Page of

Failure to wean from mechanical ventilation 

Failure to wean from mechanical ventilation
Failure to wean from mechanical ventilation
Focused Intensive Care Ultrasound

Andrew Walden



Failure to wean from mechanical ventilation is a common problem in intensive care and represents a significant burden in terms of prolonged ICU stay with associated morbidity and mortality. US examination can aid the systematic assessment of the underlying pathophysiology that is often complex and multifactorial. This chapter reviews the role of US in assessing the contribution of inadequate lung aeration, pleural effusion, and diaphragmatic and cardiac function to weaning failure.

Clinical relevance

Weaning from mechanical to spontaneous ventilation is perhaps the most important first goal on recovery from critical illness. Failure to wean is defined as an inability to pass a spontaneous breathing trial (SBT) (either via a T-piece or by using a minimal setting of pressure support) or the need for re-intubation within 48 hours following extubation. This is known to occur in 20–30% of patients and represents a significant burden in terms of prolonged ICU stay with associated morbidity and mortality.

The pathophysiology of weaning failure is often complex and multifactorial—respiratory, cardiac, and neurological problems can all occur together, especially in patients who have had prolonged stays on intensive care. It is important therefore to take a structured approach, identifying each factor and addressing it in turn. US can greatly aid this approach. Identification of diaphragmatic weakness, unknown pleural collections or ongoing evidence of interstitial oedema, and impaired lung aeration are all identified with US, as are abnormalities of cardiac function.

Respiratory causes of failure to wean

Many factors relating to the lungs can lead to failed weaning, many of which are amenable to assessment with US. Neuromuscular weakness can be detected by dynamic US of diaphragmatic function. Impaired respiratory mechanics due to pleural collections are easily diagnosed with LUS, as is loss of lung aeration due to consolidation or interstitial oedema.

Ultrasound assessment of lung aeration

LUS can be used to assess changes in lung aeration during an SBT. Loss of lung aeration during a successful SBT has been shown to predict post-extubation respiratory distress. Loss of aeration represents the integrated effects of impaired cardiac, respiratory, and diaphragmatic function and reflects associated lung derecruitment and pulmonary oedema secondary to increased PAOP. Aeration is assessed by US with a semi-quantitative score that is determined by dividing each lung into six zones (Figure 23.1). The aeration of each zone is systematically assessed, and a score allocated of between 0 (normal aeration) to 3 (complete loss of aeration), which results in a maximum total US aeration score of 36 (Table 23.1).

Figure 23.1 The regions of the lung ultrasound (LUS) score. Each hemi-thorax is divided up into six sectors, as shown.

Figure 23.1 The regions of the lung ultrasound (LUS) score. Each hemi-thorax is divided up into six sectors, as shown.

Reproduced with permission from Bouhemad et al., ‘Ultrasound for “Lung Monitoring” of Ventilated Patients’, Anesthesiology, 132: 437–447. Copyright © 2015, Wolters Kluwer Health.

Table 23.1 Lung aeration score (each of the 12 sectors is scored according to the ultrasound appearance, resulting in a maximum possible total score of 36)

Number of points






A line pattern with a maximum of two B-lines per sector

Failure to wean from mechanical ventilation


Moderate loss

Multiple B-lines regularly spaced or coalescent B-lines within a limited portion of the sector

Failure to wean from mechanical ventilation


Severe loss

Coalescent B-lines throughout sector

Failure to wean from mechanical ventilation


Total loss

Consolidated lung

Failure to wean from mechanical ventilation

Reproduced with permission from Soummer A et al. Ultrasound assessment of lung aeration loss during a successful weaning trial predicts postextubation distress Crit Care Med 2012; 40:2064–2072.

A dynamic aeration score may be calculated from the change in aeration before and after start of SBT. A change in aeration by one category scores 1, by two categories scores 3, and by three categories scores 5. For instance, a change from normal to moderate, a score of 1 is added; for a change from normal to severe, a score of 3 is added; for a change of normal to total loss, a score of 5 is added. These scores are then subtracted from the total aeration score and added where there is improvement. During an SBT, a score of <13 predicts a very low chance of post-extubation failure, whereas a score of >17 identifies an 85% chance of post-extubation failure.

