Special Feature

Trouble-Shooting the Ventilated Patient

By David J. Pierson, MD, Editor, Professor, Pulmonary and Critical Care Medicine, Harborview Medical Center, University of Washington, Seattle, is Editor for Critical Care Alert.

Late in the evening you receive a page about a patient who is "fighting the ventilator." You are cross-covering and do not know the patient. Your sign-out sheet says only that she is 68 years old, has severe COPD complicated by pneumonia, and has been in the ICU for the last 3 days. No specific problems are mentioned. The nurse tells you on the telephone that the patient was "fine" until about an hour ago, but has since become increasingly agitated. The ventilator’s high-pressure alarm has been sounding frequently, and now the pulse oximeter’s saturation readings have fallen from the mid-90s to the high-80s. What should you do?

This hypothetical scenario is familiar to everyone who cares for mechanically ventilated patients in the ICU. However, how to manage it most effectively, efficiently, and safely is not something that can be guided by high-level evidence from randomized clinical trials.1-3 Nonetheless, a logical approach, an appreciation of the physiology involved, and an understanding of some of the basics of ventilatory support can help to demystify what can be not only an anxiety-provoking situation for the clinician but also a serious threat to the patient. This article discusses some of the potential causes for acute respiratory distress that develops in a patient who was previously tolerating mechanical ventilation. Based on these causes, it then presents a straightforward, step-by-step approach for approaching the problem and its correction.

Possible Causes: A Change in the Ventilator or Other Apparatus

Table 1 lists some of the more common processes that can cause a previously stable patient to start fighting the ventilator. The first thing that the clinician at the bedside should consider is some malfunction in the ventilator or circuit. Such malfunctions may either prevent the intended tidal volume from reaching the patient or cause distress by increasing the patient’s work of breathing.4-6 Although it is uncommon, ventilators do occasionally fail. Sometimes the ventilator circuit becomes disconnected from the endotracheal or tracheostomy tube, and falls on the bedding in such as way that the low-pressure alarm is not triggered. The circuit can also be interrupted where it connects to a filter, nebulizer, temperature probe, or humidifier, or at the connections to the ventilator itself.

Apparatus includes the artificial airway, and this may also be the source of patient distress. Malposition or kinking of the tube can increase inspiratory resistance, decreasing the delivered tidal volume, or cause an inspiratory leak such that the delivered volume does not reach the patient’s lungs. Development of a leak in the cuff, or in the pilot balloon or its connecting tube, is another common cause of a sudden decrease in delivered tidal volume.

While these equipment-related problems tend to be easy to identify, picking up the fact that something in the circuit has increased the patient’s work of breathing can be more difficult. The increase in breathing effort can originate in either the inspiratory or the expiratory limb of the circuit. Simple observation of the patient’s breathing pattern can help here, although this is not always the case. If the agitation and distress began when the ventilator or circuit was changed, inappropriate assembly or malfunction of a component is suggested. However, these problems can develop more slowly, and valves may suddenly become dysfunctional during use.

Possible Causes: A Change in the Patient

Worsening of the Primary Process

A common cause for increasing intolerance of ventilatory support is worsening of the primary process that caused the patient to require intubation and mechanical ventilation. Table 1 lists what I consider to be the most common primary causes of acute respiratory failure that tend to do this. Exacerbations of COPD or asthma may worsen after admission, particularly if treatment with inhaled bronchodilators and systemic corticosteroids has been delayed or is suboptimally aggressive. Hypoxemic respiratory failure caused by community-acquired pneumonia can worsen in the initial 24 hours, particularly if the number of lobes involved turns out to be more than initially appreciated or if the patient was dehydrated on admission. Acute lung injury and the acute respiratory distress syndrome (ARDS) frequently worsen during the first 2 or 3 days after criteria for these diagnoses are first met.

There is also a difference between sudden development and sudden discovery. Sometimes a patient is perceived to have developed acute distress when in fact the worsening has been progressive but this has not been fully appreciated. Whatever acute illness caused the patient to require ventilatory support in the first place may have been more severe than initially appreciated, or it may have progressed unnoticed since admission. This is more likely to be the cause for fighting the ventilator that is noted within the first day or two after intubation. In this instance the problem may simply be that the initial ventilator settings are no longer adequate in the face of increased demands for oxygenation, ventilation, or mechanical support.

