotherapy interventions for a specific problem: secretions that a patient with an artificial airway cannot clear because the cough is bypassed, weakened or absent. None of them is a routine, and each carries real risk.
The physiological rationale
Hyperinflation
A larger-than-normal breath has three effects. It re-expands collapsed lung units and, through interdependence, pulls open neighbouring units. It shifts gas behind mucus plugs via collateral channels. And — the key mechanism — the rapid release at the end of the inflation produces a period of expiratory flow that exceeds the preceding inspiratory flow, generating a shear force at the air–mucus interface that moves secretions cephalad.
This expiratory flow bias is the active ingredient. In bench and animal work, mucus moves towards the mouth only when peak expiratory flow exceeds peak inspiratory flow by roughly 10%, or about 17 L/min; below that, and particularly when inspiratory flow dominates, mucus can be driven distally.1 A fast squeeze on the bag with a slow release is therefore not merely ineffective — it is the wrong direction.
Suction
Suction is mechanical removal, and it only reaches secretions within the catheter's reach: the trachea and main bronchi. It does nothing for peripheral secretions, which is why it is combined with techniques that move mucus proximally rather than used as a substitute for them.
When these techniques are indicated
- Audible or auscultated secretions in a patient with an artificial airway who cannot clear them.
- Rising airway pressures or falling tidal volumes on the ventilator attributable to secretions.
- A sawtooth pattern on the flow–volume loop.
- Lobar or segmental collapse thought to be secretion-related — see atelectasis.
- Deteriorating gas exchange with a plausible secretion or volume-loss explanation.
- Before extubation or decannulation where secretion load is a concern.
A scheduled time of day; a routine ward round order; a diagnosis alone; a “wet-sounding” upper airway that is actually oral secretions; or an ARDS lung, which is a disease of alveolar flooding and inflammation, not of secretion retention. Suction and hyperinflation delivered without an indication expose the patient to derecruitment, hypoxaemia, arrhythmia, mucosal injury and pain for no expected benefit.
Manual hyperinflation: technique
| Element | Practice | Why |
|---|---|---|
| Circuit | Bag with a manometer in line, PEEP valve fitted, oxygen flow 12–15 L/min | Pressure must be seen, not guessed; PEEP loss is the main cause of derecruitment |
| Inspiration | Slow, over 2–3 seconds | Promotes even distribution and collateral filling; a fast breath goes to compliant regions only |
| Inspiratory pause | Hold 2–3 seconds at the top | Time for gas to move behind secretions and to recruit slow-filling units |
| Expiration | Rapid, uninterrupted release | Creates the expiratory flow bias that moves mucus proximally |
| Pressure limit | Typically ≤40 cmH2O, lower if the lung is injured or a fresh anastomosis is present | Barotrauma and haemodynamic compromise |
| Volume | Around 50% above the set tidal volume as a guide | Larger volumes give more recruitment but more haemodynamic cost |
| Sets | About 6 breaths, then allow normal ventilation; repeat for 3–6 cycles | Limits hypocapnia and haemodynamic disturbance |
| Combine with | Positioning, expiratory chest wall vibration or compression timed to the release | Adds to the expiratory flow bias and directs it regionally |
| Monitor | SpO2, ECG, blood pressure, airway pressure, ICP if monitored, patient comfort | The intervention is stopped by the monitor, not by the protocol |
Ventilator hyperinflation: the alternative
VHI delivers the same physiology by temporarily increasing tidal volume or inspiratory pressure on the ventilator, often in pressure control with a prolonged inspiratory time. Its advantages are substantial: no disconnection, so PEEP is preserved and derecruitment, aerosol generation and infection risk are all reduced; delivered volumes and pressures are measured rather than estimated; and one operator can perform it, freeing hands for positioning or chest wall techniques.
Comparative studies show broadly equivalent effects on secretion clearance, compliance and gas exchange, with better haemodynamic and oxygenation stability using VHI.2,3 The main limitation is that the expiratory flow bias achievable depends on the ventilator and mode, and not every machine can produce a rapid enough release. Where local policy, competence and equipment allow, VHI is the reasonable default and MHI the technique reserved for circumstances requiring the tactile feedback of a bag or where the ventilator cannot deliver the required pattern.
Contraindications and cautions for hyperinflation
- Undrained pneumothorax — absolute. See pneumothorax and chest drains.
- Undrained or unresolved bronchopleural fistula, and recent lung resection or airway anastomosis — discuss with the surgical team and use reduced pressures if agreed at all.
- High PEEP dependency (broadly PEEP ≥10 cmH2O) — disconnection causes derecruitment that may take an hour to recover; use VHI or a closed system.
- Haemodynamic instability, high vasopressor requirement, uncontrolled arrhythmia — hyperinflation reduces venous return.
