book cover

Palliative Care Perspectives : Chapter 5: Non-Pain Symptom Management : Dyspnea

Dyspnea is a common and distressing symptom.58 In the National Hospice Study 70% of patients experienced dyspnea in the final six weeks of life.59 Bruera and colleagues found that 55% of 135 patients with advanced cancer experienced moderate dyspnea.60 Not surprisingly, lung involvement with cancer correlated strongly with dyspnea in this study (as did anxiety and maximal inspiratory pressure). Chronic obstructive pulmonary disease (COPD), congestive heart failure, and fluid build-up in renal failure can also cause dyspnea. Because of the high prevalence of this symptom and the severity of suffering that can be associated with it, clinicians need to become familiar with available methods for the alleviation of dyspnea.61 Dyspnea is also not rare in a variety of nonterminal illnesses. Standard medical therapies often address this symptom and are "palliative" in that treatment is not curative of the underlying disease. Examples include most treatments for COPD and congestive heart failure. Therapies effective in standard treatment may also be effective at the end of life. However, certain important distinctions need to be made.

In standard medical therapy treatment for dyspnea is directed toward both alleviating dyspnea and prolonging the patient's life. This is entirely appropriate for many patients. An example would be the administration of ACE (angiotensin converting enzyme) inhibitors for congestive heart failure. These medications both relieve dyspnea by decreasing pulmonary edema through afterload reduction and can prolong life. ACE inhibitors may also be used in patients with congestive heart failure whose primary treatment goal is comfort. However, the life-prolonging aspect of the therapy should be carefully considered. For some, such modest life-prolonging effects may be entirely acceptable, even desirable. For others, at a certain stage even modest life-prolonging "side effects" of ACE inhibitors may be considered a burden. Alternate methods of alleviating dyspnea that do not have life-prolonging side effects should be considered.

Traditional medical therapy for dyspnea focuses almost exclusively on altering physiology in the lungs and elsewhere (heart and kidneys). Bronchodilators open small airways. Diuretics reduce fluid build-up. Antibiotics kill organisms that disrupt lung function. Rarely is the sensation or other psychic aspects of the experience directly addressed. As discussed in Chapter 3, in palliative care the totality of the patient's experience that results in suffering should be addressed.

Pathophysiology of Dyspnea

Dyspnea is the experience of shortness of breath. It may or may not be associated with suffering. A runner in a race may be severely dyspneic and yet be enjoying life. A dyspneic patient dying with lung cancer with very similar physiologic parameters may be suffering greatly, as the meaning and context of the dyspnea are entirely different.

Respiration as a bodily function is virtually unique in the degree to which it is under both reflexive and volitional control. The survival advantage of such a control is obvious. We continue to breath when deeply asleep (or in a coma), and yet we can hold our breath, if need be, to escape through smoke or water. Perhaps because of this, the linkage between psychological states and breathing is tighter and more complex than it is for other physical symptoms. Dyspnea is often associated with panic and anxiety; panic may present as dyspnea, and dyspnea may induce panic.

My education about dyspnea, brief as it was in medical school, stressed the importance of blood gases. I learned that CO2 build-up and oxygen deprivation were the critical factors that result in dyspnea. Although undoubtedly important in keeping the body alive, their importance in the experience of dyspnea has been exaggerated. If an oxygen saturation monitor is attached to a dyspneic runner at the end of a race, it will register normal. This reveals an important clinical pearl: oxygen saturation is insensitive in identifying patients with dyspnea. That is, one cannot rely on the oxygen saturation to tell who is dyspneic. Patients can be very dyspneic with normal saturations. Of course, patients with low oxygen saturations (below 90%) are far more likely to be dyspneic than are patients with normal saturations. However, oxygen saturation also lacks specificity as a predictor of dyspnea; many patients with low oxygen saturations (for example, patients with chronic lung disease or those who live at high altitude) will not be dyspneic, especially at rest. Studies have suggested that hypoxia correlates best with exertional dyspnea and poor exercise tolerance. Conversely, oxygen therapy has been shown to be a most helpful method to relieve exertional dyspnea and improve exercise tolerance. Studies have been mixed in testing the relief of rest dyspnea associated with hypoxia with oxygen therapy.62,63 Oxygen levels are excellent indicators of changing pulmonary physiology. The implications of the lack of sensitivity and specificity of oxygen saturation in identifying dyspnea are profound. As with pain, we lack a "scanner" for dyspnea that can reliably identify who is short of breath. We have no choice but to ask if dyspnea is present, or at least look for signs of distress that might suggest dyspnea, such as rapid respirations or a look of panic. In fact, the oxygen saturation meter can cause dyspnea by inducing panic and fear in patients, family, and clinicians as the saturation number falls (often accompanied by an ominously lower-pitched beeping tone).

Palliative Care Note
Oxygen saturation cannot be depended upon to identify dyspnea. Patient report or signs of agitation or anxiety are the best means of identifying dyspnea.