Aside from its utility in predicting post-extubation distress following an SBT, LUS may also identify other issues which may be addressed prior to weaning from mechanical ventilation, including pulmonary oedema and atelectasis, which may be associated with pleural effusion or mucus plugging. Interventions, including inducing diuresis, pleural drainage, and physiotherapy, may help to improve lung function prior to an SBT or extubation.

Ultrasound assessment of pleural effusions

Pleural effusions are common in mechanically ventilated patients, with some studies reporting an incidence of over 50% in patients ventilated for >48 hours when examined with US. The physiological effects are varied but include:

  • An increased pleural pressure, which increases the distending pressure of the chest wall and reduces that of the lung. This results in uncoupling of the normal volume relationship between the lung and chest wall, with an increase in chest wall volume and a reduction in lung volume. A restrictive ventilatory defect is predicted with reduced vital capacity, functional residual capacity, and total lung capacity, although the loss of lung volume may not be as large as expected from the size of the effusion, as the lung may be largely floating on the effusion. Uncoupled lung recruits and derecruits during the respiratory cycle, which increases the risk of ventilator-induced lung injury. This effect can be mitigated by the application of PEEP.

  • Impaired chest wall and diaphragmatic mechanics due to outward and downward pressure—this leads to the muscles operating on an adverse portion of their length/tension curve and probably leads to subjective feelings of dyspnoea.

  • Altered ventilation–perfusion matching due to compression atelectasis of the underlying lung and the development of a shunt.

Given these adverse effects, it is reasonable to assume that thoracocentesis or pleural drainage may help to improve overall lung function and thus make weaning from mechanical ventilation more successful. There is some evidence to support this, although based on case series and surrogate outcome measures. A meta-analysis of studies showed a consistent improvement in oxygenation, and another study has demonstrated a correlation between improvements in oxygenation and volume of pleural effusion drained.

US easily identifies pleural effusion (Figure 23.2) and can be used for both qualitative and quantitative assessments of when to drain.

Figure 23.2 Large, free-flowing pleural effusion in a mechanically ventilated patient. The spleen and hemidiaphragm are clearly seen on the right-hand side of the image, with an extensive hypoechogenic area on the left-hand side of the image.

Figure 23.2 Large, free-flowing pleural effusion in a mechanically ventilated patient. The spleen and hemidiaphragm are clearly seen on the right-hand side of the image, with an extensive hypoechogenic area on the left-hand side of the image.

If there is evidence of septation or organization, then there is a compelling reason to perform a diagnostic tap. A pleural fluid pH of <7.2 or aspiration of pus are indications to drain the effusion to prevent ongoing infection or chronic scarring and lung trapping. The presence of flattening, or even inversion, of the hemidiaphragm implies it will be on an adverse portion of the length/tension curve, so inefficient and prone to cause subjective dyspnoea.

Quantitatively, it is possible to estimate the size of pleural effusions. The intra-pleural distance at the base of the lung (Figure 23.3) has been shown to correlate with volume of pleural fluid. In a supine patient, an intra-pleural distance at the base of the lung of 45 mm on the left or 50 mm on the right reliably predicts a pleural fluid volume of >800 mL. Similarly, with a patient supine with 15° elevation, the maximum intra-pleural distance when scanning in the posterior axillary line allows estimation of fluid volume. The volume approximates to the separation distance in millimetres multiplied by 20.

Figure 23.3 Measurement of the intra-pleural distance at the base of the lung.

Figure 23.3 Measurement of the intra-pleural distance at the base of the lung.

While there is no definitive volume identified at which to perform pleural drainage, the larger the effusion, the more likely there will be a benefit. There is no sense in draining small effusions, but where there is evidence of underlying lung collapse or a flattened hemidiaphragm associated with a moderate to large effusion, drainage is likely to improve respiratory mechanics, gas exchange, and subjective feelings of dyspnoea, which is likely to hasten weaning from mechanical ventilation. What is also clear is that pleural drainage using US in mechanically ventilated patients is a safe procedure with a low complication rate.

