Article Text

Download PDFPDF

Home versus outpatient pulmonary rehabilitation in COPD: a propensity-matched cohort study
  1. Claire Marie Nolan1,2,
  2. Djeya Kaliaraju2,
  3. Sarah Elizabeth Jones1,3,
  4. Suhani Patel1,
  5. Ruth Barker1,3,
  6. Jessica A Walsh1,2,
  7. Stephanie Wynne1,
  8. William Man1,2,3
  1. 1 Harefield Pulmonary Rehabilitation and Muscle Research Laboratory, Royal Brompton and Harefield NHS Foundation Trust, London, UK
  2. 2 Harefield Pulmonary Rehabilitation Unit, Royal Brompton and Harefield NHS Foundation Trust, Harefield, UK
  3. 3 National Heart and Lung Institute, Imperial College London, London, UK
  1. Correspondence to Dr Claire Marie Nolan, Harefield Pulmonary Rehabilitation and Muscle Research Laboratory, Royal Brompton and Harefield NHS Foundation Trust, London UB9 6JH, UK; c.nolan{at}


Home-based exercise has been proposed as an equivalent treatment strategy to supervised outpatient pulmonary rehabilitation (PR), but it is not known whether its implementation into clinical practice produces similar benefits to those observed in trials. We compared the real-world responses of 154 patients with COPD undergoing home-based exercise with a matched group attending supervised PR. We observed smaller improvements in exercise capacity with home-based exercise compared with PR, but similar improvements in quality of life. We propose that supervised PR remains the standard of care, with home-based exercise a less effective alternative for those unable to attend PR.

  • chronic obstructive pulmonary disease
  • home rehabilitation
  • pulmonary rehabilitation

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Pulmonary rehabilitation (PR) is an integral component of the management of COPD.1 Although the traditional model of PR typically comprises outpatient supervised exercise and multidisciplinary education,1 there are well-recognised problems with uptake and adherence. Consequently, there has been growing interest in alternative PR models with recent trials comparing home-based exercise to traditional PR.2–5 However, potential limitations of these trials include selective populations,4 lack of patient equipoise2 3 and lower than anticipated benefits in the PR arms,2–5 making non-inferiority analysis difficult to interpret. Furthermore, it is not known whether implementation of home-based exercise into clinical practice produces similar benefits to those observed in controlled trials.

The aim of our study was to observe the real-world responses of patients choosing to undergo home-based exercise to a matched control group choosing to undergo PR, in terms of exercise capacity, dyspnoea, health-related quality of life and programme completion.


Participants were patients with a physician diagnosis of COPD, who were consecutively referred to Harefield PR Unit between 2012 and 2015. Those with comorbidities that would make exercise unsafe were excluded (eg, unstable angina, severe aortic stenosis, abdominal aortic aneurysm >5.5 cm, uncontrolled cardiac arrhythmias). The participants provided informed consent.

All patients underwent a face-to-face assessment and were allowed a free choice of either PR or home-based exercise. To balance baseline characteristics, we used 1:1 propensity score matching (probability model)6 to match patients undergoing home-based exercise to those attending PR, accounting for age, gender, FEV1 %predicted, Medical Research Council Dyspnoea score, body mass index, incremental shuttle walk (ISW) distance and Chronic Respiratory Questionnaire (CRQ) total score. Gender was treated as a categorical variable and the others as continuous variables.

Home-based exercise was an 8-week structured programme with weekly telephone follow-up. It included individually tailored exercise which the patient was encouraged to perform at least 3 days per week, as well as educational material. The type of aerobic exercise was chosen by the patient and usually involved walking. Patients were encouraged to exercise to achieve a Borg Dyspnoea Scale score of 3–4 and progression included increasing the time to achieve 30 min of aerobic exercise three times per week. No specialist equipment was provided but upper and lower limb training was prescribed based on the American College of Sports Medicine endurance resistance training guidelines7 with resistance provided by body weight and home-made weights, for example, water bottles. Further information is provided in the online supplementary file. PR was an 8-week supervised outpatient programme delivered in a gymnasium or community hall according to the British Thoracic Society Quality Standards for PR.1 The programme comprised two supervised sessions per week and at least one unsupervised home exercise session was encouraged per week. Details of the aerobic and resistance training programme have been described in detail elsewhere8 as well as in the online supplementary file.

Supplemental material

The primary outcome measure was exercise capacity (ISW). Secondary outcome measures were dyspnoea (CRQ-dyspnoea domain), health-related quality of life (CRQ-total score) and programme completion (participation in ≥50% of weekly telephone calls or PR sessions and attendance at the postintervention assessment). Following PR, patients completed a Global Rating of Change Questionnaire: How do you feel your condition has changed after rehabilitation?. Responses were rated on a five-point Likert scale ranging from ‘much better to much worse.