Elevations in carbon dioxide levels appear to stimulate dyspnea more than do low oxygen levels. Elevated partial pressure of arterial carbon dioxide (PaCO2) levels have been found to be an independent stimulus of dyspnea.64 However, increased respiratory drive does not necessarily result in dyspnea if it occurs unimpeded. Patients with certain forms of increased respiratory drive, such as diabetic ketoacidosis and pregnancy, may not experience dyspnea. What does cause dyspnea is an imbalance between the perceived need to breathe and the perceived ability to breathe. Elevated PaCO2 levels may be one among a number of factors that contribute to the brain's perception of a need to breathe.

Palliative Care Note
Dyspnea results when there is an imbalance between the perceived need to breathe and the perceived ability to breathe.

Recent studies have suggested that the body's ability to determine whether breathing is occurring normally relies on considerably more than high CO2 levels or low oxygen levels.64 Nerves in the nose sense the passage of air. Stretch receptors in and about the lungs signal expansion and contraction of the lungs. Thoracic muscles and ribs signal that they are moving, a good sign breathing is occurring normally. These nerves tell the brain "all is well" and allow respirations to continue automatically and largely unconsciously in the nondyspneic person. The brain senses no imbalance between the ability to breathe and the need to breathe. If there is a sudden cessation of airflow or respiratory muscle movement, as measured by these nerves, the brain quickly goes into alarm mode - before any measurable change in blood gases occurs.

Imagine you are being held one foot below water. Go ahead - hold your breath. Note how quickly there is a desire to breathe. You can suppress this need for a while, although it will build rapidly. (Your oxygen saturation may fall slightly and your CO2 saturation may rise slightly by the time you must breathe, contributing a bit to the need to breathe.) If you were truly underwater, you would panic almost immediately. This psychological state would result in a severe imbalance between the perceived need to breathe and the perceived ability to breathe. You would desperately struggle to get to the surface. When you finally breathe again, notice how quickly your dyspnea is relieved. Do you really believe it was because your blood gases were so quickly normalized?

Patients more commonly are dyspneic while breathing. What is happening here? The perception of respiratory fatigue is a key component. Although our understanding of both peripheral and central receptor involvement in fatigue is poor, it is clear that the brain is able to sense fatigue, much as you can sense any overworked muscle. However, unlike wobbly leg muscles that signal you to stop running, tired thoracic muscles must continue to work in an attempt to meet a perceived need to breathe. The brain senses a mismatch: the need to breathe continues, but the ability of the body to meet that need with tired muscles is in doubt. Dyspnea is a wake-up call to this mismatch.

Short of altering lung physiology, how can this mismatch be addressed? Decreasing the perceived need through energy conservation can help. Patients with emphysema, who have minimal thoracic expansion, may ventilate adequately but receive a signal from the thoracic muscles that they are not moving enough, much as in the underwater example above. A number of studies have demonstrated that fooling these muscles by activating vibrating devices during inspiration can lessen dyspnea.65-67 The perception that the body is not able to meet this need can also be addressed by inhibiting the perception of muscular fatigue with opioids and other drugs (see below).

As should have been apparent in the breath-holding experiment, perceived need is also intimately connected to the psyche. Perceived need and perceived inability to meet that need can quickly result in panic. Panic, in turn, can stimulate increased ventilatory effort, which can result in more fatigue and increased panic - a vicious cycle. Panic, fear, and anxiety are common affective components of dyspnea. Beyond shear panic, dyspneic patients think about what their dyspnea means. Does dyspnea reflect a good run, as it might to a jogger, or impending death? If dyspnea occurs with increasingly mild exertion, the patient is reminded of growing dependence on others. As with other symptoms, patients monitor the trend of their dyspnea - is it getting better or worse? One reason the runner does not suffer from dyspnea is that he or she knows it will end when the running ends. The patient dying of COPD or lung cancer projects into a future wherein dyspnea worsens. These cognitive processes commonly trigger affective responses. In addition to panic, fear, and anxiety, patients may become depressed and angry.

Dyspnea Treatment Principles

Here I address only those therapies not commonly employed in the traditional treatment of dyspnea. Traditional palliative measures focus on altering lung physiology by using beta agonists, diuretics, or tapping pleural effusions. Understanding the physiology of dyspnea as presented above suggests that when altering lung physiology fails, altering the perceived need and perceived ability to meet that need should improve the sensation of dyspnea. Addressing affective and cognitive components of dyspnea should alleviate its psychic components.

Opioids and benzodiazepines are the mainstays of palliative therapy for dyspnea. In choosing between starting an opioid or a benzodiazepine treatment for a dyspneic patient, the patient may give you a hint as to which is likely to be more effective. A patient who complains of hard work breathing but who lacks associated anxiety is more likely to benefit from an opioid. A patient who complains of anxiety associated with breathing is usually better first treated with a benzodiazepine. Having said this, often both are tried to determine which drug class is better for the patient. Prediction based on this rule of thumb is imperfect. Many patients will need both classes of medication and may be able to state clear preferences for "breakthrough" doses for exacerbations of dyspnea.