- Raised intracranial pressure or unstable cerebral perfusion — both hyperinflation and suction raise ICP; coordinate with sedation and monitor directly.
- Severe bronchospasm and dynamic hyperinflation — risk of breath stacking and haemodynamic collapse.
- Bullous emphysema, and severe ARDS where recruitment is being managed by a protective ventilation strategy.
- Unstable spinal or rib injury where chest wall techniques form part of the treatment.
Endotracheal suction: doing it well
| Parameter | Practice |
|---|---|
| Indication | Assessed need, not a schedule |
| Catheter size | External diameter no more than half the internal diameter of the tube |
| Suction pressure | Lowest effective; commonly 80–150 mmHg in adults, lower in children |
| Duration | Under 15 seconds per pass, with recovery between passes |
| Depth | Shallow — to the tip of the tube, or just beyond only if secretions are not reached. Routine deep suction damages the carinal mucosa |
| Pre-oxygenation | Consider 100% oxygen before and after in patients who desaturate; avoid prolonged hyperoxia |
| System | Closed in-line suction preserves PEEP and reduces derecruitment and aerosol exposure; preferred in high-PEEP, hypoxaemic or infectious patients |
| Saline instillation | Not routine. It does not reliably thin secretions, causes desaturation and distress, and may propel organisms distally4,5 |
| Humidification | The real answer to thick secretions. Check the humidifier before increasing suction frequency |
| Analgesia and sedation | Suction is painful and frightening. Plan for it; explain it even to the apparently unresponsive patient |
Subglottic secretion drainage, where the tube has that facility, reduces ventilator-associated pneumonia and should be used and checked rather than forgotten.
Complications
- Hypoxaemia and derecruitment — the most common, largely preventable by closed suction, pre-oxygenation and limiting duration.
- Arrhythmia and bradycardia — vagal stimulation and hypoxaemia.
- Haemodynamic disturbance — both hypertension from stimulation and hypotension from reduced venous return.
- Raised intracranial pressure — clinically important in acute brain injury; cluster interventions and pre-treat.
- Mucosal trauma and bleeding — from excessive pressure, oversized catheters and deep suction.
- Bronchospasm, particularly in reactive airways.
- Infection — breaks in the circuit and poor technique contribute to ventilator-associated pneumonia.
- Barotrauma from unmonitored hyperinflation pressures.
- Pain and distress — consistently reported by survivors as among the worst memories of intensive care.
Positioning and the rest of the treatment
Hyperinflation and suction work best when the lung region being treated is positioned to advantage, and they are frequently less important than the position itself. Side lying with the affected lung uppermost for secretion drainage, prone positioning for oxygenation in severe hypoxaemic failure, and simply sitting the patient up all change regional ventilation more than any single bagged breath. Where the patient can participate, active techniques and early mobilisation displace passive ones entirely — an extubated patient sitting on the edge of the bed does not need bagging.
Role of the physiotherapist
These are advanced, high-risk techniques and in Australian intensive care they are core physiotherapy scope. Four responsibilities define good practice.
Deciding whether to treat at all. The most important skill is the negative decision. Assessment — auscultation, the ventilator waveform, imaging, gas exchange, the volume and character of aspirate, the trend — determines whether an intervention is indicated. Documenting the reasoning for not treating is as valuable as documenting a treatment.
Delivering the technique precisely. Slow in, hold, fast out; measured pressures; PEEP preserved; sets not marathons; monitoring watched by someone whose only job is to watch it. The difference between effective and harmful hyperinflation is a matter of seconds and centimetres of water.
Protecting against the routine. Challenging standing orders for scheduled suction, unnecessary saline instillation, and “chest physio TDS” requests that have outlived their indication. The evidence supports assessment-based intervention; the culture often supports the timetable.
Moving upstream. Humidification, positioning, sedation minimisation, early mobilisation and cough augmentation reduce the need for these techniques — see assisted cough and ICU-acquired weakness. The best measure of a respiratory physiotherapy service in intensive care is not how much bagging it does.
Evidence summary
Framing. Manual hyperinflation and endotracheal suction sit at the boundary between essential and harmful. Their physiological effects are demonstrable and immediate; their trial evidence for patient-centred outcomes is thin; and their capacity to destabilise a critically ill patient is well documented. The literature does not support their routine use, and it does not support abandoning them. It supports precise indication-based delivery by a competent operator, and preference for the least destabilising route to the same physiology.
Expiratory flow bias is the mechanismWork by Kim and later by Maxwell and colleagues established that mucus transport direction depends on the relationship between peak expiratory and peak inspiratory flow: a bias favouring expiration of roughly 10%, or about 17 L/min, is required to move mucus cephalad, while an inspiratory-dominant pattern moves it distally.1 This single finding explains most technique guidance — the slow inspiration, the pause and the rapid release — and explains why an operator who squeezes fast and releases slowly may be actively worsening the problem. It also explains why chest wall compression or vibration timed to the release adds to the effect: it increases expiratory flow, not inspiratory volume.
