• Consider oxygen therapy in people with chronic dyspnoea who need symptomatic treatment despite having tried simple measures, such as improving air circulation by using a fan.
• When selecting people for oxygen therapy, carefully consider the potential risks and benefits in each case.
• The person's need for oxygen therapy should be clinically assessed by the palliative care team in primary care if possible, taking into account, for example:
• The person's response to simple measures.
• If pulse oximetry indicates the person is hypoxic, or whether the person has to make a constant physical effort to avoid hypoxia.
• The underlying cause of the dyspnoea.
• The risks from oxygen use (e.g. does the person or any other member of the household smoke?).
• A trial of short-burst oxygen therapy may be initiated in primary care if considered appropriate following assessment (preferably after discussion with a specialist), and should be tailored to a person's needs.
• Long-term oxygen therapy should be initiated by a respiratory specialist.
• Oxygen therapy may also be appropriate in people with cancer who have dyspnoea of non-malignant causes. For more information, see Managing dyspnoea of known cause.
• For more information see the section on oxygen in Prescribing information.

• Breathlessness with and without oxygen therapy, at rest or on exertion, may be measured using:
• Exercise tests (e.g. in which people exercise at a fixed rate for as long as possible, until they are limited by symptoms).
• Objective measurement of oxygenation (e.g. pulse oximetry, in which a convenient, non-invasive device measures peripheral haemoglobin oxygen saturation by fluctuations of light absorption in well-vascularized tissue).
• However, pulse oximetry is not as accurate as blood gas measurement, so should be used as a guide, to monitor change in levels pre- and post-exertion, to ensure improvement with oxygen therapy, and to titrate dose of oxygen (e.g. to prevent under- or over-dosing).
• Assessment tools (e.g. visual analogue scale, on which people self-report symptom severity).
• Little evidence supports the use of short-burst oxygen therapy immediately before and after exercise.

[Booth and Dudgeon, 2006]

• However, expert opinion suggests that short-burst oxygen therapy may also be useful for pre-planned activity which triggers breathlessness (e.g. washing).
• Short-burst oxygen therapy should be initiated at 2 L/minute. Some experts suggest an initial duration of treatment of between 15 and 30 minutes, but other experts suggest continuing until the person feels benefit.
• Specialist initiation of long-term oxygen therapy (15 hours or more a day) can be considered in people with severe disabling breathlessness due to cancer or other progressive, life-threatening diseases [Twycross et al, 2002].

• Appropriate use of oxygen therapy is recommended for people with dyspnoea and cancer in palliative care guidelines [International Association for Hospice and Palliative Care, 2004; Lothian Palliative Care Guidelines Group, 2004b; NHMRC, 2004; Pan-Glasgow Palliative Care Algorithm Group, 2005; NCCN, 2006] and by published palliative care literature [Doyle et al, 2004; Regnard and Hockley, 2004b; Booth and Dudgeon, 2006]. Recommendations were also based on the expert opinion of palliative care specialists.
• When assessing people, focus on the relevance of oxygen therapy in the clinical context rather than on assessment tools used in research [Booth and Dudgeon, 2006].
• It is very difficult to select people who will benefit significantly from oxygen therapy:
• Oxygen does not relieve breathlessness in all people, and evidence indicates that the people who gain acute benefit from oxygen are not necessarily those who experience an increase in quality of life by using oxygen therapy at home [Booth and Dudgeon, 2006].
• CKS found no studies that identified which people are most likely to benefit significantly from short-burst oxygen therapy. Short-burst oxygen therapy can relieve symptoms in people with cancer and dyspnoea at rest (oxygen saturation =< 90%), although it is uncertain whether oxygen is better than air in people with advanced cancer [Doyle et al, 2004; Gallagher and Roberts, 2004].
• A recent randomized controlled trial (n = 51) investigated the effect of oxygen on dyspnoea in people with advanced cancer and found that symptoms improved with both air and oxygen, and that the treatments did not differ significantly [Philip et al, 2006]. This supports the suggestion that the sensation of airflow is an important determinant of benefit [Twycross et al, 2002].
• In people with oxygen saturation greater than 90%, the role of oxygen is controversial because response to oxygen therapy varies greatly [Twycross et al, 2002].
• Not all breathless people are hypoxaemic, and not all hypoxaemic people benefit from oxygen therapy. Therefore, a simple trial of oxygen is often the best guide to therapy [Booth and Wade, 2003], particularly when pulse oximetry is not available.