Design

A randomised controlled superiority trial, with blinding of participants, clinicians and researchers, in 260 people with fILD who desaturate during walking. The trial will be conducted in Australia and Sweden, with trial sites detailed on the study registration at ClinicalTrials.gov. Participants will be randomised 1:1 into two groups: group 1: ambulatory oxygen therapy using a POC (oxygen group), group 2: sham therapy using an identical POC (air group). The allocated treatment will be delivered for 6 months. We have previously demonstrated the feasibility of the trial methodology in a pilot study.21 Figure 1 shows the participant flow through the trial. The overview of the study procedures follows the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) checklist22 (see online supplemental file 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Study flow. 6MWT, 6-minute walk test; LTOT, long-term oxygen therapy; SpO₂, oxyhaemoglobin saturation.

Participants

People with fILD who receive their care at the study sites will be invited to participate. Patients will be eligible for inclusion if they are (1) aged 18 years and over; (2) have a physician-confirmed diagnosis of fILD, such as IPF, connective tissue disease-associated ILD, fibrotic hypersensitivity pneumonitis, idiopathic non-specific interstitial pneumonia, unclassifiable idiopathic interstitial pneumonia, environmental/occupational lung disease or sarcoidosis, with features of diffuse fILD of >10% extent on high-resolution CT, with ILD being the predominant pathological process23; (3) have had stable pharmacotherapy over the last 3 months; and (4) exhibit exertional desaturation, defined as oxyhaemoglobin saturation (SpO₂) ≤88% for at least 10 consecutive seconds during a 6-minute walk test (6MWT) performed on room air. Only fILD will be included as these conditions are often characterised by chronic progressive fibrosis, as opposed to other ILDs with differing pathophysiology (eg, inflammatory or granulomatous) where different mechanisms may underlie exertional desaturation.

Participants will be excluded if they: (1) are currently using or eligible for long-term oxygen therapy (LTOT), with eligibility defined as arterial oxygen pressure ≤55 mm Hg at rest on room air, or 56–59 mm Hg with evidence of right heart failure10 as it is not ethical to withhold oxygen therapy in this group for whom it is strongly recommended24; (2) are current smokers, due to the risk of oxygen use near flames; (3) have predominantly obstructive lung disease, with forced expiratory ratio less than the lower limit of normal; (4) are pregnant; (5) are cognitively unable to consent; (6) are non-ambulant; (7) have been admitted to an acute care hospital within the last 30 days; or (8) if death or transplant is anticipated within the study period. Participants currently participating in pulmonary rehabilitation will not be enrolled, and participation in pulmonary rehabilitation during the 6-month trial period will be avoided where possible, as this may impact on both primary and secondary outcomes.

Recruitment

Potential participants will be identified by their treating healthcare team. If the participant is interested in obtaining more detailed information, they will be contacted by the trial coordinator or site coordinator, who is not in a dependent relationship with the patient and will provide further information. Patients will be informed that participation in the study is voluntary, their decision about participation will not affect their treatment or relationship with their healthcare team, their data will be held securely and they will not be identified in any study publications. Only patients who provide written, informed consent will undertake the study procedures outlined in this protocol.

Patient and public involvement

We interviewed patients with fILD and physicians who cared for them, in order to understand experiences and role of oxygen therapy.7 19 Patients emphasised the need for oxygen devices that were lighter and easier to use. Our subsequent work showed that patients preferred using a POC over traditional oxygen cylinders.16 In our feasibility trial we interviewed participants about their experiences of trial participation. Participants expressed positive experiences of the study and stated they would recommend such trial participation to others with ILD.21 These experiences underpinned the design of the protocol for the current trial.

Randomisation

Randomisation will occur following completion of the baseline assessment, including measures of physical activity. Participants will be randomly allocated to groups using a computer-generated, permuted block randomisation schedule with stratification for (1) desaturation during 6MWT (<80% vs ≥80%) as this is a powerful predictor of physical activity, HRQL and mortality,4 5 25 and (2) site of recruitment. Sequence generation will be performed by an individual independent of the research team and the allocation sequence will be concealed using a secure online randomisation service.