A New Pulmonary Process

Barotrauma is always a possibility during invasive mechanical ventilation, and it may not become clinically evident for many hours after intubation. Pneumothorax, the most potentially lethal of the forms of extra-alveolar air encountered during ventilatory support, can quickly become physiologically important, so having this complication first on the list of possibilities is appropriate. Another, more subtle complication of positive-pressure ventilation is dynamic hyperinflation and auto-PEEP.7,8 Auto-PEEP is especially common in patients with obstructive lung disease, but it may occur in others as well and should always be thought of when patients appear to be struggling to breathe. Circumstances in which dynamic hyperinflation and auto-PEEP are especially likely are listed in Table 2.9 The use of ventilator graphics monitoring can be helpful in identifying the presence of auto-PEEP, particularly when the patient is triggering the ventilator, preventing assessment by end-expiratory airway occlusion.10,11

The appearance or worsening of wheezing should prompt consideration of bronchospasm as a potential cause for fighting the ventilator. However, wheezing is a sign of airway obstruction, not necessarily of bronchospasm. It can be produced by secretions or other causes of airway obstruction, and is likely to worsen whenever there is increased expiratory effort. Pulmonary thromboembolism should always be on the list of possible causes for new hypoxemia or other deterioration in a ventilated patient. However, other causes are much more common, and a computed tomographic angiogram should seldom be the first procedure undertaken in evaluating the problem. Thromboembolism is more likely if the patient has obvious predisposing conditions such as a history of thrombosis, vascular injury, or malignancy, or if deep-venous thrombosis prophylaxis has not been given, and also if a search for other likely explanations for the respiratory distress is negative.

After the first 2 or 3 days, the likelihood of developing ventilator-associated pneumonia increases the longer a patient has been intubated. This serious complication is suggested when respiratory secretions increase in quantity or become more purulent, when the patient develops fever and/or new leukocytosis, and there is a new and persistent area of opacification on the chest radiograph. Finally, patient agitation is a frequent finding when acute lung injury or ARDS develops during the course of mechanical ventilation. Even when the risk factor associated with ARDS was present on admission, meeting the criteria for the syndrome often takes 2 or 3 days, or even longer.

A Non-Pulmonary Problem

Although manifested by respiratory distress, fighting the ventilator may be caused by a new, non-pulmonary problem, as listed in Table 1. The latter includes changes in volume status. Volume depletion in a patient receiving positive-pressure ventilation most often causes tachycardia and hypotension. However, central hypovolemia can be subtle, particularly when high levels of PEEP are used, and it should be considered in the assessment of newly-recognized agitation and respiratory distress. More commonly, respiratory deterioration is associated with volume overload. Pulmonary congestion and early pulmonary edema can cause tachypnea, air hunger, and increased oxygen requirements, which may be interpreted as fighting the ventilator. Hypervolemia is particularly a threat in patients who initially require volume resuscitation and subsequently have infusions of isotonic fluid continued for one or more days thereafter. Volume overload may be associated with wheezing and the development of auto-PEEP as well as with the other signs mentioned.

Cardiac ischemia or myocardial infarction should be considered, particularly in patients with known heart disease. An intracranial hemorrhage or other cerebrovascular event may announce itself by a change in breathing pattern or respiratory rate, and decreased responsiveness or other new neurological findings may initially be overlooked. The onset of sepsis may produce restlessness and tachypnea before other signs appear.

Drugs can cause patients to fight the ventilator in 2 general ways. The first of these is an adverse drug reaction. Many drugs administered to mechanically ventilated patients can cause delirium, hallucinations, and other alterations in perception, and these things are often manifested by agitation and distress. Other drugs such as central-nervous system stimulants and methylxanthines increase ventilatory drive and the sensation of air hunger. Recovery from the effects of muscle relaxants may be delayed, leading to respiratory distress in patients receiving only partial ventilatory support. Corticosteroids, especially when given in conjunction with muscle relaxants or for prolonged periods, are associated with muscle weakness and inability to wean from ventilatory support. Less often, idiosyncratic reactions to aminoglycosides and some other drugs can produce ventilatory muscle weakness.