Opioids are very effective in relieving dyspnea, although the exact mechanism is not understood. Contrary to common belief, this effect does not result through inhibition of respiratory drive. Relief from the "work of breathing" is a function of steady-state opioid levels, much like steady-state opioid levels relieve pain. Inhibition of respiratory drive results primarily from rising opioid serum levels. Studies have demonstrated significant relief of dyspnea from opioids without significant effects on ventilation or pCO2 levels in common therapeutic doses.68,69 Having said this, patients with dyspnea are fragile. Respiratory drive suppression can occur if serum opioid levels rise rapidly. Thus, when initiating therapy with opioids for dyspnea, one should start with a low dose and raise the dose slowly as needed.

Morphine is the best studied of the opioids for relief of dyspnea, although relief has been observed with other agents, such as oxycodone, fentanyl, and methadone. There is no demonstrated advantage of one opioid over another. Generally, a lower dose of opioid is required to relieve dyspnea than is needed to relieve pain.

Nebulized morphine has been used by some to treat dyspnea, although this is not an FDA-approved route of administration in the United States.70 Mu receptors, to which morphine binds, have been identified in the lung, and it has been theorized that binding of these peripheral receptors may relieve dyspnea at lower serum levels of morphine than when it is given via other routes. Studies have demonstrated that aerosolized morphine is effective in the relief of dyspnea, although no clear advantage of this route has yet been proven other than its rapid onset of action. Few bioavailability studies have been done. From 5% to 100% bioavailability has been reported.71 I believe it safest to assume 100% bioavailability, at least with initial dosing. One advantage of the aerosolized route is that peak serum levels occur rapidly - roughly equal to the time it takes to deliver the aerosol. This can be an advantage over oral dosing (peak effect in one hour) and parenteral administration, which may not be feasible in certain settings, such as in the home.

Morphine can cause histamine release, thereby inducing bronchospasm when given by aerosol. (See chapter 4, section on morphine.) It may be wise to give a trial dose under observation and to watch for bronchospasm. It seems advisable to ensure that patients are able to tolerate nonaerosolized morphine with no histamine release before attempting aerosol therapy. There has been some concern that the preservatives for regular IV morphine may trigger bronchospasm, although, to my knowledge, this is only theoretical. Thus, some authors have recommended using preservative-free morphine when using aerosol. I have successfully used injectable (not oral solution) morphine in a number of patients with no evidence of adverse effect.

In my practice I have used this route primarily when a patient is already receiving other nebulized medications to which morphine can be added or when rapid, nonparenteral (IV, SC) dosing is desired. Responses appears to be somewhat idiosyncratic. Some patients love it, while others are unimpressed. I recommend starting with a low dose, 2-4 mgs. I have not read of experiences with other opioids administered via this route.


Benzodiazepines are very helpful in relieving panic and anxiety associated with dyspnea, if present, but are not helpful if panic and anxiety are absent.6 Lorazepam is most commonly used in low doses. Longer-acting benzodiazepines may also be used for chronic dyspnea. This effect also does not appear to be directly related to suppression of respiratory drive. As with opioids, respiratory drive can be suppressed by benzodiazepines if they are escalated rapidly and given in high doses. Thus, rapid escalation of benzodiazepines, especially via the IV route, should be avoided. Start low and go slow.

Other agents

Identifying an agent that reliably alleviates dyspnea without risk of respiratory drive suppression would be the palliative care equivalent of finding the Holy Grail. So far, efforts have largely failed. It was hoped that local anesthetics, such as lidocaine, would numb stretch receptors and in so doing alleviate dyspnea, but this has not been found to be the case.72 Further research in this area is needed.

Oxygen may relieve dyspnea, especially when significant hypoxia (O2 sat <90) is present during exercise. In addition, oxygen may relieve dyspnea via other mechanisms. Airway resistance may be decreased when oxygen is administered, thereby reducing the work of breathing.73 The flow of gas over the nasal mucosa may itself provide a dampening effect on the perception of dyspnea. Liss and Grant randomized patients to receive oxygen or air via nasal prongs at 2 or 4 liters. Dyspnea was equally relieved by both therapies. Dyspnea increased when the nasal mucosa was anesthetized.74 This study can be criticized for having treated patients with relatively high mean pO2 levels (67 mm Hg) and for having treated only rest dyspnea.62 In a later double-blind cross-over study by Bruera and colleagues, 14 hypoxemic cancer patients were randomized to receive oxygen or air by mask. Twelve of fourteen patients consistently preferred oxygen. Visual analogue scale reports of dyspnea were significantly less on oxygen compared to air.63 While oxygen therefore may be very helpful, patients who are confused or distressed may not tolerate oxygen administration, especially via mask, as masks may be constricting. Such patients often respond better to a gentle fan or cool breeze.

<<< Previous Next >>> [ Go Up ]

book cover

Palliative Care Perspectives

James L. Hallenbeck, M.D.

Order hardback from
Order paperback from

Copyright 2003 by Oxford University Press, Inc.

The online version of this book is used with permission of the publisher and author on web sites affiliated with the Inter-Institutional Collaborating Network on End-of-life Care (IICN), sponsored by Growth House, Inc.