Interventions

All participants will be informed that the aim of using a POC is to assist them to be more active, with fewer symptoms. They will be encouraged to use the POC at all times when they are moving about, including walking at home or in the community, during exercise or during other activities. The POC should not be used when sitting still or sleeping. Written and verbal education will be provided. Participants will be encouraged to use their allocated POC during physical activity for the 6-month study period. Participants will be provided with a standard POC carry bag, worn over one shoulder, but will have the option to use a backpack if required. The method by which each participant chooses to carry their POC will be recorded. Informed by our pilot study, the Inogen One G3 HF POC will be used at its maximum pulse flow setting of 5 for both groups, as this delivered similar oxygen saturation during walking to a portable cylinder delivering 5 L/min of oxygen on continuous flow.16 No titration of flow rates will be performed, in order to maintain blinding of clinicians, researchers and participants.

All participants will be contacted monthly by telephone by a blinded investigator to encourage POC use and answer any questions. These calls will also collect adverse events, healthcare utilisation data for economic analyses and information on concurrent therapies, including changes to medications. It is likely that some participants will commence LTOT during the 6-month study period. This will occur if participants meet the usual LTOT eligibility criteria10 and it is recommended by their treating physician. On commencement of LTOT the participant will cease using the allocated POC. The number of participants in each group who commence LTOT will be recorded, outcome measures will be collected as per the trial protocol and data will be analysed according to allocated treatment group, as per intention-to-treat principles. Similarly, if a participant deteriorates during the 6-month study period then they may commence new/additional pharmacotherapies at their physician’s discretion. The number of participants in each group who commence new pharmacotherapies and their nature will be recorded, outcome measures will be collected as per the trial protocol and data will be analysed according to allocated treatment group, as per intention-to-treat principles. The hours of usage downloaded from POC flash memory will be evaluated every 3 months. Reason for cessation of therapy will be recorded where relevant (patient request, commencement of LTOT, other).

Blinding

Participants, clinicians and researchers will be blinded to group allocation. The Inogen One G3 HF POCs for ambulatory oxygen therapy and air groups will be identical in appearance, display, weight and operation, with the only difference being the gas delivered. The POCs will be coded by the distributor, who will not be involved in trial conduct. We successfully used this system to maintain blinding in a previous trial using gas cylinders in chronic obstructive pulmonary disease15 and our feasibility trial did not identify any safety issues in randomising participants with fILD and exertional desaturation to POCs delivering air.21 The intervention code will only be available to the randomisation centre. All participants will be advised against measuring oxygen saturation at home during the duration of the trial, as this does not represent usual clinical practice in any of the centres and may unblind the participants. At the conclusion of the trial, participants will be asked two questions to evaluate the success of blinding: (1) Which treatment do you think you were receiving, oxygen or air? and (2) Did you have a pulse oximeter at home over the last 6 months? If yes, how did you use it?

Outcome measures

Outcome measures will be collected at baseline (two visits, 1 week apart), 3 and 6 months following randomisation (table 1), by an assessor who is blinded to group allocation. Three months were selected as this is sufficient to achieve change in the primary outcome26 and 6 months will provide robust data for economic analyses.

The primary outcome is change in physical activity, measured by the number of steps per day. Steps per day is an objective measure of physical activity in people with fILD that has strong relationships to respiratory function, exercise capacity, exertional desaturation, HRQL and fatigue.5 12 Physical inactivity, defined as less than 3300 steps per day, is associated with poor survival (30% vs 70% over 3 years).5 Steps per day is responsive to changes following non-pharmacological interventions in chronic lung disease26 and the minimal important difference has been defined as 599 steps.27 Steps per day is a direct measure of how a patient functions in daily life and thus fulfils the criteria for meaningful endpoints in fILD clinical trials.28 Steps per day will be measured using the StepWatch activity monitor (SAM) (Modus Health, Washington DC, USA) which is reliable and valid in chronic lung disease.29 It accurately detects slow walking speeds and is sensitive to small changes in step rate.30 The SAM will be worn on the ankle continuously for 7 days (except for bathing) between the first and second baseline visits, and then for 7 days following the 3-month and 6-month assessments. Seven days of monitoring is required for optimum reliability.29