The second general mechanism by which drugs can contribute to fighting the ventilator is withdrawal. Agitation and respiratory distress developing 2 to 5 days after admission may be caused by withdrawal from alcohol, opioids, or other agents the patient may have been taking regularly prior to hospitalization. In patients with prolonged critical illness, withdrawal from drugs administered in the hospital should also be considered a possibility.

Possible Causes: Inappropriate Ventilator Settings

The ventilator settings chosen by the clinician may not match the patient’s desired breathing pattern, although this is typically evident from the beginning. However, patients who are obtunded or heavily sedated when admitted to the ICU may become more alert and first show signs of respiratory distress hours later. In addition, ventilator settings may be altered either inadvertently or during procedures, causing subsequent patient distress.

Table 1 lists several of the ventilator settings that may affect a patient’s perception of the ease or adequacy of breathing. The use of lung-protective ventilation in the management of acute lung injury or ARDS12 involves tidal volumes much smaller than what many patients want.13 Because the proven mortality benefit of lung-protective ventilation14,15 is believed to be related to both tidal volume and airway pressure, respiratory distress associated with this approach must be managed by means other than increasing ventilation volumes and pressures. Patients in respiratory distress are tachypneic, and lung-protective ventilation allows the rate to be increased, but only up to a point. Respiratory rates above 35 breaths/min tend to be associated with increasing auto-PEEP, which should be avoided.12

Triggering sensitivity should be maintained at 1.0 to 1.5 cm H2O. When the threshold for initiation of inspiratory flow exceeds this range effort increases, and this is uncomfortable for most patients. For patients who are awake and acutely aware of their breathing, the inspiratory flows commonly used in ventilatory support may be inadequate. Although peak inspiratory flow during quiet normal breathing is about 50 L/min, normal inspiratory flows during activity are typically 2 to 4 times greater than this, and peak inspiratory flow in healthy individuals exceeds 400-500 L/min. Thus, dyspneic patients may be further distressed if inspiratory flow is less than 70-80 L/min. One potential advantage of pressure-targeted over volume-targeted ventilation is the higher peak inspiratory flows achievable with the former—as high as 200 L/min in some instances.

Possible Causes: Inadequate Sedation

The need for administration of more sedatives is an important cause of respiratory distress and patient-ventilator dyssynchrony. As mentioned, the low tidal volumes associated with lung-protective ventilation may require more sedation in some patients than would otherwise be the case. However, to assume that fighting the ventilator is simply a sign of the need for increased doses of sedatives, particularly when respiratory distress develops in a patient who previously tolerated ventilatory support, is a potentially dangerous mistake. Such development should be approached as a diagnostic problem as well as a management challenge. Inadequate sedation and ICU delirium should always be diagnoses of exclusion, and the processes discussed above should be considered and excluded. A patient should never be paralyzed just to achieve synchrony with the ventilator; such an approach does nothing either to clarify the cause of the distress or to relieve it.

An Approach to Ventilator Trouble-Shooting

Given the possible causes for fighting the ventilator and the potential seriousness of several of them, it is important to approach the problem in a logical and systematic fashion. Table 3 outlines such an approach.

To exclude equipment malfunction and to assure that the artificial airway is patent, the patient should first be disconnected from the circuit and manually ventilated with supplemental oxygen. This maneuver may instantly relieve the patient’s distress and reveal its likely cause. Whether the patient is easy or difficult to hand-ventilate is also an important observation. Especially if inspiratory obstruction is suggested during manual ventilation, the airway should next be suctioned, both to assure its patency and to assess any change in secretions. The respiratory therapist should then perform a quick inspection of the ventilator and its circuit to make sure that there are no obvious failures or defects.

