The benefits of supplemental oxygen use in patients with severe hypoxemia are well established and include increased survival rates,1 decrease in hospitalizations2 improved neuropsychological functions,3 exercise capacity4 and health related quality of life (QoL).5 The benefits were has also been found to be in proportion to the number of hours of daily use3 which justifies the prescription of continuous oxygen therapy. However, the use of oxygen supplementation for such long periods is very demanding, since patients have to be connected to an oxygen supply for all or most of the day, inevitably interfering with their activities, and forcing them to stay at home.

Ambulatory oxygen (AO) consists of oxygen supplementation to patients during exercise and daily activities using a portable device, which means that they are no longer housebound.

Standards of care and guidelines have established criteria for the assessment and prescription of oxygen. In particular, AO therapy prescription is recommended for active patients who are on long-term oxygen therapy (LTOT) for extensive periods and for those that although not eligible for LTOT, desaturate with exercise and show an acute response to oxygen.6,7 In Portugal the existing guidelines (Circular Normativa da Direcção Geral de Saúde N°: 06/DSPCS de 07/06/2006) define the criteria for LTOT prescription- exercise desaturation is one of them, but the specific indications of when to use AO are not explicit. Despite the benefits of oxygen use in enhancing activity8 and improving their QoL, patients have shown low tolerance to its use and are poorly compliant.9 In order to assess the appropriateness of AO therapy, it is important to understand whether patients are in fact using and benefiting from it. The purpose of this study was to determine patterns of AO use among patients with chronic pulmonary disease (COPD) and interstitial lung diseases (ILD) and analyze the effects of this therapy on daily activities and QoL.

The aim of this study was to assess the efficacy and patterns of AO usage by liquid oxygen in patients with chronic respiratory diseases independently of LTOT therapy prescription. LTOT benefits have been proven mainly in studies of COPD patients,1,3 with the results being extrapolated to hypoxemic patients with other diagnosis.13 Consequently oxygen therapy became routine treatment for all severely hypoxemic patients. The same applied to patients not fulfilling LTOT criteria but desaturating with exercise, when it comes to AO. One study found out that improvement in exercise ability and reduction in breathlessness with AO could be achieved both in COPD and ILD patients.14 Our patients therefore included several diagnoses which were grouped according to their ventilatory pattern.

Although exercise capacity is a good prognosis factor for survival in COPD patients,15 we cannot expect survival benefits of AO in the context of current knowledge because studies assessing long-term benefits of AO do not allow us to draw any firm conclusions concerning the effectiveness of AO.16

When LTOT is prescribed, it is intended to be used for long periods and is traditionally provided by 2 types of large stationary systems: a gaseous-oxygen cylinder or an oxygen concentrator. To facilitate movement in and about the home, patients can be connected to their stationary system by up to 15 meters of supply tubing.17 But to ambulate beyond that distance, either within the home or out into the community, a portable oxygen device will be required. The major goals of AO are thus to enable the patients to leave their homes for a longer period of time, to improve health status, and to increase daily activities.

AO can also be provided by 2 ways: small gaseous-oxygen cylinders that can be pulled on a cart, but have limited oxygen capacity18; or a liquid oxygen system, which was used by our patients. The latter consists of a main base unit container (referred to as a “dewar”) that stores oxygen in its liquid state at -170°C and a small portable oxygen-delivery system, that the patient can refill from the stationary dewar prior to ambulation. This portable option has several advantages, it is small in size and has a larger capacity, and therefore more cost-effective for patients with high daily activity.18 If patients are also on LTOT, they can use the stationary dewar in the same way as the gaseous oxygen container or concentrator to receive oxygen during rest. In this way, liquid oxygen can became the only source of supplemental oxygen, used for both purposes.

Our results showed that patients had a positive acute response to oxygen, assessed by the 6MWT, validating OA prescriptions.7 Nevertheless, the subjective sensation of dyspnea assessed by Borg scale, did not improve, at least during the in-laboratory test. Previous studies are contradictory with regard to the benefits of acute oxygen supply for exercise tolerance and the associated dyspnea. A systematic review of 31 randomized controlled trials demonstrated that AO was effective in increasing exercise capacity and reducing breathlessness in patients with exercise desaturation.19 More recently, Héraud et al concluded that oxygen supply does not systematically improve exercise tolerance in all patients. In about one third a negative response with a paradoxical increase in breathing frequency and dyspnea was observed.20