Secondary outcomes are change in functional exercise capacity, HRQL, breathlessness, fatigue, anxiety, depression, time spent in moderate to vigorous physical activity, sedentary time, oxygen saturation in daily life, and plasma markers of skeletal muscle metabolism, systemic inflammation and oxidative stress.

Functional exercise capacity will be measured with the 6-minute walk distance31 which is responsive to change with acute administration of oxygen13 and is a strong predictor of survival in fILD.32 The 6MWT will be performed according to international standards, including performance of two tests at each time point to control for the known learning effect, with the best distance recorded.31 All tests will be performed breathing room air so that a valid comparison can be made across time points,31 and to maintain blinding. HRQL will be measured using three instruments: the St George’s Respiratory Questionnaire, a disease specific HRQL measure that is valid and responsive in fILD33; the EQ-5D-5L, a validated generic quality of life measure which is used to derive health utilities for economic analyses; and the King’s Brief Interstitial Lung Disease Questionnaire, a validated disease-specific health status questionnaire.14 The Dyspnoea-12 will be used to capture both the physical and affective components of breathlessness and is a reliable and valid questionnaire in fILD.34 Fatigue will be measured with the Fatigue Severity Scale, a valid and sensitive questionnaire in fILD.35 Anxiety and depression will be evaluated using the Hospital Anxiety and Depression Scale, a validated and widely used tool for assessing psychological distress. Anxiety and depression are common in fILD and are increased in users of LTOT.36 The GeneActiv (GENEActiv, Cambridgeshire, UK) will measure time spent in moderate to vigorous physical activity and sedentary time. This wrist-worn, triaxial accelerometer has been validated in IPF.37 Seven days of monitoring are required to accurately capture all activity intensities.38 Oxygen saturation in daily life will be measured using a Nonin 3150 Wrist Oximeter. The wrist oximeter will be worn during waking hours on 2 consecutive weekdays, with the display turned off so that participants remain blinded to their oxygen saturation during POC use. Examination of plasma markers of skeletal muscle metabolism (xanthine, hypoxanthine); systemic inflammation (interleukin-6, tumour necrosis factor-a, C reactive protein) and oxidative stress (8-isoprostane, thiobarbiuric acid reactive substrates), as previously published,11 will be performed.

Health economic analyses

We will undertake a comparison of per-person costs, including direct (health system) and indirect (personal) healthcare costs, of ambulatory oxygen therapy compared with air. Direct costs will include staff time, consumables, communications and overheads. Intervention costs will include staff inputs by duration, type and resource use (including troubleshooting and support) and equipment (POC and consumables). Personal costs will include transportation, travel time and impact of the intervention on the economic activities of other household members. Health system costs will include visits to the general practitioner, specialist or emergency department, including any telemedicine visits; use of chronic disease services and hospitalisation. We will collect healthcare utilisation data from hospital records, Medicare Benefits Schedule and Pharmaceutical Benefits Scheme data (Australia) and the National Patient Registry of the Swedish Board of Health and Welfare, as well as directly from participants via monthly telephone calls. Sensitivity analyses will use different assumptions about personal healthcare costs across countries.

An incremental cost-effectiveness analysis will be undertaken to compare differences in costs with differences in: (a) quality-adjusted life years (QALYs): a single preference-based utility score will be derived from the EQ-5D-5L. This will be converted to QALYs on the assumption that the duration of each status is exactly one-half of the time between two measurement intervals39 ; (b) the number of hospital admissions per enrolled person in the 6-month follow-up period. The indicator will be the incremental cost of averting an additional hospitalisation.

