Abstract
Ambulatory oxygen may be cost-effective in improving quality of life in fibrotic lung disease. To be more conclusive, we need to understand societal willingness to pay for quality of life improvements and whether improvements are sustained. http://bit.ly/2pAiBJi
To the Editor:
Fibrotic interstitial lung diseases (ILDs) are chronic and often progressive conditions resulting in substantial impact on morbidity, health-related quality of life (HRQoL), and health system costs. Ambulatory oxygen (AO) used during routine daily activities could lead to improved exercise performance, reduced symptoms and improved mobility in daily life. A UK prospective, multicentre, mixed method, randomised controlled crossover trial in patients with fibrotic ILD (AmbOx trial: NCT02286063), the first study on AO effects in daily life, reported improved HRQoL after 2 weeks of AO compared to no intervention, when measured by the King's Brief ILD (K-BILD) questionnaire [1–3]. Although AO is used in ILD, evidence supporting its health-economic impact is absent. Here, we evaluate the cost-effectiveness of AO in patients with ILD, using data collected alongside the AmbOx Trial.
AmbOx included adults with fibrotic ILD with isolated exertional hypoxia [1, 2]. Participants (mean age 67.9 years; 31% female) were randomised to receive either AO during routine activities of daily living (n=41) or no intervention (n=43) first. After a two-week period, the groups were crossed over to the alternative. AO for 2 weeks was associated with a significant improvement in total K-BILD score compared to no oxygen (mean difference adjusted for treatment order 3.7, 95% CI 1.8 to 5.6).
For this economic evaluation, data were analysed on a complete case basis (n=74; of the 76 participants completing AmbOx, two had missing K-BILD scores). We estimated resource use and costs incurred from a National Health Service perspective. Costs of AO were based on the number of cylinders used (median 2.75 cylinders per week, range 0–14) [1, 2] and assumed use of one nasal cannula per participant [4, 5]. Rental prices for oxygen cylinders (GBP 0.25 per day), refills (GBP 10.56 per refill) and delivery (GBP 16.90 per delivery) were not available from UK suppliers (commercial in confidence) and so were assumed based on online information for an Australian medical gas cylinder company [6]. As AO was not anticipated to have any effect on disease progression in the short-term, costs related to any unplanned health professional visits or hospital admissions were not included. The mean cost for AO for 2 weeks was estimated to be GBP 91.02 (95% CI GBP 77.83–104.21) per participant. Costs are expressed in GBP at 2017 rates, after conversion using AUD 1 equivalent to GBP 0.608.
We estimated the incremental cost per unit improvement in total K-BILD score, the primary trial outcome [1]. A one-point improvement in K-BILD score over a 2-week period was estimated to cost an additional GBP 25.21 (bootstrapped 95% CI GBP 15.21–69.48). Sensitivity analyses using the intention to treat cohort (with multiple imputation to adjust for missing values) gave a similar incremental cost of GBP 27.38 (95% CI GBP 15.68–86.77). Likewise, the estimated cost-effectiveness was not substantially impacted if the costs of the intervention were changed by up to ±80% of that assumed in the primary analysis (mean estimated cost per unit improvement in K-BILD GBP 14.42–46.72 over a 2-week period).
We also undertook post hoc analyses to estimate the cost-effectiveness of AO in providing a benefit based on the number of participants who reported a minimum of an 8- or 4-unit improvement in total K-BILD score (these values have been reported as the minimal clinically important difference [3, 7]), or an improvement in the patient-reported global assessment of change in breathlessness and walking ability at the end of each 2-week treatment period, one of the pre-defined major secondary trial outcomes [2]. An 8-unit improvement in K-BILD score was reported for 13 (17.6%, number needed to treat (NNT) 5.85) and a 4-unit improvement for 27 (36.5%, NNT 2.81) of participants [2]. Given an incremental cost of GBP 91.02 (95% CI GBP 77.83–104.21) per person treated for 2 weeks, it is estimated to cost an additional GBP 532.47 (95% CI GBP 455.31–609.63) to achieve one additional “responder” of at least 8 units or GBP 255.77 (95% CI GBP 218.70–292.83) to achieve one additional “responder” of at least 4 units on the K-BILD, over a 2-week period.