Concerning the long-term benefits in QoL and time spent in outdoor activities our results do not clearly support the prescription of AO. No improvements were noticed in QoL score values, except for emotional role (32.4% vs. 48.6%, p=0.032), and patients spent fewer hours per day outdoors at the time of the study than before AO treatment. Reported side-effects like embarrassment and device weight could account for some of the inhibition to get outside, but the disease aggravation (a proportion of 35% patients described worsening of global health state) could also be responsible. The failure of AO to increase activity or improve QoL has been previously reported9,21–23. A recent randomized trial of domiciliary AO (cylinder air versus cylinder oxygen) in patients with COPD and dyspnea but without hypoxemia also showed no benefits for either groups in terms of dyspnea, QoL, functional status or cylinder utilization. Statistically significant but clinically small improvements in dyspnea and depression were observed in the study group as a whole over the 12 weeks of the study, probably representing a placebo benefit.24 In contrast, Eaton et al, in a study of 41 COPD patients with exercise desaturation reported a statistically significant, albeit clinically small, improvement in QoL, anxiety and depression.25,26

There are some positive aspects related to QoL. In spite of the complex interplay of functional and psychosocial disability that affects these patients, emotional role was improved, and 46.0% referred to a subjective improvement in their health global state. Patient perspectives about the use of AO should be taken into account and it is interesting that 4 patients said that they would never leave their home without oxygen, although this point was not included in the questionnaire. Their experience may be attributed to a learned oxygen-related decrease in exertional dyspnea in daily activities, despite dyspnea relief was not sensed in 6MWT.9 As suggested by other studies, a placebo effect cannot be excluded.27

The number of hours of AO useper day (4.1h) was similar to the hours spent outside the home (3.5h), and was quite high. It is remarkable that patients left the house almost every day, despite their respiratory morbidity. Other studies have reported a lack of use of AO,28 or its inappropriate use at rest,29 which did not seem to occur with our patients. The non-compliance rate was low and all of the compliant patients, with 3 exceptions, used AO more than 1h/day, which can be accepted as a reasonable use.30

Pépin J et al reported a more effective oxygen use with the duration of the treatment29, which the authors suggested could be due to the deterioration in respiratory function over time but could also be related to the effects of technical and medical education of the patient. Vergeret et al29 have also shown the association between good compliance and the severity of the patient's condition. We consider that our patients did in fact have a high level of compliance (83.8%), for various reasons, including severity of the disease (78.4% were on LTOT), more effective use with time (mean time of AO prescription of 14.7 months) and more repeated education from health-care professionals. However, since compliance was assessed in a subjective fashion, there is the possibility that patient may have overestimated their use of AO.28,29

No significant differences appear to exist in the efficacy of AO (hours of use and compliance) between patients on LTOT compared with AO alone. In other studies AO prescription seemed to lead to a more effective use of LTOT29 and increased compliance,28 but the number of patients in our study does not allow us to take firm conclusions on this.

One limitation of our work was that some Borg scale data were not available, which has conditioned the identification of significant responses to AO. Taking into account previous studies reporting conflicting results about dyspnea relief29 we hypothesized that patients with the best improvements in Borg dyspnea scale could be more prone to use and benefit from AO, but this could not be proved in our study. Used oxygen flow rates were not evaluated, which could also have interfered with the response to AO, since low flow rates may provide inadequate relief of exercise dyspnea.23 The type of protocol used to assess the exercise performance can explain some of the variability in reported benefits of supplemental oxygen.8 Davidson et al reported that the endurance walking test was equally sensitive as a more formal assessment on a cycle ergometer and more reliable than the commonly used 6MWT. All these insights must be considered in future studies.

Time spent away from home and QoL baseline evaluation were collected retrospectively, with some patients having to remember things from up to 4 years earlier, which may have led to recall bias. A prospective study may overcome such limitations and in addition may explore the effect of time and education on AO use.

As far as we know, this is the first study in our country to evaluate the use and efficacy of AO. Since AO therapy does have its side-effects and costs, this information is of prime importance to patients, clinicians and health-care policy programs, as is the ability to distinguish patients who would benefit from this treatment. We observed that the acute improvement in 6MWT parameters did not help to predict the long-term benefits related to the enhancement of outdoor activities and QoL, as we would have expected. Desaturation in 6MWT also does not seem to be a reliable criterion for selecting the patients who should be prescribed AO. As well as acute response to oxygen, other factors like improved Borg dyspnea scale, the severity of disease, patient motivation or education, could influence the use and benefits of AO. In this way, a detailed assessment of individual patient response with a standardized follow-up with more frequent medical and technical monitoring, might achieve evidence based AO benefits. Our results point to the need of broader and more detailed studies evaluating these parameters and including more accurate assessment of patients’ activities, AO use and efficacy.

The authors have no conflicts of interest to declare.