Diaphragmatic function

US visualization of the hemidiaphragm has been described in Chapter 16. There are two main ways of visualizing diaphragmatic function, either by examining the zone of apposition where the hemidiaphragm ‘inserts’ into the abdominal wall or by M-mode examination of the dome of the diaphragm over the liver and spleen in the mid-clavicular line. Dynamic visualization of the zone of apposition allows determination of the fractional thickening [(thickness at end-inspiration – thickness at end-expiration)/thickness at end-expiration]. Similarly, visualization of the dome of the hemidiaphragm allows an estimate of the total diaphragmatic excursion. Both these systems have been validated and perform well against correlates of diaphragmatic function, such as transdiaphragmatic pressure measurements, but notably this is in spontaneously breathing individuals. The application of PEEP increases lung volume and so affects the normal doming of the hemidiaphragm in expiration, thus limiting the utility of diaphragmatic excursion as a measure of function in mechanically ventilated patients. Fractional thickening probably remains a good measure of diaphragmatic efficiency, although at higher levels of PEEP and lung volume, it may become less useful.

Diaphragmatic US can be used to identify phrenic nerve injury, which may occur following cardiac surgery, by looking for asymmetric diaphragmatic dysfunction. During spontaneous breathing, the paralysed diaphragm will be noted to move cephalad during inspiration. If the function in the other hemidiaphragm is preserved, it appears that patients will still manage to wean successfully. A maximal diaphragmatic excursion of >25 mm in the non-affected diaphragm predicts a high likelihood of successful weaning, while if it is <25 mm, it suggests bilateral dysfunction and low likelihood of rapid weaning from mechanical ventilation.

In most other situations, diaphragmatic weakness is generalized due to systemic disorders. Critical illness polyneuromyopathy and specific neuromuscular disorders, such as Guillain–Barré syndrome, will lead to bilateral diaphragmatic weakness, meaning that unilateral assessment of function is adequate. Similarly, in conditions such as chronic obstructive pulmonary disease, high lung volumes due to gas trapping and increases in total thoracic capacity due to bullous disease can lead to impaired diaphragmatic function by placing the diaphragm on an adverse portion of the length/tension generation curve. In these situations, right-sided examination of the diaphragm can serve as an adequate proxy for whole diaphragmatic function.

Diaphragmatic excursion has been used in a general ICU population to predict successful weaning. In one study, a cut-off value of maximal excursion of 1.1 cm had a sensitivity of 84% and a specificity of 83% to predict re-intubation at 72 hours, and in another, a value of <10 mm for either hemidiaphragm defined dysfunction and was associated with a longer time to liberation from mechanical ventilation and re-intubation rates.

More recently, fractional thickening has been used as an effective tool to predict weaning failure. In patients with a fractional thickening of ≥30%, the positive predictive value for successful extubation was >90%, regardless of whether assessment was performed by SBT or on pressure support. Very similar results have been found using a cut-off value of 36% in patients with tracheostomy tubes in situ at being successfully weaned from respiratory support within 48 hours.

Cardiac failure—systolic and diastolic

A role for cardiac dysfunction in weaning failure is supported by evidence of a progressive reduction in mixed venous oxygen saturations, due to a fall in cardiac output, in patients who fail an SBT, compared to those who do not. The transition from PPV to spontaneous breathing is associated with a number of physiological effects that may predispose patients to a deterioration in cardiac function. The loss of positive intrathoracic pressure leads to an increase in venous return and ventricular preload, as well as effects on left ventricular wall stress that can lead to increases in afterload. This is often coupled with hypertension, as a patient’s sedation is reduced or stopped, which, in turn, can lead to both systolic and diastolic dysfunction. It is difficult to be certain what role pure cardiac dysfunction has on weaning failure, as it is difficult to separate the effects on increased cardiac loading from the increases in respiratory loading, but it is certainly a factor in over 40% of cases.