Data on global assessment of change was available for 76 participants. More participants perceived an improvement in their walking ability (51, 67.1%) or breathlessness (52, 68.4%) after receiving oxygen as compared to no oxygen (1, 1.3% in each case). This corresponds to an NNT of 0.02 for one additional patient to perceive an improvement in walking ability or breathlessness, and an estimated incremental cost of GBP 1.82 (95% CI GBP 1.56–2.08) over 2 weeks to achieve one additional person perceiving an improvement.
This economic evaluation is the first to examine the cost-effectiveness of AO in fibrotic ILD, often a devastating and progressive group of diseases with substantial impact on patients' HRQoL and limited treatment options. Whether or not AO is considered to be of acceptable cost-effectiveness depends on society's (unknown) willingness to pay for an improvement in K-BILD score or the global assessment measures. The analysis suggests a much lower number needed to treat (and therefore cost for one additional responder) to obtain a perceived improvement in breathlessness or walking ability according to the global assessment measures, than to achieve an improvement in HRQoL according to the K-BILD. This apparent responsiveness for the global assessment of change might be related to the comparative bluntness of the single item measure and the possible impact of non-blinding on the participants' self-reported perception of improvement.
There are limitations associated with our study. Costs for oxygen cylinders were assumed, although sensitivity analysis suggests this assumption did not substantially impact findings. The AmbOx trial was open label, had a relatively small sample size and a short duration of oxygen use (2 weeks), making it challenging to extrapolate the costs or benefits to ILD patients in the longer term. It is possible that people using AO would be more likely to access long-term oxygen therapy as they become accustomed to the idea of using oxygen, with yet unknown benefits and drawbacks. This analysis should be considered as indicative only, until data from a larger study with longer follow-up is available to support more conclusive assertions. Finally, lightweight oxygen cylinders were used for all patients to standardise the intervention and to allow for higher oxygen flow rates in patients with more severe exertional hypoxia [2], but further studies are needed to assess whether portable oxygen concentrators may be more beneficial for the subset with milder ILD.
We had intended to evaluate cost per quality-adjusted life year (QALY) by deriving QALYs using utility values from the existing medical literature for health states described by the K-BILD or the St George's Respiratory Questionnaire, another secondary outcome measure [1]. However, appropriate HRQoL data in ILD patients on which to base an estimate of QALY gain are not available. Further research to derive a preference-based utility index for the K-BILD instrument is required to support accurate assessment of the benefits of treatment targeting HRQoL in ILD. Moreover, future trials should consider collecting a preference-based measure of health (such as the EQ-5D-5L) as an outcome [8, 9].
Nevertheless, despite these limitations, this study is the first to provide an indication of the cost-effectiveness of AO for improving HRQoL outcomes in ILD. Further evidence for the long-term effectiveness of AO, conversion of HRQoL outcomes in ILD to QALYs, and societal willingness to pay for HRQoL improvements are required to ensure the benefits of AO are accurately captured in economic evaluation and can be interpreted in resource allocation decisions.
Shareable PDF
Supplementary Material
This one-page PDF can be shared freely online.
Shareable PDF ERJ-01157-2019.Shareable
Acknowledgements
The study was supported by the NIHR Biomedical Research Unit at Royal Brompton and Harefield NHS Foundation Trust and Imperial College London. Members of the DMC: Joanna Porter, Chair; Sunita Rehal, independent statistician; John S. Wort. Members of the TSC: Stephen Durham, Vicky Tsipouri, Ira Jakupovic, Elisabetta Renzoni, Dina Visca, Voon Ong, Arnab Datta. We thank John Tayu Lee for providing input into the design of the economic evaluation component of the study. We also thank the RBH Research and Finance Office for their help in organising and supporting the infrastructure needed for the study, in particular Patrick Petterson and Alla Kashif. The AmbOx study is funded in full by the Research for Patient Benefit Programme National Institute for Health Research (Ref: PB-PG-0712-28073). Additional infrastructure support for the study is being provided by the Royal Brompton NIHR-funded Biomedical Research Unit.