A completer analysis was performed. For continuous data, within-group and between-group differences were compared using paired samples t-test and independent samples t-test, respectively. For categorical data, between-group differences were compared using Pearson’s χ2 test or Pearson’s χ2 test for trend.


In total, 154 patients (10% of eligible patients) chose home-based exercise and the baseline characteristics of the two groups are presented in table 1.

Table 1

Baseline characteristics (n=308)

Following PR, both groups achieved significant improvements in exercise capacity, but the home-based group demonstrated smaller increases (table 2). Furthermore, 54% of the PR group achieved the minimum clinically important difference of the ISW,9 compared with 34% in the home-based group; p=0.008. Both programmes produced clinically and statistically significant improvements in the CRQ with no significant between-group differences (table 2).

Table 2

Response to the intervention

Regarding self-reported perception of the intervention effect, more positive responses favoured PR: 53% reported feeling ‘much better’ in contrast to 37% of the home-based group; p=0.03. Further information is provided in the online supplementary file. Nonetheless, 83% of the home-based group reported feeling ‘much’ or ‘a little better’.

The programme completion rates were low in both groups (PR: 64%, home: 56%; p=0.16). The reasons for non-completion were broadly similar across the groups, with illness cited as the most common reason (PR: 39%, home: 34%).


To our knowledge, this is the first study to compare the real-world responses of patients with COPD choosing to undergo PR or home-based exercise. A distinct finding was the significantly greater mean improvements in exercise capacity with PR (59 m) compared with home-based exercise (29 m), with the former comparable to data reported in national PR data sets.10 In contrast, recent trials comparing home-based exercise with PR have demonstrated lower than expected improvements in exercise outcomes in the PR arms,2 5 which may have been driven by participant bias or design. For example, Bourne et al, 4 reported the effects of home-based exercise conducted through an online platform. Although the comparator arm was supervised and took place in an outpatient setting, the participants did not undergo traditional PR but followed a programme of 10 exercise stations that was identical to that in the online programme, with minimal aerobic exercise.

A potential advantage of our study was that we compared patients who chose to undergo home-based exercise with a matched group of patients who chose to attend PR. As such, our results may be generalisable and were not affected by a lack of equipoise. In contrast, a limitation of previous studies is selective trial populations, where significant proportions of eligible participants were not randomised as they wanted to participate in traditional PR.2 3 Furthermore, it is plausible that these trials may have lacked equipoise; for example, the majority of participants questioned by Horton et al , ‘would have chosen the home programme 3. In contrast, out of the 1593 patients eligible for our study, only 154 (10%) chose home-based exercise. This may have reflected referral bias or bias of the clinical team when offering treatment options, but we suspect that patients choosing home-based exercise were primarily those who did not wish to exercise in a group setting, or those with mobility or transport difficulties who were unable to travel to PR sessions.

A novel finding from our study is that the majority of patients in the home-based group (83%) reported feeling ‘much’ or ‘a little better’ following intervention. We also observed significant improvements in health-related quality of life following home-based exercise, corroborating recent meta-analysis data.11 However, we only evaluated the short-term effects and exploration of the long-term impact may be relevant.

In conclusion, traditional PR remains the standard of care and first-line option, with home-based exercise a less effective alternative for patients with COPD unable to attend PR.


We would like to express our gratitude to the patients who participated in this research and to the staff of Harefield Pulmonary Rehabilitation Unit for their role in delivering the interventions and help with data collection.



  • Contributors CMN and WM contributed to the concept and design of study and drafted the manuscript. CMN, DK, SEJ, SP, RB, JAW and SW contributed to the acquisition of data. CMN, DK and WM contributed to the analysis of data. All authors contributed to the revision of manuscript critically for important intellectual content and approval of the final manuscript.

  • Funding This study was funded by a project grant from the NIHR Collaboration for Leadership in Applied Health Research and Care Northwest London (CLAHRC for Northwest London). CMN, JAW and SCW are supported by an NIHR Research for Patient Benefit Grant (PB-PG-0816-20022); SEJ is supported by a NIHR Doctoral Research Fellowship (DRF-2015-08-004); SP is supported by a British Lung Foundation IPF Project Grant (IPFPG17-15) and REB is supported by a NIHR Clinical Doctoral Research Fellowship (CDRF-2017-03-018).

  • Competing interests WM reports grants from National Institute for Health Research during the conduct of the study; grants from Pfizer, non-financial support from GSK, personal fees from Mundipharma, personal fees from Novartis, outside the submitted work.

  • Patient consent for publication Not required.

  • Ethics approval The study was approved by London-Camberwell St Giles Research Ethics Committee (11/LO/1780).

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles

  • Airwaves
    The Triumvirate