TTE is useful in identifying both systolic and diastolic dysfunction, which can predict the likelihood of successful weaning or liberation from mechanical ventilation and may therefore indicate the need for diuretics, beta-blockers, or angiotensin-converting enzyme inhibitors to optimize cardiac function. Left ventricular systolic function has been discussed in Chapter 7 and is a skill that can be acquired relatively easily. Assessment of left ventricular diastolic function is an advanced echocardiographic skill, which involves spectral and tissue Doppler examination of MV inflow and annulus, respectively. PW Doppler of MV inflow in diastole provides information on the E wave during early diastole and the A wave due to atrial systole in late ventricular diastole. Tissue Doppler analysis of MV annulus allows determination of the e′ wave of left ventricular relaxation. The E/e′ ratio can then be calculated, which correlates with the left ventricular filling pressure (see Chapter 7 for more detailed explanation).

Several studies have examined the echocardiographic assessment of ventricular function prior to extubation and its relationship to failure of weaning. The following are associated with a higher likelihood of failure of weaning:

  • EF of <40%.

  • E/A ratio of >2 in the presence of impaired left ventricular function.

  • E/e′ ratio of >12 in the presence of preserved left ventricular function.

Echocardiography undertaken during an SBT may demonstrate a cardiac cause for a failed trial of weaning that was not apparent when assessed while the patient was receiving ventilatory support due to the change in left ventricular loading conditions. Increases in E/A ratio and E/e′ ratio correlate with measurement of increased PAOP, with a reasonable sensitivity and specificity, where the E/A ratio is >0.95 and the E/e′ ratio >8.5.

Chapter 23



1. US examination has the following role in the assessment of weaning:

  1. A Lung aeration can be assessed with a 3-zone US assessment in each lung

  2. B The presence of a pleural effusion predicts SBT failure

  3. C An inverted diaphragm associated with a pleural effusion signifies a significant mechanical disadvantage

  4. D The number of B-lines noted in each lung zone correlates with the loss of aeration and severity of pulmonary oedema

  5. E An intra-pleural distance of >50 mm suggests an effusion of at least 500 mL

2. In the echocardiographic assessment of weaning failure, the following are true:

  1. A Preserved systolic function excludes a cardiac cause of weaning failure

  2. B Echocardiography undertaken during an SBT has increased sensitivity in identifying a cardiac cause for weaning failure

  3. C AV spectral and tissue Doppler is used to assess diastolic function

  4. D E/e′ correlates with the measured PAOP

  5. E E/e′ of >12 is associated with increased likelihood of SBT failure


1. US examination has the following role in the assessment of weaning:

  1. A FALSE. A 6-point assessment has been used.

  2. B FALSE. Very common and may not have a significant physiological effect if small.

  3. C TRUE. An inverted diaphragm indicates that the pleural pressure is significantly raised and that drainage should lead to improved pulmonary mechanics and gas exchange.

  4. D TRUE. The number of B-lines correlates with measurements of extravascular lung water and can be used to monitor therapy.

  5. E FALSE. It indicates an effusion of >800 mL.

2. In the echocardiographic assessment of weaning failure, the following are true:

  1. A FALSE. Diastolic dysfunction may be an important cause of weaning failure.

  2. B TRUE. Increased E/eat SBT has been shown to predict weaning failure.

  3. C FALSE. MV inflow and tissue Doppler of MV annulus is used.

  4. D TRUE. These PW and tissue Doppler variables require considerable echo experience to measure reliably.

  5. E TRUE. This indicates diastolic dysfunction.

Further reading

Goligher E, Leis J, Fowler R, Pinto R, Adhikari N, Ferguson N. Utility and safety of draining pleural effusions in mechanically ventilated patients: a systematic review and meta-analysis. Critical Care 2011;15:R46.Find this resource:

Ho CY, Solomon SD. A clinician’s guide to tissue Doppler imaging. Circulation 2006;113;10:e396–8.Find this resource:

Soummer A, Perbet S, Brisson H, et al. Ultrasound assessment of lung aeration loss during a successful weaning trial predicts postextubation distress. Critical Care Medicine 2012;40:2064–72.Find this resource:

Walden AP, Jones QC, Matsa R, Wise MP. Pleural effusions on the intensive care unit; hidden morbidity with therapeutic potential. Respirology 2013;18:246–54.Find this resource:

Copyright © 2022. All rights reserved.