Footnotes
Support statement: This manuscript presents independent research funded by the National Institute for Health Research (NIHR) under its Research for Patient Benefit (RfPB) Programme (Grant Reference Number PB-PG-0712-28073). The funder has no role in the trial design, collection, management, analysis or interpretation of data, writing of reports or submission for publication. Jennifer Whitty is supported by the National Institute for Health Research (NIHR) Collaboration for Leadership in Applied Health Research and Care (CLAHRC) East of England. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health. Funding information for this article has been deposited with the Crossref Funder Registry.
Conflict of interest: J.A. Whitty reports grants from Research for Patient Benefit Programme National Institute for Health Research (ref: PB-PG-0712-28073), during the conduct of the study.
Conflict of interest: J. Rankin reports grants from Research for Patient Benefit Programme National Institute for Health Research (ref: PB-PG-0712-28073), during the conduct of the study.
Conflict of interest: D. Visca reports grants from Research for Patient Benefit Programme National Institute for Health Research (ref: PB-PG-0712-28073), during the conduct of the study.
Conflict of interest: V. Tsipouri reports grants from Research for Patient Benefit Programme National Institute for Health Research (ref: PB-PG-0712-28073), during the conduct of the study.
Conflict of interest: L. Mori reports grants from Research for Patient Benefit Programme National Institute for Health Research (ref: PB-PG-0712-28073), during the conduct of the study.
Conflict of interest: L. Spencer reports grants from Research for Patient Benefit Programme National Institute for Health Research (ref: PB-PG-0712-28073), during the conduct of the study.
Conflict of interest: H. Adamali has nothing to disclose.
Conflict of interest: T.M. Maher has, via his institution, received industry-academic funding from GlaxoSmithKline R and D and UCB, and has received consultancy or speakers fees from Apellis, AstraZeneca, Bayer, Biogen Idec, Boehringer Ingelheim, Galapagos, GlaxoSmithKline R and D, Indalo, Pliant, ProMetic, Roche, Samumed and UCB.
Conflict of interest: N.S. Hopkinson reports grants from Research for Patient Benefit Programme National Institute for Health Research (ref: PB-PG-0712-28073), during the conduct of the study.
Conflict of interest: S.S. Birring reports grants from Research for Patient Benefit Programme National Institute for Health Research (ref: PB-PG-0712-28073), during the conduct of the study; fees (to Kings College Hospital) for using KBILD from Boehringer Ingleheim, Roche, Galapogos and Novartis, outside the submitted work.
Conflict of interest: M. Farquhar reports grants from Research for Patient Benefit Programme National Institute for Health Research (ref: PB-PG-0712-28073), during the conduct of the study.
Conflict of interest: A.U. Wells reports grants from Research for Patient Benefit Programme National Institute for Health Research (ref: PB-PG-0712-28073), during the conduct of the study; personal fees for lectures and advisory board work from Boehringer Ingelheim, Roche and Bayer, outside the submitted work.
Conflict of interest: P. Sestini reports grants from Research for Patient Benefit Programme National Institute for Health Research (Ref: PB-PG-0712-28073), and Royal Brompton and Harefield NHS Foundation and Trust, during the conduct of the study.
Conflict of interest: E.A. Renzoni reports grants from Research for Patient Benefit Programme National Institute for Health Research (ref: PB-PG-0712-28073), during the conduct of the study; personal fees for lectures and advisory board work from Boehringer Ingelheim and Roche, personal fees for lectures from Mundipharma, outside the submitted work.
- Received January 25, 2019.
- Accepted October 14, 2019.
- Copyright ©ERS 2020