Abernethy, A.P., McDonald, C.F., Frith, P.A., Clark, K., Herndon, J.E., 2nd, Marcello, J., . . . Currow, D.C. (2010). Effect of palliative oxygen versus room air in relief of breathlessness in patients with refractory dyspnoea: A double-blind, randomised controlled trial. Lancet, 376(9743), 784-793.

The objective of the study is to compare the symptomatic effectiveness of palliative oxygen against room air in providing relief for patients with life-limiting illness, refractory breathlessness, and PaO₂ greater than 7.3 kPa (54.75 mmHg).

Eligible participants underwent arterial blood gas assessment in an outpatient clinic or at home. Those with PaO₂ greater than 7.3 kPa were assigned in a 1:1 ratio to receive oxygen or room air delivery by a concentrator and nasal cannula. Medical gas was administered continuously at 2 liters per minute via nasal cannula, and participants were instructed to use concentrators for at least 15 hours per day and record the amount of “breathlessness right now” twice a day (within 30 minutes of waking up in the morning and going to bed in the evening). Diaries also captured secondary outcomes, including average dyspnea in the previous 24 hours, worst breathlessness in the previous 24 hours, relief of dyspnea during the previous 24 hours, and ordered categorical scales for functional impact, sleep disturbance, drowsiness, anxiety, nasal irritation, and nose bleeds. Quality of life also was assessed every day, as well as functional changes .

The study reported on a sample of 239 patients; 120 were in the oxygen group, 119 were in the room air group, and 13 withdrew from the study before it started and any data were collected.

The sample was 62% male and 38% female.

Key disease characteristics were chronic obstructive pulmonary disease (COPD) (152), restrictive lung disease (14), bronchiectasis (7), primary pulmonary hypertension (3), primary lung cancer (33), known secondary lung cancer (5), pleural effusion (2), end-stage cardiomyopathy (7), and other (16).

Patients were excluded if they met international eligibility guidelines for long-term oxygen therapy, had a history of hypercarbic respiratory failure with oxygen, had anemia (Hgb less than 100 g/L ), were hypercarbic, or had cognitive impairment.

The median age for the oxygen group was 73 years, and the median age for the room air group was 74 years.
 

The study was conducted in a multi-site, outpatient setting at outpatient pulmonary, palliative care, oncology, and primary clinics at five sites in Australia, two sites in the United States, and two sites in the United Kingdom.

• Patients were undergoing end-of-life care.
• The study has clinical applicability for end-of-life and palliative care.
   

The study was a double-blinded, randomized, controlled trial.

• Folstein mini-mental status examination score to assess for cognitive impairment during screening process, prior to study commencement
• Eastern Cooperative Oncology Group (ECOG) performance status scale
• Numerical Rating scale (NRS) from 0-10 adjusted with 1-point reduction in self-reported dyspnea
• McGill Quality of Life Questionnaire (MQoLQ)
• Participant diary entries capturing secondary outcomes (average dyspnea in the previous 24 hours, worst breathlessness in the previous 24 hours, relief of dyspnea during the previous 24 hours, sleep disturbance, drowsiness, anxiety, nasal irritation, and nosebleeds) using ordered categorical scales
• Medical Research Council (MRC) 4-point categorical dyspnea functional scale to assess functional changes and a 4-point categorical dyspnea exertion scale measured daily
• 5-point Likert-type categorical scale to measure side-effects
   

Thirteen patients (5%) withdrew before the study started, and no assessments were completed. Fifteen patients (6%) withdrew before the day six assessment and completed data. Longitudinal analyses measuring the clinical effect of the interventions found significant improvement in morning and evening dyspnea in both oxygen and room air groups (time p < 0.0001), but the primary outcome of breathlessness did not differ between groups at any time during the study period. Quality-of-life measures were similar between groups. Reports of participants’ worst level of functioning on the MRC dyspnea scale and sleep disruption from breathlessness decreased during the seven-day study with little difference between groups. Baseline dyspnea also seemed to predict evening response, regardless of intervention.

Based on study results, palliative oxygen does not provide benefit over room air for relief of breathlessness in patients who were not hypercarbic.

Exact times of morning and evening assessments and times during which participants used prescribed gases were not recorded. Due to the heterogeneous nature of the patient population, assessing which patient subgroup (e.g., patients with COPD versus patients with cancer) may have experienced better symptomatic relief from palliative oxygen is difficult. Because most participants had an ECOG performance status of 2 or 3 with no breathlessness at rest, the patient population may not be representative of the sickest patients in palliative care who would frequently receive palliative oxygen. Also, more randomized participants withdrew from the room air group than the oxygen group, causing potentially skewed results. The authors also question the clinical significance of demonstrated benefit, when considering their means of defining symptomatic relief (i.e., NRS with 1-point change), and note the possibility that secondary analyses might be underpowered. Similarly, objective measures of dyspnea (oxygen saturation, hemodynamics, and sleep) had not been recorded for comparison to subjective results of the study. Gauging participant compliance with instructed use of interventions for the prescribed 15 hours per day also was difficult. While authors note a slightly lower usage (14 hours per day), most responses occurred within the first 24 hours, and they note the unlikelihood that stricter adherence would change outcomes. Finally, regardless of baseline NRS, all participants were relegated to 2 liters per minute of either oxygen or room air by nasal cannula, thus calling into consideration the benefits of titrating oxygen to higher delivery levels. 

Prescription of palliative oxygen therapy may not be a cost-effective and scientifically based clinical intervention for the relief of breathlessness. Oxygen is expensive, flammable, and should be monitored carefully when prescribed for patients with potentially hypercarbic states and central hypoventilation syndromes.