Manual versus ventilator hyperinflationBerney and Denehy compared MHI and VHI in ventilated patients and found comparable sputum yield and static compliance, with VHI producing more stable haemodynamics and oxygenation.2 A subsequent systematic review of VHI concluded that it improves compliance and secretion clearance with a safety profile at least equivalent to MHI, while avoiding circuit disconnection and PEEP loss.3 Delivered volumes during MHI vary substantially between operators even when experienced, which is an argument for in-line manometry and for VHI where equipment permits. The remaining case for MHI is tactile feedback on compliance and the ability to shape the flow pattern in ways some ventilators cannot.
Suction technique and salineThe AARC clinical practice guideline recommends suction only when clinically indicated rather than routinely, shallow rather than deep suction, catheters occluding no more than half the airway lumen, brief duration, and pre-oxygenation in patients who desaturate; it does not recommend routine normal saline instillation.6 Saline instillation has been shown to cause greater oxygen desaturation than suction alone,4 and while one large single-centre trial reported fewer microbiologically confirmed ventilator-associated pneumonias with instillation, the finding has not been replicated and is outweighed in most units by the demonstrated physiological cost.5 Thick secretions are, in the great majority of cases, a humidification problem.
Physiotherapy implications- Use a manometer every time. Unmeasured hyperinflation pressures vary widely between operators and between breaths.
- Default to VHI where competence and equipment allow, particularly at PEEP ≥10 cmH2O, in hypoxaemic patients, and where aerosol generation matters.
- Slow in, hold, fast out. If the technique feels like resuscitation, the flow bias is wrong.
- Treat thick secretions by fixing humidification before increasing suction frequency or reaching for saline.
- Cluster interventions in patients with raised intracranial pressure, and pre-treat with sedation or analgesia rather than accelerating the technique.
- Reassess the indication at every attendance. A standing order for chest physiotherapy is not an indication, and discontinuing an ineffective intervention is a clinical decision worth documenting.
Three questions before touching the circuit. First: is this a secretion problem? Falling compliance in ARDS, pulmonary oedema and abdominal distension all mimic it and none respond. Second: what is the least destabilising way to achieve the same physiology — a position change, a ventilator adjustment, closed suction, or a cough in a patient who could be woken and sat up? Third: what will tell me it worked — sputum yield, compliance, oxygenation, the waveform — and when will I stop if it does not? Physiotherapy in intensive care earns its place through the accuracy of these decisions, not through the vigour of the treatment.
Evidence gapsNo trial has evaluated MHI or VHI against usual care with duration of ventilation, ventilator-associated pneumonia or mortality as primary endpoints, and the surrogate outcomes used — sputum wet weight, static compliance — correlate uncertainly with anything patients experience. Optimal hyperinflation dose, pressure ceiling and frequency are conventions. The interaction with lung-protective ventilation strategies is unstudied, so practice in the injured lung is extrapolated. Paediatric and neonatal parameters are largely borrowed from adult work. And the long-term consequences of repeated suction-related mucosal injury, including in patients with long-term tracheostomy, have not been characterised.
References & evidence base
- Maxwell L, Ellis ER. The effect of circuit type, volume delivered and “rapid release” on flow rates during manual hyperinflation. Aust J Physiother 1998;44:23–9.
- Berney S, Denehy L. A comparison of the effects of manual and ventilator hyperinflation on static lung compliance and sputum production in intubated and ventilated intensive care patients. Physiother Res Int 2002;7:100–8.
- Anderson A, Alexanders J, Sinani C, et al. Effects of ventilator vs manual hyperinflation in adults receiving mechanical ventilation: a systematic review of randomised clinical trials. Physiotherapy 2015;101:103–10.
- Ridling DA, Martin LD, Bratton SL. Endotracheal suctioning with or without instillation of isotonic sodium chloride solution in critically ill children. Am J Crit Care 2003;12:212–19.
- Caruso P, Denari S, Ruiz SAL, et al. Saline instillation before tracheal suctioning decreases the incidence of ventilator-associated pneumonia. Crit Care Med 2009;37:32–8.
- American Association for Respiratory Care. AARC clinical practice guidelines: endotracheal suctioning of mechanically ventilated patients with artificial airways 2010. Respir Care 2010;55:758–64.
References are numbered in citation order (Vancouver/BMJ style) and were current at the time of writing. Guidelines are living documents — verify against the latest version before clinical use.
Corrections: If something on this page is wrong, out of date or unclear, we want to know. Email reception@inspireclinic.au with the page name and what you believe is incorrect. Substantive corrections are made promptly, and the guide’s version and last-updated date are changed to reflect it.