Once these initial, potentially life-saving maneuvers have been completed, the responding clinician should quickly gather information available at the bedside that may point to the cause of the problem. Taking report from the patient’s nurse and respiratory therapist may provide clear indications of the nature of the problem. Was the onset of respiratory distress abrupt or gradual? What happened first? Is it intermittent? The clinical setting for ventilatory support is also important. Is this routine ventilation after surgery, or is the patient ventilated because of primary respiratory failure due to pneumonia, obstructive lung disease, or ARDS? Does the patient have severe underlying pulmonary or cardiac disease? In addition, it is important to know whether the patient is receiving full or partial ventilatory support, and volume- vs pressure-targeted ventilation.9

The clinician should then undertake a brief physical examination, targeted at the manifestations of respiratory distress and their most likely causes. The amount of inspiratory effort being expended by the patient with every breath should specifically be assessed, as should the symmetry of chest expansion, air movement, and lung sounds. To the visual and auscultatory impression of breathing effort can be added evidence from ventilator graphics, if this type of monitoring display is available.16 Other recent monitoring data should be examined next, looking for trends in vital signs, inspiratory and expiratory airway pressures, and minute ventilation over the last few hours. The respiratory therapists at my institution monitor every ventilated patient for auto-PEEP at least once each shift, and this important assessment should be done if not already recorded.

Pertinent recent laboratory data—especially oxygenation and acid-base status as revealed by arterial blood gases, and complete blood count to look for signs of new infection or blood loss—should next be examined. To complete the initial assessment, the patient’s most recent chest X-ray should be reviewed, and a new one obtained at the bedside if none has been done recently or the cause for the problem is still not apparent. This last step is especially important, as most of the potentially life-threatening problems in Table 1 are identified or at least suggested by radiographic findings.

As the clinician proceeds through the assessment algorithm, the likely cause of the patient’s acute distress should become more apparent—or, at least, the most life-threatening possibilities should be able to be excluded. At this point, the patient’s level of sedation can be increased with less likelihood of overlooking or obscuring important findings. It may also be justifiable, if the preceding measures have not revealed the reason for distress or relieved it, to try modifying the mode, inspiratory flow, or other ventilator settings in an attempt to achieve better patient comfort.

Two words of caution are in order here, however. First, the volume and pressure targets of lung-protective ventilation should not be altered if the patient has acute lung injury or ARDS, as these have been proven to save lives.14,15 And, finally, current ICU ventilators have new mode combinations and variations that are unfamiliar to many clinicians, and empirically switching to some unfamiliar ventilatory approach should be avoided unless everyone who will be involved in the patient’s management has been thoroughly instructed in its use and potential complications.


  1. Tobin MJ. Respir Care. 1992;37:1081-1096.
  2. Keith RL, Pierson DJ. Clin Chest Med. 1996;17:439-451.
  3. Hess DR, Thompson BT. Crit Care Med. 2006;34:231-233.
  4. Austin PN, et al. Respir Care. 2002;47:667-674.
  5. Younes M, et al. Am J Respir Crit Care Med. 2002;166:21-30.
  6. Racca F, et al. Respir Care Clin N Am. 2005;11:225-245.
  7. Brochard L. Intensive Care Med. 2002;28:1552-1554.
  8. Ranieri VM, et al. Clin Chest Med. 1996;17:379-394.
  9. Pierson DJ. Invasive mechanical ventilation. In: Albert RK, Spiro SG, Jett JR, eds. Clinical Respiratory Medicine. London/Philadelphia, Elsevier Health Sciences, 2nd edition, 2004:189-209.
  10. Blanch L, et al. Respir Care. 2005;50:110-123.
  11. Dhand R. Respir Care. 2005;50:246-261.
  12. Kallet RH, et al. Respir Care. 2001;46:1024-1037.
  13. Kallet RH, Luce JM. Respir Care. 2002;47:183-185.
  14. The Acute Respiratory Distress Syndrome Network. N Engl J Med. 2000;342:1301-1308.
  15. Kallet RH, et al. Crit Care Med. 2005;33:925-929.
  16. Nilsestuen JO, Hargett KD. Respir Care. 2005;50:202-234.