The purpose of this study is to determine the effect that oxygen has when administered to patients complaining of shortness of breath, where the underlying cause of this symptom is advanced cancer. The study tests the hypothesis that oxygen improves shortness of breath more than air in this population. Both oxygen and air will be administered to patients in random order and in a blinded fashion, with patients asked to rate their shortness of breath before and after each gas. Finally patients will be asked which gas they prefer.

Dyspnoea is a devastating symptom in patients with advanced cancer. Management strategies are limited and include behavioural therapies such as relaxation, and pharmacological therapies such as opioids and anxiolytics. The latter are associated with problematic side effects in many patients. Inhalational oxygen is frequently administered but there are few studies in this population which define its role and benefits.

Main Aim: To compare patient preference for inhalational oxygen versus air for relief of dyspnoea.

Specific aims:

1. To compare the patient preference for inhalational oxygen versus compressed air in the relief of dyspnoea.
2. To compare the response to oxygen and air in improvement of dyspnoea in patients with advanced cancer.
3. To compare the response to oxygen and air in improvement of dyspnoea in those patients with advanced cancer with documented hypoxia.
4. To identify factors other than hypoxia which impact on the sensation of dyspnoea and its relief, when patients are administered oxygen and air.

Hypotheses

1. Oxygen improves dyspnoea in patients with cancer more than compressed air.
2. Patients with cancer who are hypoxic are more likely than those who are not hypoxic to have improvement of dyspnoea with oxygen administration.
3. The cause of the dyspnoea may affect whether dyspnoea improves more with oxygen than with air.

Using a randomised, double blind, crossover study design, patients will be adminstered air and oxygen for 15 minutes and be asked to rate dyspnoea scores before and after each gas. Measures of oxygen saturation will be simultaneously measured, and finally the patient preferences for the gases will be sought at trial completion.

Inclusion Criteria:

• Patients who have dyspnoea mainly due to advanced cancer. Patients with a history of COAD will be eligible for participation in this study as long as the main mechanism of current dyspnoea is related to tumor.
• intensity of dyspnoea of at least 3 on a 0-10 visual analogue scale at the time of treatment.
• Regular Bronchodilators and corticosteroids and other adjuvant medications for dyspnoea will be allowed to continue during the study. Inhaled bronchodilator steriods may not be used during the study period.
• Patients may be receiving regular oral or parenteral opioids and opioid dose must be stable for 24 hours.
• Patients must have normal cognitive status defined as normal state of arousal and absence of obvious clinical findings of confusion, memory or concentration deficit according to Blessed Orientation Memory & Concentration mental status examination (score<10).
• Patients must be 18 years of age or older.
• Patients must have no contraindications to oxygen.
• Patients must sign written informed consent.

Exclusion Criteria:

• Patients who have evidence of acute respiratory distress.
• Patients who are currently oxygen dependent
• Patients who refuse to participate or are deemed incapable of completing the research.