Schedule of assessments

Outcomes will be obtained by a blinded assessor at baseline, 3 months and 6 months (table 1). At the baseline visit, an arterial blood gas will be performed in participants with resting SpO₂ <93% to exclude resting hypoxaemia, as this is an indication for LTOT. At each assessment two 6MWTs will be performed according to international standards with continuous pulse oximetry31 while the participant breathes room air. The nadir oxyhaemoglobin from the longest 6MWT will be used to determine eligibility at baseline.31 Questionnaires will be administered and blood for biomarkers will be obtained. The StepWatch and the GENEActiv activity monitors will be given to participants to wear over the following 7 days and the Nonin 3150 Wrist Oximeter will be worn on 2 consecutive weekdays. The monitors will be returned to the investigators by post. Spirometry is performed every 6 months in usual care, to document disease progression. To minimise patient burden, we will not repeat this test separately for the trial.

Safety and adverse events

Adverse events will be defined according to Good Clinical Practice guidelines. Adverse events of specific interest will be defined according to the criteria used in the recent Long-term Oxygen Treatment Trial: worsening of fILD (worsening of lung function, development of resting hypoxaemia); exacerbation of fILD; burns (from smoking while using a POC, using the POC around an open flame or equipment that sparks); nosebleed or dry nose; musculoskeletal injury from tripping on a POC; hospitalisation or death.40 Adverse events will be identified during monthly telephone calls and three monthly assessment visits, or by reports from the treating medical team. Participants who experience an adverse event will receive all necessary medical care from their local healthcare team.

Sample size

A total of 220 participants (110 per group) will provide 80% power to detect, at the two-sided 5% level, a clinically important difference between groups in the primary outcome of 599 steps per day.27 This assumes an SD of 1582 steps, based on physical activity data previously collected at our centre in 52 patients with fILD. Our previous trials had less than 10% attrition.15 41 Previous experience suggests that 5% of participants could start LTOT (and cease POC use) over 6 months. We will therefore randomise 260 participants to ensure that 220 participants complete the study.

Over 1000 patients with fILD are currently managed at our centres, with an additional 300 new patients seen each year. Approximately half of these patients exhibit exertional desaturation and are not using LTOT.42 Based on the rate of recruitment in our feasibility study,21 we anticipate recruiting the required sample of 260 participants over 3 years.

Analysis

Mean differences for continuous variables will be analysed using linear mixed models, controlling for baseline values as required. Generalised linear mixed models will be used for binary or count outcomes. Estimates will be presented with 95% CIs and two-sided p values reported. Results will be displayed graphically where it will illuminate. All data will be analysed by intention to treat, including all randomised participants in the groups to which they were allocated, regardless of adherence.

Data integrity and management

Hardcopy data collection forms will be stored in a locked filing cabinet within a locked office, and electronic data will be stored in a purpose-built online database (www.adeptrs.com), with encryption and password protection. The online database will be protected by encryption enabled at up to 256-bits and SSL (Secure Sockets Layer) certificate, and hosted on a dedicated SSL cluster. No identifying information will be stored in the online database or on hardcopy data forms. Electronic data for all sites will be accessible by the principal investigator and the trial coordinator. Site-specific investigators will only have access to data relating to their individual site. Information will be stored indefinitely, in accordance with Human Research Ethics Committees requirements for interventional studies.

Data monitoring

The Data Safety and Monitoring Committee (DSMB) will meet two times yearly, chaired by a respiratory physician who is independent of the study team and trial sites. The DSMB will include a biostatistician. The DSMB will report its findings to the trial steering committee, consisting of the chief investigators.

Ethics and dissemination

Ethical approval has been granted in Australia by Alfred Hospital Human Research Ethics Committee (HREC/18/Alfred/42) with governance approval at all Australian sites, and in Sweden (Lund Dnr: 2019-02963). The study will be conducted and reported according to the SPIRIT guidelines22 and the Consolidated Standards of Reporting Trials statement.43 Results will be published in peer-reviewed journals and presented at conferences. We will also disseminate our results to people with fILD through lay publications and seminars.